System and method for assessing disease burden and progression

Abstract

What is presented herein are systems and methods that provide semi-automated and / or automated analysis of medical image data, provide patient risk and / or disease pictures, determine and / or communicate values of measurements. The techniques described herein analyze medical image data to provide a snapshot of a patient's disease burden at a particular time, evaluate quantitative measurements, and / or analyze images taken over time to create a longitudinal dataset that provides pictures of how a patient's risk and / or disease evolves over time during surveillance and / or in response to treatment. Measurements calculated via the image analysis tools described herein may themselves be used as quantitative measures of disease burden and / or may be linked to clinical endpoints.

Description

Technical Field

[0001] (Cross - Reference to Related Applications) This application claims the benefit and priority of U.S. Provisional Application No. 63 / 350,211, filed on June 8, 2022, U.S. Provisional Application No. 63 / 458,031, filed on April 7, 2023, and U.S. Provisional Application No. 63 / 461,486, filed on April 24, 2023, each of which is incorporated herein by reference in its entirety.

[0002] The present invention generally relates to systems and methods for the generation, analysis, and / or presentation of medical image data. More specifically, in certain embodiments, the present invention relates to systems and methods for automated analysis of medical images for identifying and / or characterizing cancerous lesions and / or prognosis or risk for a subject.

Background Art

[0003] Nuclear medicine imaging involves the use of radiolabeled compounds, referred to as radiopharmaceuticals. Radiopharmaceuticals are administered to a patient and depend on the biophysical and / or biochemical properties of the tissues therein, such as those affected by the presence and / or state of a disease such as cancer, and thus accumulate in various regions within the body in a manner that indicates it. For example, a certain radiopharmaceutical accumulates in regions of abnormal bone formation associated with malignant bone lesions indicating metastasis following administration to a patient. Other radiopharmaceuticals can bind to specific receptors, enzymes, and proteins within the body that are modified during the progression of a disease. After administration to the patient, these molecules circulate in the blood until they find their intended targets. The bound radiopharmaceuticals remain at the disease site while the remainder of the agent exits the body.

[0004] Nuclear medicine imaging techniques capture images by detecting the radiation emitted from the radioactive portion of a radiopharmaceutical. The accumulated radiopharmaceutical serves as a beacon so that images depicting the location and concentration of the disease can be obtained using commonly available nuclear medicine modalities. Examples of nuclear medicine imaging modalities include bone scan imaging (also referred to as scintigraphy), single photon emission computed tomography (SPECT), and positron emission tomography (PET). Bone scan, SPECT, and PET imaging systems are found in most hospitals around the world. The selection of a particular imaging modality depends on and / or is determined by the particular radiopharmaceutical used. For example, technetium 99m ( 99m Tc) labeled compounds are compatible with bone scan imaging and SPECT imaging, while PET imaging often uses fluorine compounds labeled with 18F. The compound 99m Tc methylene diphosphonate ( 99m Tc MDP) is a commonly used radiopharmaceutical used for bone scan imaging to detect metastatic cancer. 99m Tc labeled 1404 and PyL TM (also referred to as [18F]DCFPyL) and other radiolabeled prostate specific membrane antigen (PSMA) targeting compounds can each be used in combination with SPECT and PET imaging, offering the potential for highly specific detection of prostate cancer.

[0005] Thus, nuclear medicine imaging is a valuable technique for providing physicians with information that can be used to determine the presence and extent of disease within a patient. Physicians can use this information to provide a treatment course recommended for the patient and to track the progression of the disease.

[0006] For example, an oncologist may use nuclear medicine images from a patient's examination as input for an assessment of whether the patient has a particular disease, such as prostate cancer, what stage of the disease is evident, (if applicable) what the recommended course of treatment would be, whether surgical intervention is necessary, and perhaps also a prognosis. The oncologist may use the radiologist's report in this assessment. The radiologist's report is a technical evaluation of the nuclear medicine images prepared by a radiologist for the physician who requested the imaging examination and includes, for example, the type of examination performed, clinical history, comparison between images, techniques used to perform the examination, the radiologist's opinion and findings, as well as the overall impression and recommendations that the radiologist may have based on the imaging examination results. The signed radiologist's report is sent to the physician who ordered the examination for review, and subsequent discussions between the physician and patient regarding the results and treatment recommendations follow.

[0007] Accordingly, the process involves steps of having a radiologist perform an imaging examination on a patient, analyzing the acquired images, creating a radiologist's report, automatically forwarding the report to the requesting physician, having the physician compile an assessment and treatment recommendations, and having the physician communicate the results, recommendations, and risks to the patient. The process may also involve steps of repeating the imaging examination due to inconclusive results or ordering additional tests based on the initial results. If the imaging examination indicates that the patient has a particular disease or condition (e.g., cancer), the physician discusses various treatment options, including surgery, as well as the risks of doing nothing, or adopting a watchful waiting or active surveillance approach instead of undergoing surgery.

[0008] Therefore, the process of reviewing and analyzing multiple patient images over time plays an important role in cancer diagnosis and treatment. Thus, there is a significant need for improved tools that facilitate and enhance the accuracy of image review and analysis for cancer diagnosis and treatment. Improving the toolkits thus utilized by physicians, radiologists, and other healthcare providers provides a significant improvement in standard of care and patient experience. SUMMARY OF THE INVENTION MEANS FOR SOLVING THE PROBLEM

[0009] What is presented herein are systems and methods that provide semi-automated and / or automated analysis of medical image data, provide values of metrics that provide photographs of patient risks and / or diseases, and determine and / or communicate the values of the metrics. The techniques described herein analyze medical image data and evaluate quantitative metrics that provide a snapshot of a patient's disease burden at a particular time and / or analyze images taken over time to create a longitudinal dataset that provides a photograph of how a patient's risk and / or disease evolves over time during surveillance and / or in response to treatment. The metrics calculated via the image analysis tools described herein may themselves be used as quantitative measures of disease burden and / or may be linked to clinical endpoints that seek to measure and / or stratify patient outcomes. Thus, the image analysis techniques of the present disclosure may be used to inform clinical decision-making, evaluate treatment effectiveness, and predict patient response.

[0010] In one embodiment, the value of a patient index that quantifies disease burden is calculated by analyzing a 3D nuclear medicine image of a subject to identify and quantify sub-regions, referred to as hot spots, that indicate the presence of underlying cancerous lesions. Various quantitative measurements can be calculated for individual hot spots to reflect the severity and / or size of the underlying lesion that they represent. These individual hot spot quantification measurements can then be aggregated to calculate values of various patient indices that provide a measure of the disease burden and / or risk for the subject as a whole and / or within specific tissue regions or lesion subclasses.

[0011] In one aspect, the present invention is a method for automatically processing a 3D image of a subject and determining the value of one or more patient indices (plural indices) that measure the (e.g., overall) disease burden and / or risk for the subject, the method comprising: (a) receiving, by a processor of a computing device, a 3D functional image of the subject obtained using a functional imaging modality; (b) partitioning, by the processor, a plurality of 3D hot spot volumes within the 3D functional image, each 3D hot spot volume corresponding to a local region of elevated intensity relative to its surroundings and representing a potential cancerous lesion within the subject, thereby obtaining a set of 3D hot spot volumes; (c) calculating, by the processor, for each particular one of one or more individual hot spot quantification measurements, a value of the particular individual hot spot quantification measurement for each individual 3D hot spot volume of the set; and (d) determining, by the processor, the value of one or more patient indices (plural indices), wherein at least a portion of each patient index is associated with one or more specific individual hot spot quantification measurements and is a function of at least a portion (e.g., substantially all, e.g., a particular subset) of the values of the one or more specific individual hot spot quantification measurements calculated with respect to the set of 3D hot spot volumes.

[0012] In one embodiment, at least one specific patient index of one or more patient index values is associated with a single specific individual hot spot quantification measurement value and calculated as a function (e.g., mean, median, mode, sum, etc.) of substantially all (e.g., all, e.g., excluding only statistical outliers) values of the specific individual hot spot quantification measurement values calculated with respect to a set of 3D hot spot volumes.

[0013] In one embodiment, a single specific individual hot spot quantification measurement value is an individual hot spot intensity measurement value that quantifies the intensity within a 3D hot spot volume (e.g., calculated with respect to an individual 3D hot spot volume as a function of the intensity of the voxels of the 3D hot spot volume).

[0014] In one embodiment, the individual hot spot intensity measurement value is the mean hot spot intensity (e.g., calculated with respect to an individual 3D hot spot volume as the mean value of the intensities of the voxels within the 3D hot spot volume).

[0015] In one embodiment, a specific patient index is calculated as the sum of substantially all values of the individual hot spot intensity measurement values calculated with respect to a set of 3D hot spot volumes.

[0016] In one embodiment, a single specific individual hot spot quantification measurement value is the lesion volume (e.g., calculated with respect to a specific 3D hot spot volume as the sum of the volumes of each individual voxel within the specific 3D hot spot volume).

[0017] In one embodiment, the value of a specific patient index is calculated as the sum of substantially all lesion volume values calculated with respect to a set of 3D hot spot volumes (e.g., such that the specific patient index value provides a measurement of the total lesion volume within the subject).

[0018] In one embodiment, a particular one of one or more overall patient indices (plural indices) is associated with two or more specific individual hot spot quantification measurements and is calculated as a function (e.g., weighted sum, weighted average, etc.) of substantially all values of two or more specific individual hot spot quantification measurements calculated with respect to a set of 3D hot spot volumes.

[0019] In one embodiment, two or more specific individual hot spot quantification measurements comprise (i) individual hot spot intensity measurements and (ii) lesion volume.

[0020] In one embodiment, an individual hot spot intensity measurement is an individual lesion index that maps a value of hot spot intensity to a value on a standardized scale.

[0021] In one embodiment, a particular patient index (value) is calculated as a sum value of intensity-weighted lesions (e.g., hot spots) volumes by, for each individual 3D hot spot volume of substantially all 3D hot spot volumes, weighting a value of lesion volume by a value of an individual hot spot intensity measurement (e.g., calculating a product of the lesion volume value and the value of the individual hot spot intensity measurement), thereby calculating a plurality of intensity-weighted lesion volumes, and calculating a sum value of substantially all intensity-weighted lesion volumes as the value of the particular patient index.

[0022] In one embodiment, one or more individual hot spot quantification measurements comprise one or more individual hot spot intensity measurements that quantify intensity within a 3D hot spot volume (e.g., calculated with respect to an individual 3D hot spot volume as a function of intensities of voxels of the 3D hot spot volume).

[0023] In one embodiment, one or more individual hot spot quantification measurement values comprise one or more members selected from the group consisting of an average hot spot intensity (e.g., calculated with respect to a particular 3D hot spot volume as the average value of the intensities of the voxels within the particular 3D hot spot volume), a maximum hot spot intensity (e.g., calculated with respect to a particular 3D hot spot volume as the maximum value of the intensities of the voxels within the particular 3D hot spot volume), and a median hot spot intensity (e.g., calculated with respect to a particular 3D hot spot volume as the median value of the intensities of the voxels within the 3D hot spot volume).

[0024] In one embodiment, one or more individual hot spot intensity measurement values comprise the peak intensity of the 3D hot spot volume [e.g., with respect to a particular 3D hot spot volume, the value of the peak intensity is calculated by: (i) identifying the maximum intensity voxel within the particular 3D hot spot volume; (ii) identifying the voxels within a sub-region centered on the maximum intensity voxel (e.g., comprising the voxels within a particular threshold distance thereof) and within the particular 3D hot spot; and (iii) calculating the average value of the intensities of the voxels within the sub-region as the corresponding peak intensity].

[0025] In one embodiment, one or more individual hot spot intensity measurement values comprise an individual lesion index that maps the value of the hot spot intensity to a value on a standardized scale.

[0026] In one embodiment, the method comprises: identifying, by a processor, in a 3D functional image, one or more 3D reference volumes, each corresponding to a specific reference tissue region; determining, by the processor, one or more reference intensity values, each associated with a specific 3D reference volume of the one or more 3D reference volumes and corresponding to a measured intensity value within the specific 3D reference volume; in step (c), for each 3D hot spot volume in the set, determining, by the processor, a corresponding value of a specific individual hot spot intensity measurement value (e.g., average hot spot intensity, median hot spot intensity, maximum hot spot intensity, etc.); and determining, by the processor, a corresponding value of an individual lesion index based on the specific individual hot spot intensity measurement value and the corresponding value of the one or more reference intensity values.

[0027] In one embodiment, the method comprises: mapping each of the one or more reference intensity values to a corresponding reference index value on a scale; for each 3D hot spot volume, using the reference intensity value and the corresponding reference index value to determine a corresponding value of an individual lesion index and interpolating, based on the corresponding value of the specific individual hot spot intensity measurement value, the corresponding individual lesion index value onto the scale.

[0028] In one embodiment, the reference tissue region comprises one or more members selected from the group consisting of the liver, aorta, and parotid gland.

[0029] In one embodiment, the first reference intensity value is a blood reference intensity value (i) associated with a reference volume corresponding to an aortic portion and (ii) mapped to a first reference index value, and the second reference intensity value is a liver reference intensity value (i) associated with a reference volume corresponding to the liver and (ii) mapped to a second reference index value, the second reference intensity value being greater than the first reference intensity value and the second reference index value being greater than the first reference index value.

[0030] In one embodiment, the reference intensity value comprises a maximum reference intensity value that is mapped to a maximum reference index value, and is a 3D hot spot volume for which the corresponding value of a particular individual hot spot intensity measurement exceeds the maximum reference intensity value, and the 3D hot spot volume is assigned an individual lesion index value equal to the maximum reference index value.

[0031] In one embodiment, the method includes, within a set of 3D hot spot volumes, identifying one or more subsets, each associated with a particular tissue region and / or lesion classification, and for each particular subset of the one or more subsets, calculating a corresponding value of one or more particular patient indices (plural) using the values of individual hot spot quantification measurements calculated for the 3D hot spot volumes within the particular subset.

[0032] In one embodiment, one or more subsets are associated with a particular one of one or more tissue regions, and the method includes, for each particular tissue region, identifying a subset of 3D hot spot volumes located within the volume of interest corresponding to the particular tissue region.

[0033] In one embodiment, one or more tissue regions comprise one or more members selected from the group consisting of a skeletal region, a lymphatic region, and a prostate region, comprising one or more bones of the subject.

[0034] In one embodiment, each of one or more subsets is associated with a particular one of one or more lesion subtypes [e.g., according to a lesion classification scheme (e.g., miTNM classification)], and the method includes, for each 3D hot spot volume, determining the corresponding lesion subtype and assigning the 3D hot spot volume to one or more subsets according to those corresponding lesion subtypes.

[0035] In one embodiment, the method uses at least a portion of the value(s) of one or more patient index(es) as an input to a prognostic model (e.g., a statistical model such as a regression, e.g., a classification model by which a patient is assigned to a particular class based on a comparison of one or more patient index values to one or more thresholds, e.g., a machine learning model that receives as input the value(s) of one or more patient index(es)) to generate, as an output, an expected value and / or a range (e.g., a class) (e.g., time, e.g., expressed in months such as expected survival rate, time to progression, time to radiological progression, etc.) that indicates a value with a high likelihood of a particular patient outcome.

[0036] In one embodiment, the method uses at least a portion of the value(s) of one or more patient index(es) as an input to a prediction model (e.g., a statistical model such as a regression, e.g., a classification model by which a patient is assigned to a particular class based on a comparison of one or more patient index values to one or more thresholds, e.g., a machine learning model that receives as input the value(s) of one or more patient index(es)) to generate, for each of one or more treatment options (e.g., abiraterone, enzalutamide, apalutamide, darolutamide, sipuleucel-T, Ra223, docetaxel, cabazitaxel, pembrolizumab, olaparib, rucaparib, 177 Lu-PSMA-617, etc.) and / or class of therapeutic agents [e.g., androgen biosynthesis inhibitors (e.g., abiraterone), androgen receptor inhibitors (e.g., enzalutamide, apalutamide, darolutamide), cellular immunotherapies (e.g., sipuleucel-T), internal radiation therapy treatments (Ra223), anti-neoplastic agents (e.g., docetaxel, cabazitaxel), immune checkpoint inhibitors (pembrolizumab), PARP inhibitors (e.g., olaparib, rucaparib), PSMA binders], an efficacy score, where the efficacy score for a particular treatment option and / or class of therapies indicates a prediction of whether a patient will benefit from a particular treatment and / or class of therapies.

[0037] In one embodiment, the method includes (e.g., automatically) generating a report [e.g., an electronic document, e.g., within a graphical user interface (e.g., for verification / approval by a user)] comprising at least a portion of the value(s) of one or more patient metrics.

[0038] In one embodiment, the method includes using one or more machine learning modules [e.g., one or more neural networks (e.g., one or more convolutional neural networks)] to perform one or more functions selected from the group consisting of: detecting a plurality of hot spots, wherein each of at least a portion of the plurality of 3D hot spot volumes corresponds to a particular detected hot spot and is created by partitioning the particular detected hot spot; partitioning at least a portion of the plurality of 3D hot spot volumes; and classifying (e.g., determining the likelihood that each 3D hot spot volume represents an underlying cancerous lesion) at least a portion of the 3D hot spot volumes.

[0039] In one embodiment, the 3D functional image comprises a PET or SPECT image acquired following administration of an agent to a subject. In one embodiment, the agent comprises a PSMA binder. In one embodiment, the agent comprises 18 F. In one embodiment, the agent comprises [18F]DCFPyL. In one embodiment, the agent comprises PSMA-11. In one embodiment, the agent comprises 99m Tc, 68 Ga, 177 Lu, 225 Ac, 111 In, 123 I, 124 I, and 131 I and comprises one or more members selected from the group consisting of.

[0040] In another aspect, the present invention is a method for automated analysis of a subject's time series of medical images [e.g., three-dimensional images, e.g., nuclear medicine images (e.g., bone scans (scintigraphy), PET, and / or SPECT), e.g., anatomical images (e.g., CT, X-ray, MRI), e.g., combined nuclear medicine and anatomical images (e.g., overlaid)], the method comprising: (a) receiving and / or accessing, by a processor of a computing device, the subject's time series of medical images; and (b) identifying, by the processor, a plurality of hot spots within each of the medical images, and determining, by the processor, one, two, or all three of the following: (i) a change in the number of identified lesions, (ii) a change in the overall volume of the identified lesions (e.g., a change in the sum of the volumes of each identified lesion), and (iii) a change in the PSMA (e.g., lesion index)-weighted total volume (e.g., the sum of the products of the lesion index and the lesion volume for all lesions within the region of interest) [e.g., the changes identified in step (b) are used to (1) identify a disease status [e.g., progression, regression, or no change], (2) make a treatment management decision [e.g., active surveillance, prostatectomy, anti-androgen therapy, prednisone, radiation, radiotherapy, radiopharmaceutical PSMA therapy, or chemotherapy], or (3) identify treatment effectiveness (e.g., if the subject has started or is continuing treatment with a drug or other therapy following an initial set of images in the time series of medical images)] [e.g., step (b) includes using a machine learning module / model].

[0041] In another aspect, the present invention is a method for analyzing a plurality of medical images of a subject (e.g., for assessing a disease state and / or progression within the subject), the method comprising: (a) receiving and / or accessing, by a processor of a computing device, a plurality of medical images of the subject, and obtaining, by the processor, a plurality of 3D hot spot maps, each corresponding to a particular medical image, and identifying one or more hot spots (e.g., representing potential underlying physical lesions within the subject) in the particular medical image; (b) for each particular one (medical image) of the plurality of medical images, determining, by the processor, using a machine learning module [e.g., a deep learning network (e.g., a convolutional neural network (CNN))], a corresponding 3D anatomical segmentation map that identifies a set of organ regions (e.g., representing one or more of soft tissue and / or bone structures within the subject, such as cervical vertebrae, thoracic vertebrae, lumbar vertebrae, left and right lumbar bones, sacrum, and coccyx, left and right ribs and scapulae, left and right thighs, skull, brain, and mandible) in the particular medical image, thereby generating a plurality of 3D anatomical segmentation maps; (c) determining, by the processor, using (i) the plurality of 3D hot spot maps and (ii) the plurality of 3D anatomical segmentation maps, an identification of one or more lesion correspondences, each of which (lesion correspondence) is determined (e.g., by the processor) to identify two or more corresponding hot spots in different medical images and represent the same underlying physical lesion within the subject; and (d) determining, by the processor, based on the plurality of 3D hot spot maps and the identification of one or more lesion correspondences, values of one or more measurements {e.g., one or more hot spot quantification measurements and / or changes therein [e.g., quantifying changes in properties such as volume, radiopharmaceutical uptake rate, shape, etc. of individual hot spots and / or the underlying physical lesions they represent (e.g., over time / between a plurality of medical images)], e.g., a patient index (e.g., measuring the overall disease burden and / or status and / or risk for the subject) and / or changes therein, e.g., a value for classifying the patient (e.g.,belonging to and / or having a particular disease state, progression, category, etc., e.g., a prognostic measure [e.g., indicating and / or quantifying the likelihood of one or more clinical outcomes (e.g., disease state, progression, likely survival rate, treatment effectiveness, and the like) (e.g., overall survival rate)], e.g., a predictive measure (e.g., indicating a predicted response to a therapy and / or other clinical outcome)}, and a step of determining, a method is targeted.

[0042] In certain embodiments, the plurality of medical images comprises one or more anatomical images (e.g., CT, X-ray, MRI, ultrasound, etc.).

[0043] In certain embodiments, the plurality of medical images comprises one or more nuclear medicine images [e.g., bone scan (scintigraphy) (e.g., obtained following administration of a radiopharmaceutical such as 99mTc-MDP, etc.), PET (e.g., obtained following administration of a radiopharmaceutical such as [18F]DCFPyL, [68Ga]PSMA-11, [18F]PSMA-1007, rhPSMA-7.3(18F), [18F]-JK-PSMA-7, etc.), or SPECT (e.g., obtained following administration of a radiopharmaceutical such as a 99mTc-labeled PSMA binder, etc.)].

[0044] In certain embodiments, the plurality of medical images comprises one or more composite images (e.g., overlaid / co-registered with each other, e.g., obtained with respect to the subject substantially simultaneously) (e.g., one or more PET / CT images), each of which comprises an anatomical and nuclear medicine pair.

[0045] In certain embodiments, the plurality of medical images is or comprises time-series medical images, and each medical image in the time series is associated with a different specific time and is obtained at that time.

[0046] In certain embodiments, the time series medical images comprise a first medical image obtained prior to administering a particular therapeutic agent (e.g., a PSMA binder (e.g., PSMA-617, e.g., PSMA I&T), e.g., a radiopharmaceutical, e.g., a radionuclide-labeled PSMA binder (e.g., 177Lu-PSMA-617, e.g., 177Lu-PSMA I&T)) (e.g., for one or more cycles) to a subject, and a second medical image obtained after administering the particular therapeutic agent (e.g., for one or more cycles) to the subject.

[0047] In certain embodiments, the method includes classifying the subject as a responder and / or non-responder to a particular therapeutic agent based on the value of one or more measurements determined in step (d).

[0048] In certain embodiments, step (a) includes generating each hot spot map by (e.g., automatically) partitioning at least a portion of the corresponding medical image (e.g., a sub-image thereof such as a nuclear medicine image) (e.g., using a second hot spot partitioning machine learning module [e.g., the hot spot partitioning machine module comprises a deep learning network (e.g., a convolutional neural network (CNN))]).

[0049] In certain embodiments, each hot spot map comprises one or more labels identifying one or more assigned anatomical regions and / or lesion subtypes (e.g., miTNM classification labels) for at least a portion of the hot spots identified therein.

[0050] In one embodiment, the plurality of hot spot maps comprises (i) a first hot spot map corresponding to a first medical image (e.g., identifying a first set of one or more hot spots therein), and (ii) a second hot spot map corresponding to a second medical image (e.g., identifying a second set of one or more hot spots therein), the plurality of 3D anatomical partitioning maps comprises (i) a first 3D anatomical partitioning map identifying a set of organ regions within the first medical image, and (ii) a second 3D anatomical partitioning map identifying a set of organ regions within the second medical image, and step (c) comprises using the first 3D anatomical partitioning map and the second 3D anatomical partitioning map to register (e.g., using the set of organ regions and / or one or more subsets thereof as landmarks within the first and second 3D anatomical partitioning maps to determine one or more registration fields (e.g., a full 3D registration field, e.g., point-to-point registration), and using one or more of the determined registration fields to co-register the first and second hot spot maps) (i) the first hot spot map and (ii) the second hot spot map.

[0051] In one embodiment, step (c) comprises determining, for two or more hot spots, each a member of a different hot spot map and identified in different medical images, one or more lesion correspondence metric values (e.g., volume overlap, e.g., centroid distance, e.g., lesion type match), and determining, based on the one or more lesion correspondence metric values, two or more hot spots of the group to represent the same underlying physical lesion, thereby including the two or more hot spots of the group within one of the one or more lesion correspondences.

[0052] In one embodiment, step (d) is as follows, namely, (i) a change in the number of identified lesions, (ii) a change in the overall volume of the identified lesions (e.g., a change in the sum value of the volumes of each identified lesion), and (iii) a change in the PSMA (e.g., lesion index) weighted total volume (e.g., the sum value of the product of the lesion index and the lesion volume for all lesions within the region of interest) [e.g., the changes identified in step (b) are used to (1) identify the disease status [e.g., progression, regression, or no change], (2) make treatment management decisions [e.g., active surveillance, prostatectomy, anti-androgen therapy, prednisone, radiation, radiotherapy, radiation PSMA therapy, or chemotherapy], or (3) identify treatment effectiveness (e.g., when the subject has started or is continuing treatment with a drug or other therapy following an initial set of medical images in the time-series medical images)] and includes the step of determining one, two, or all three of (i), (ii), and (iii) as described above.

[0053] In one embodiment, the method includes the step of determining the value of one or more prognostic measurements indicative of the disease state / progression and / or treatment (e.g., based on the value of one or more measurements in step (d), for example) [e.g., the step of determining the overall survival rate (OS) (e.g., the predicted number of months) expected for the subject].

[0054] In one embodiment, the method is based on the value of one or more measurements (e.g., a change in tumor volume, SUV mean , SUV max), as an output, generate an expected value and / or range (e.g., class) that indicates a value highly likely to be a specific patient outcome (e.g., time, e.g., expressed in months such as the expected survival rate, time to progression, time to radiological progression, etc.), and use it as an input to a prognostic model (e.g., a statistical model such as regression, e.g., a classification model by which a patient is assigned to a specific class based on the comparison of one or more patient index values with one or more thresholds, e.g., a machine learning model that receives as input one or more patient index values).

[0055] In certain embodiments, the method includes using, as an input, the value of one or more measured values (e.g., change in tumor volume, SUV mean SUV max ), PSMA score, number of new lesions, number of disappeared lesions, total number of lesions followed up), as an output, generate a classification (e.g., binary classification) indicating the patient response to treatment, and use it as an input to a response model (e.g., a statistical model such as regression, e.g., a classification model by which a patient is assigned to a specific class based on the comparison of one or more patient index values with one or more thresholds, e.g., a machine learning model that receives as input one or more patient index values).

[0056] In certain embodiments, the method includes using, as an input, the value of one or more measured values (e.g., change in tumor volume, SUV mean SUV max, PSMA score, number of new lesions, number of disappeared lesions, total number of lesions followed up), as output, one or more treatment options (e.g., abiraterone, enzalutamide, apalutamide, darolutamide, sipuleucel-T, Ra223, docetaxel, cabazitaxel, pembrolizumab, olaparib, rucaparib, 177Lu-PSMA-617, etc.) and / or classes of therapeutic agents [e.g., androgen biosynthesis inhibitors (e.g., abiraterone), androgen receptor inhibitors (e.g., enzalutamide, apalutamide, darolutamide), cellular immunotherapy (e.g., sipuleucel-T), internal radiotherapy treatment (Ra223), antineoplastic agents (e.g., docetaxel, cabazitaxel), immune checkpoint inhibitors (pembrolizumab), PARP inhibitors (e.g., olaparib, rucaparib), PSMA binders], for each, a step of using as input to a prediction model (e.g., a statistical model such as regression, e.g., a classification model by which a patient is assigned to a specific class based on the comparison of one or more patient index values with one or more thresholds, e.g., a machine learning model that receives as input one or more values of a patient index), and the efficacy score for a specific treatment option and / or therapeutic class indicates a prediction of whether a patient will benefit from a specific treatment and / or therapeutic class.

[0057] In another aspect, the present invention is a method for analyzing a plurality of medical images of a subject, comprising: (a) obtaining (e.g., receiving, accessing, and / or generating) a first 3D hot spot map for the subject by a processor of a computing device; (b) obtaining (e.g., receiving, accessing, and / or generating) a first 3D anatomical segmentation map associated with the first 3D hot spot map by the processor; (c) obtaining (e.g., receiving, accessing, and / or generating) a second 3D hot spot map for the subject by the processor; (d) obtaining (e.g., receiving, accessing, and / or generating) a second 3D anatomical segmentation map associated with the second 3D hot spot map by the processor; (e) determining a registration field (e.g., a full 3D registration field, e.g., point-wise registration) using and / or based on the first 3D anatomical segmentation map and the second 3D anatomical segmentation map by the processor; (f) registering the first 3D hot spot map and the second 3D hot spot map using the determined registration field by the processor, thereby generating a pair of co-registered 3D hot spot maps; (g) determining the identification of one or more lesion correspondences using the pair of co-registered 3D hot spot maps by the processor; and (h) storing and / or providing the identification of one or more lesion correspondences for display and / or further processing by the processor.

[0058] In another aspect, the present invention is a method for analyzing a plurality of medical images of a subject (e.g., evaluating a disease state and / or progression within the subject), the method comprising: (a) receiving and / or accessing, by a processor of a computing device, a plurality of medical images of the subject; (b) for each particular one (medical image) of the plurality of medical images, determining, by the processor, a corresponding 3D anatomical partitioning map that identifies a set of organ regions within the particular medical image [e.g., representing soft tissue and / or bone structures within the subject (e.g., one or more of cervical vertebrae, thoracic vertebrae, lumbar vertebrae, left and right lumbar bones, sacrum, and coccyx, left ribs and left scapula, right ribs and right scapula, left thigh, right thigh, skull, brain, and mandible)] using a machine learning module [e.g., a deep learning network (e.g., a convolutional neural network (CNN))], thereby generating a plurality of 3D anatomical partitioning maps; (c) determining, by the processor, using the plurality of 3D anatomical partitioning maps, one or more alignment fields (e.g., a full 3D alignment field, e.g., point-wise alignment), applying the one or more alignment fields, and aligning the plurality of medical images, thereby generating a plurality of aligned medical images; (d) for each particular one of the plurality of aligned medical images, determining, by the processor, a corresponding aligned 3D hot spot map that identifies one or more hot spots within the particular aligned medical image [e.g., representing a potential underlying physical lesion within the subject], thereby generating a plurality of aligned 3D hot spot maps; (e) determining, by the processor, using the plurality of 3D aligned hot spot maps, one or more lesion correspondences, each (lesion correspondence) being determined (e.g., by the processor) to identify two or more corresponding hot spots in different medical images and representing the same underlying physical lesion within the subject; (f) based on the plurality of 3D hot spot maps and the determination of one or more lesion correspondences, by the processor,A step of determining one or more measured values {e.g., one or more hot spot quantification measured values and / or changes therein [e.g., quantifying changes in properties such as (e.g., over time / between multiple medical images) individual hot spots and / or the volume of underlying physical lesions they represent, radiopharmaceutical uptake rate, shape, etc.], e.g., patient indices (e.g., measuring overall disease burden and / or status and / or risk for a subject) and / or changes therein, e.g., values for classifying a patient (e.g., belonging to and / or having a particular disease state, progression, category, etc.), e.g., prognostic measured values [e.g., indicating and / or quantifying one or more clinical outcomes (e.g., disease state, progression, likely survival rate, treatment effectiveness, and the like) (e.g., overall survival rate)], e.g., predictive measured values (e.g., indicating a predicted response to therapy and / or other clinical outcomes)} is targeted for a method.,

[0059] In another aspect, the present invention is a method for analyzing a plurality of medical images of a subject, comprising: (a) obtaining (e.g., receiving, accessing, and / or generating) a first 3D anatomical image (e.g., CT, X-ray, MRI, etc.) and a first 3D functional image [e.g., nuclear medicine image (e.g., PET, SPECT, etc.)] of the subject by a processor of a computing device; (b) obtaining (e.g., receiving, accessing, and / or generating) a second 3D anatomical image and a second 3D functional image of the subject by the processor; (c) obtaining (e.g., receiving, accessing, and / or generating) a first 3D anatomical segmentation map based on (e.g., using) the first 3D anatomical image by the processor; (d) obtaining (e.g., receiving, accessing, and / or generating) a second 3D anatomical segmentation map based on (e.g., using) the second 3D anatomical image by the processor; (e) determining a registration field (e.g., a full 3D registration field, e.g., point-to-point registration) using and / or based on the first 3D anatomical segmentation map and the second 3D anatomical segmentation map by the processor; (f) registering (aligning) the second 3D functional image with the first 3D functional image using the registration field, thereby generating a registered version of the second 3D functional image by the processor; (g) obtaining a first 3D hot spot map associated with the first functional image by the processor; (h) determining a second 3D hot spot map using the registered version of the second 3D functional image, wherein the second 3D hot spot map is thereby aligned with the first 3D hot spot map, by the processor; (i) using the first 3D hot spot map and the second 3D hot spot map aligned therewith by the processor,Determining the identification of one or more lesion responses; and (j) storing and / or providing, by a processor, the identification of one or more lesion responses for display and / or further processing. A method is contemplated.

[0060] In another aspect, the invention is a method for evaluating the effectiveness of an intervention, comprising: (a) for each particular subject in a test population presenting and / or at risk of a particular disease (e.g., prostate cancer (e.g., metastatic castration-resistant prostate cancer)), and / or registered in a clinical trial, e.g., comprising a plurality of subjects, performing the method of any one of the preceding claims on a plurality of medical images of the particular patient, wherein the plurality of medical images of the particular patient comprises time-series medical images acquired over a period spanning the intervention under test (e.g., before, during, and / or after), and one or more risk indices comprise one or more endpoints indicative of a patient response to the intervention under test, thereby determining a plurality of values for each of the one or more endpoints across the test population; and (b) determining the effectiveness of the intervention under test based on the values of the one or more endpoints across the test population. A method is contemplated.

[0061] In another aspect, the present invention is a method for treating a subject having and / or at risk of a particular disease (e.g., prostate cancer (e.g., metastatic castration-resistant prostate cancer)), the method comprising administering a first cycle of a therapeutic agent to the subject and, based on the subject being imaged (e.g., before, during, and / or after the first cycle of the therapeutic agent) and identified as a responder to the therapeutic agent using, for example, any one of the aspects and embodiments described herein in the above paragraphs (e.g., paragraphs

[0003] -

[0052] ), administering a second cycle of the therapeutic agent to the subject (e.g., the subject is identified / classified as a responder based on a value of one or more risk indices determined using, for example, any one of the aspects and embodiments described herein in the above paragraphs (e.g., paragraphs

[0003] -

[0052] )).

[0062] In another aspect, the present invention is a method for treating a subject having and / or at risk of a particular disease (e.g., prostate cancer (e.g., metastatic castration-resistant prostate cancer)), the method comprising administering a first cycle of a first therapeutic agent to the subject and, based on the subject being imaged (e.g., before, during, and / or after the first cycle of the first therapeutic agent) and identified as a non-responder to the first therapeutic agent using, for example, any one of the aspects and embodiments described herein in the above paragraphs (e.g., paragraphs

[0003] -

[0052] ), administering a second cycle of a second therapeutic agent to the subject (e.g., the subject is identified / classified as a non-responder based on a value of one or more risk indices determined using, for example, any one of the aspects and embodiments described herein in the above paragraphs (e.g., paragraphs

[0003] -

[0052] )) (e.g., thereby transitioning the subject to a potentially more effective therapy).

[0063] In another aspect, the present invention is a method for treating a subject having and / or at risk of a particular disease (e.g., prostate cancer (e.g., metastatic castration-resistant prostate cancer)), the method comprising administering a cycle of a therapeutic agent to the subject and, for example, imaging the subject (e.g., before, during, and / or after the first cycle of the therapeutic agent) using the method described in any one of the aspects and embodiments described herein in the above paragraphs (e.g., paragraphs

[0003] -

[0052] ), and discontinuing administration of the therapeutic agent to the subject based on the subject being identified as a non-responder to the therapeutic agent (e.g., the subject is identified / classified as a non-responder based on a value of one or more risk indices determined using the method described in any one of the aspects and embodiments described herein in the above paragraphs (e.g., paragraphs

[0003] -

[0052] )) (e.g., thereby potentially transitioning the subject to a more effective therapy).

[0064] In another aspect, the present invention is a method for automated or semi-automated whole-body evaluation of a subject suffering from metastatic prostate cancer [e.g., metastatic castration-resistant prostate cancer (mCRPC) or metastatic hormone-sensitive prostate cancer (mHSPC)] for assessing disease progression and / or treatment efficacy, the method comprising: (a) receiving, by a processor of a computing device, a first prostate-specific membrane antigen (PSMA)-targeted positron emission tomography (PET) image (a first PSMA-PET image) of the subject and a first 3D anatomical image of the subject [e.g., a computed tomography (CT) image, e.g., a magnetic resonance image (MRI)], wherein the first 3D anatomical image of the subject is acquired simultaneously with, immediately after, or immediately before (e.g., on the same date as) the first PSMA-PET image such that the first 3D anatomical image and the first PSMA PET image correspond to a first date, and the images depict a sufficiently large area of the subject's body to cover regions of the body to which metastatic prostate cancer has spread (e.g., a whole-body image or a full-body image covering multiple organs) {e.g., the PSMA-PET image is PYLARIFY®, F-18 piflufolastat F(18) (i.e., 2-(3-{1-carboxy-5-[(6-[18F]fluoropyridine-3-carbonyl)amino]-pentyl}ureido)-pentanedioic acid, also known as [18F]F-DCFPyL), or Ga-68(a) obtaining, using PSMA-11, or another radiolabeled prostate-specific membrane antigen inhibitor contrast agent, a first PSMA-PET image of a subject, obtained on a first date; (b) receiving, by a processor, a second PSMA-PET image of the subject and a second 3D anatomical image of the subject, both obtained on a second date following the first date; (c) automatically determining, by the processor, a registration field (e.g., a full 3D registration field, e.g., point-to-point registration) using landmarks automatically identified within both the first and second 3D anatomical images (e.g., an identified region representing one or more of cervical vertebrae, thoracic vertebrae, lumbar vertebrae, left and right lumbar bones, sacrum, and coccyx, left ribs and left scapula, right ribs and right scapula, left thigh, right thigh, skull, brain, and mandible), and aligning, by the processor, the first and second PSMA-PET images using the determined registration field [e.g., either before or after segmentation of the CT and / or PSMA-PET images for identifying organ and / or bone boundaries, and either before or after automatic hot spot (e.g., lesion) detection from the PSMA-PET images]; and (d) automatically detecting (e.g., staging and / or quantifying), by the processor, a change in disease (e.g., its progression or remission) from the first date to the second date using the first and second PSMA-PET images thus aligned [e.g., automatically identifying one or both of (i) and (ii) as follows: (i) a change in the number of lesions {e.g., one or more new lesions (e.g., organ-specific lesions), or elimination of one or more lesions (e.g., organ-specific)}, and (ii) a change in tumor size {e.g., an increase in tumor size (PSMA-VOL increase / decrease), e.g., total tumor size, or a decrease in tumor size (PSMA-VOL decrease)} {e.g., a change in the volume of each of one or more specific lesions, or an overall change in the volume of a specific type of lesion (e.g., organ-specific tumor), or a change in the total volume of the identified lesions}, and / or tagging or labeling the same as such].

[0065] In certain embodiments, the method includes one or more members selected from the group consisting of lesion location assignment, tumor staging, lymph node staging, distant metastasis staging, assessment of intraprostatic lesions, and determination of PSMA expression score.

[0066] In certain embodiments, the subject is administered, one or more times from a first date to a second date (after the first image is acquired and before the second image is acquired), a therapy {e.g., hormonal therapy, chemotherapy, and / or radiation therapy, e.g., androgen ablation therapy, e.g., a 177Lu-containing compound, e.g., 177Lu-PSMA radioligand therapy, e.g., 177Lu-PSMA-617, e.g., lutetium Lu177 vipivotide tetraxetan (Pluvicto), e.g., cabazitaxel} for the treatment of metastatic prostate cancer so that the method is used to assess treatment efficacy.

[0067] In certain embodiments, the method further includes, subsequent to the second date, acquiring one or more additional PSMA PET images and 3D anatomical images of the subject, aligning the additional PSMA PET images using the corresponding 3D anatomical images, and assessing disease progression and / or treatment efficacy using the aligned additional PSMA PET images.

[0068] In certain embodiments, the method further includes, at least in part by a processor, determining and rendering a predicted PSMA-PET image depicting a predicted progression (or remission) of the disease for a future date (e.g., a future date after the second date or any other subsequent date after the date on which the PSMA-PET image was acquired) based on the detected changes in the disease from the first date to the second date.

[0069] In another aspect, the present invention is a method for quantifying and reporting disease (e.g., tumor) burden in a patient having and / or at risk of cancer, the method comprising: (a) obtaining a medical image of the patient by a processor of a computing device; (b) detecting, by the processor, one or more (e.g., a plurality of) hot spots in the medical image, each hot spot corresponding to (e.g., being the specific 3D volume or comprising the specific 3D volume) a specific 3D volume in the medical image [e.g., a 3D hot spot volume, e.g., voxels of the 3D hot spot volume having an elevated intensity relative to their surroundings (e.g., and / or otherwise indicating an increased radiopharmaceutical uptake rate)] and representing a potential underlying physical lesion within the subject; (c) for each specific lesion class representing a specific tissue region and / or lesion subtype, identifying, by the processor, a corresponding subset of one or more hot spots as belonging to the specific lesion class (e.g., based on a determination made by the processor that the hot spot is located within a specific tissue region and / or represents an underlying physical lesion belonging to a specific lesion subtype represented by the specific lesion class), and determining, by the processor, a value of one or more patient indices (plural indices) that quantify the disease (e.g., tumor) burden within and / or associated with the specific lesion class based on the corresponding subset of hot spots; (d) causing, by the processor, a graphical representation (e.g., a summary table listing each lesion class and the calculated patient index value for each lesion class) of the patient index values calculated for each of the plurality of lesion classes to be displayed, thereby providing the user with a graphical report summarizing the tumor burden within a specific tissue region and / or associated with a specific lesion subtype.

[0070] In one embodiment, the plurality of lesion classes include, hereinafter, namely, (i) identifying potential lesions and / or portions thereof located within one or more local tumor-associated tissue regions associated with and / or adjacent to a localized (e.g., primary) tumor site within a patient, and the corresponding subset of hot spots representing the same, a local tumor class (e.g., “T” or “miT” class) [e.g., the cancer is prostate cancer, and one or more local tumor-associated tissue regions include the prostate and optionally one or more adjacent structures (e.g., seminal vesicles, external sphincter, rectum, bladder, levator ani, and / or pelvic wall); e.g., the cancer is breast cancer, and one or more local tumor-associated tissue regions include the breast; e.g., the cancer is colorectal cancer, and one or more local tumor-associated tissue regions include the colon; e.g., the cancer is lung cancer, and one or more local tumor-associated tissue regions include the lung], (ii) identifying potential lesions located within local lymph nodes adjacent to and / or proximal to the original (e.g., primary) tumor site, and the corresponding subset of hot spots representing the same, a local lymph node class (e.g., “N” or “miN” class) [e.g., the cancer is prostate cancer, and the local lymph node class identifies hot spots representing lesions located within one or more pelvic lymph nodes (e.g., internal iliac, external iliac, obturator, presacral, or other pelvic lymph nodes)], and (iii) identifying potential metastases (e.g., lesions that have spread outside the original (e.g., primary) tumor site) and / or subtypes thereof, and the corresponding subset of hot spots representing the same, one or more (e.g., distant) metastatic tumor classes (e.g., one or more “M” or “miM” classes) [e.g., the cancer is prostate cancer, and one or more metastatic tumor classes identify hot spots representing potential metastatic lesions located outside the patient's pelvic region (e.g., as defined by the pelvic brim, according to the American Joint Committee on Cancer Staging Manual)].

[0071] In one embodiment, one or more metastatic tumor classes identify, hereinafter, i.e., metastatic to distant lymph nodes, potential lesions, and a corresponding subset of hot spots represents a distant lymph node metastasis class (e.g., “Ma” or “miMa” class) [e.g., the cancer is prostate cancer, and the distant lymph node region class identifies hot spots representing lesions located within extra-pelvic (e.g., outside the pelvic region) lymph nodes (e.g., common iliac, retroperitoneal, supra-diaphragmatic, inguinal, and other extra-pelvic lymph nodes)], identify potential lesions located within one or more bones (e.g., distant bones) of the patient, and a corresponding subset of hot spots represents a distant bone metastasis class (e.g., “Mb” or “miMb” class), and identify potential lesions located within other non-lymphoid soft tissue regions outside one or more organs or local tumor-associated tissue regions, and a corresponding subset of hot spots represents a visceral (also referred to as distant soft tissue) metastasis class (e.g., “Mc” or “miMc” class) (e.g., the cancer is prostate cancer, and the visceral metastasis class identifies hot spots representing potential lesions located within extra-pelvic organs such as the patient's brain, lung, liver, spleen, and kidney), and includes one or more of the above.

[0072] In one embodiment, step (c) determines, for each specific lesion class, the value of one or more of the following patient indices, namely, a lesion count (e.g., calculated as the number of hot spots in the corresponding subset) that quantifies the number of lesions (e.g., distinct) represented by a subset of hot spots corresponding to the specific lesion class, a maximum uptake value that quantifies the maximum uptake rate within the corresponding set of hot spots (e.g., calculated as the maximum individual voxel intensity over all voxels within the corresponding subset of hot spot volumes, according to equation (13a)), an average uptake value that quantifies the overall average uptake rate within the corresponding subset of hot spots (e.g., calculated as the overall average intensity over all voxels within the (total combination) hot spot volume of the corresponding subset, according to equation (13b)), a total volume lesion volume that quantifies the total volume of lesions belonging to the specific lesion class (e.g., calculated as the sum value of all individual lesion (e.g., hot spot) volumes of the corresponding subset, according to equation (13c)), and an intensity-weighted tumor volume (ILTV) score (e.g., a PSMA score) calculated as the weighted sum value of all individual lesion volumes weighted (e.g., multiplied) by their intensity measurements [e.g., the intensity measurements are lesion indices that quantify the hot spot intensity on a normalized scale based on comparison with one or more reference intensities indicating the physiological (e.g., normal, non-cancer associated) radiopharmaceutical uptake rate within the corresponding reference tissue region such as the aorta portion and the liver] [e.g., calculated according to equation (13d)].

[0073] In one embodiment, the method includes, for each lesion class, determining an alphanumeric code that classifies the overall burden within a particular lesion class (e.g., (i) along with a particular lesion class, (ii) a corresponding subset of hot spots, and thus one or more numbers and / or alphanumeric codes indicating a particular number, size, spatial extent, spatial pattern, and / or sublocation of the underlying physical lesions they represent, a miTNM staging code), and optionally, in step (e), for each particular lesion class, causing the occurrence and / or display of a representation of the alphanumeric code.

[0074] In one embodiment, the method further includes determining an overall disease stage (e.g., an alphanumeric code) for the patient that indicates the overall disease status and / or burden for the patient, based on the plurality of lesion classes and their corresponding subsets of hot spots, and causing, by a processor, the rendering of a graphical representation of the overall disease stage (e.g., an alphanumeric code) for inclusion within a report.

[0075] In one embodiment, the method further includes determining, by a processor, one or more reference intensity values indicative of a physiological (e.g., normal, non-cancer related) uptake rate of a radiopharmaceutical, each calculated based on the intensity of image voxels within a corresponding reference volume identified within a medical image within a particular reference tissue region (e.g., an aortic portion, e.g., the liver) within the patient, and in step (d), causing, by a processor, the rendering of a representation (e.g., a table) of the one or more reference intensity values for inclusion within a report.

[0076] In another aspect, the present invention is a method for characterizing and reporting individual lesions detected based on an imaging assessment of a patient having cancer and / or at risk thereof, the method comprising: (a) obtaining, by a processor of a computing device, a medical image of the patient; (b) detecting, by the processor, a set of one or more (e.g., multiple) hot spots within the medical image, each hot spot of the set corresponding to (e.g., being, or comprising) a particular 3D volume within the medical image [e.g., a 3D hot spot volume, e.g., voxels of the 3D hot spot volume having increased intensity relative to their surroundings (e.g., and / or otherwise indicating an increased rate of radiopharmaceutical uptake)] and representing a potential underlying physical lesion within the subject; (c) assigning, by the processor, one or more lesion class labels to each of one or more hot spots of the set, each lesion class label class representing a specific tissue region and / or lesion subtype and identifying the hot spot as representing a potential lesion located within the specific tissue region and / or belonging to the lesion subtype; (d) calculating, by the processor, for each particular one of one or more individual hot spot quantification measurements, the value of the particular individual hot spot quantification measurement for each individual hot spot of the set; and (e) causing, by the processor, for at least some particular hot spots of the set of hot spots, a graphical representation [e.g., a summary table (e.g., a scrollable summary table) listing, for each individual hot spot as a row, the assigned lesion class and hot spot quantification measurements as columns] to be displayed, with the identification of the particular hot spot (e.g., a row in a table, optionally, an alphanumeric identification such as a number identifying the particular hot spot).

[0077] In one embodiment, the lesion class label comprises one or more of the following, namely: (i) a local tumor class (e.g., a "T" or "miT" class) that identifies a potential lesion and / or a part thereof that is located within one or more local tumor-associated tissue regions associated with and / or adjacent to a localized (e.g., primary) tumor site within a patient, and for which a corresponding subset of hot spots represents, e.g., the cancer is prostate cancer and one or more local tumor-associated tissue regions include the prostate and optionally one or more adjacent structures (e.g., seminal vesicles, external sphincter, rectum, bladder, levator ani, and / or pelvic wall), e.g., the cancer is breast cancer and one or more local tumor-associated tissue regions include the breast, e.g., the cancer is colorectal cancer and one or more local tumor-associated tissue regions include the colon, e.g., the cancer is lung cancer and one or more local tumor-associated tissue regions include the lung; (ii) a local lymph node class (e.g., an "N" or "miN" class) that identifies a potential lesion located within a local lymph node that is adjacent to and / or in proximity to the original (e.g., primary) tumor site, and for which a corresponding subset of hot spots represents, e.g., the cancer is prostate cancer and the local lymph node class identifies hot spots representing lesions located within one or more pelvic lymph nodes (e.g., internal iliac, external iliac, obturator, presacral, or other pelvic lymph nodes); and (iii) one or more (e.g., distant) metastatic tumor classes (e.g., one or more "M" or "miM" classes) that identify a potential metastasis (e.g., a lesion that has spread outside the original (e.g., primary) tumor site) and / or a subtype thereof, and for which a corresponding subset of hot spots represents, e.g., the cancer is prostate cancer and one or more metastatic tumor classes identify hot spots representing potential metastatic lesions located outside the patient's pelvic region (e.g., as defined by the pelvic brim, according to the American Joint Committee on Cancer Staging Manual).

[0078] In one embodiment, one or more metastatic tumor classes identify, and corresponding subsets of hot spots represent, potential lesions that have metastasized to remote lymph nodes, i.e., a remote lymph node metastasis class (e.g., the “Ma” or “miMa” class) [e.g., the cancer is prostate cancer, and the remote lymph node region class identifies hot spots representing lesions located within extra-pelvic (e.g., outside the pelvic region) lymph nodes (e.g., common iliac, retroperitoneal, supra-diaphragmatic, inguinal, and other extra-pelvic lymph nodes)], potential lesions located within one or more bones (e.g., remote bones) of the patient, and corresponding subsets of hot spots represent a remote bone metastasis class (e.g., the “Mb” or “miMb” class), and potential lesions located within other non-lymphatic soft tissue regions outside of one or more organs or tissue regions associated with the local tumor, and corresponding subsets of hot spots represent a visceral (also referred to as remote soft tissue) metastasis class (e.g., the “Mc” or “miMc” class) (e.g., the cancer is prostate cancer, and the visceral metastasis class identifies hot spots representing potential lesions located within extra-pelvic organs such as the patient's brain, lungs, liver, spleen, and kidneys), and includes one or more of these.

[0079] In one embodiment, the lesion class label comprises one or more tissue labels that identify a particular organ or bone (e.g., one or more of the organs or bone regions listed in Table 1) in which (the lesion represented by the hot spot) has been determined to be located (e.g., based on a comparison of the hot spot with an anatomical compartmentalization map).

[0080] In one embodiment, one or more individual hot spot quantification measurements are, i.e., the maximum intensity (e.g., SUV max )(e.g., determined according to any one of equations (1a), (1b), or (1c)), and the peak intensity (e.g., SUV peak)(determined, for example, according to any one of equations (3a), (3b), or (3c)), and the average intensity (e.g., SUV mean )(determined, for example, according to any one of equations (2a), (2b), (2c)), the lesion volume (determined, for example, according to any one of equations (5a) or (5b)), and the lesion index (measuring the intensity of the hot spot on the standardized scale) (determined, for example, according to equation (4)), including one or more than one of them.

[0081] In another aspect, the present invention is a method for quantifying and reporting over time the progression and / or risk of a disease (e.g., a tumor) in a patient having cancer and / or at risk thereof, the method comprising: (a) obtaining, by a processor of a computing device, a plurality of medical images of the patient, each medical image representing a scan of the patient (e.g., a longitudinal dataset) obtained at a particular time; (b) for each particular one of the plurality of medical images, detecting, by the processor, a corresponding set of one or more (e.g., a plurality of) hot spots within the particular medical image, each hot spot corresponding to (e.g., being, or comprising) a particular 3D volume within the medical image [e.g., a 3D hot spot volume, e.g., voxels of the 3D hot spot volume having an elevated intensity relative to their surroundings (e.g., and / or otherwise increased radiopharmaceutical uptake rate)] and representing a potential underlying physical lesion within the subject; (c) for each particular one of one or more (e.g., overall) patient indices for measuring (e.g., quantifying) the overall disease (e.g., tumor) burden within the patient at a particular time, determining, by the processor, a value of the particular (e.g., overall) patient index for each particular medical image of the plurality of medical images based on the corresponding set of hot spots detected for the particular medical image, thereby determining a set of values for tracking over time the change in disease burden as measured by the particular patient index value for each particular one of one or more patient indices; and (d) causing, by the processor, a graphical representation of a set of values for at least a portion (e.g., a particular one, a particular subset) of one or more of the patient index values to be displayed, thereby communicating a measure of the progression of the disease over time for the patient.

[0082] In one embodiment, one or more patient indices (plural indices) correspond to a particular medical image and quantify the number of lesions (e.g., distinct) represented by a set of hot spots detected therein (e.g., at a particular time point) (e.g., calculated as the number of hot spots within the corresponding set of hot spots), a maximum uptake value that quantifies the maximum uptake rate within the corresponding set of hot spots for the particular medical image (e.g., calculated as the maximum individual voxel intensity over all voxels within the hot spot volume of the corresponding set of hot spots for the particular medical image, according to equation (7a) or (7b)), an average uptake value that quantifies the overall average uptake rate within the corresponding set of hot spots (e.g., calculated as the overall average intensity over all voxels within the (total combination) hot spot volume of the corresponding set, according to equation (10a) or (10b)), a total volume lesion volume that quantifies the total volume of lesions detected within the subject at a particular time point (e.g., calculated as the sum value of all individual hot spot volumes of the corresponding set of hot spots detected within the particular medical image), and an intensity-weighted tumor volume (ILTV) score (e.g., an aPSMA score) calculated as the weighted sum value of all individual lesion volumes, where each individual lesion volume is weighted (e.g., multiplied) by its measured intensity value [e.g., the measured hot spot intensity is a lesion index that quantifies the hot spot intensity on a standardized scale based on comparison to one or more reference intensities indicative of the physiological (e.g., normal, non-cancer associated) radiopharmaceutical uptake rate within one or more corresponding reference tissue regions such as the aorta portion and the liver] [e.g., calculated according to equation (12)].

[0083] In one embodiment, the method further includes, for each particular medical image of a plurality of medical images, determining an overall disease stage (e.g., an alphanumeric code) indicative of the overall disease status and / or burden for the patient at a particular point in time based on the corresponding set of hot spots, and causing, by a processor, a graphical rendering of the overall disease stage (e.g., an alphanumeric code) at each point in time.

[0084] In one embodiment, the method further includes, by a processor, for each of a plurality of medical images, determining a set of one or more reference intensity values indicative of the physiological (e.g., normal, non-cancer related) uptake rate of a radiopharmaceutical within a particular reference tissue region (e.g., a portion of the aorta, e.g., the liver), each calculated based on the intensity of the image voxels within the corresponding reference volume identified within the medical image and within the patient, and causing, by the processor, a rendering of a representation (e.g., a table, e.g., a trace within a graph) of the one or more reference intensity values.

[0085] In another aspect, the present invention is a method for automatically processing a 3D image of a subject and determining a value of one or more patient indices (plural indices) that measure the (e.g., overall) disease burden and / or risk associated with the subject, the method comprising: (a) receiving, by a processor of a computing device, a 3D functional image of the subject acquired using a functional imaging modality; (b) partitioning, by the processor, a plurality of 3D hot spot volumes within the 3D functional image, each 3D hot spot volume corresponding to a local region of elevated intensity relative to its surroundings and representing a potential cancerous lesion within the subject, thereby obtaining a set of 3D hot spot volumes; (c) calculating, by the processor, for each particular one of one or more individual hot spot quantification measurements, a value of the particular individual hot spot quantification measurement for each individual 3D hot spot volume in the set, wherein for a particular individual 3D hot spot volume, each hot spot quantification measurement is a particular function of the properties (e.g., intensity, volume, etc.) of the particular 3D hot spot volume and the intensity and / or number of individual voxels within the particular 3D hot spot volume (e.g., calculated as such); and (d) determining, by the processor, a value of one or more patient indices (plural indices), wherein at least some of each patient index is associated with one or more specific individual hot spot quantification measurements and is calculated using a (e.g., same) particular function with the intensity and / or number of voxels within a combined hot spot volume that comprises at least a portion (e.g., substantially all, e.g., a particular subset) of the set of 3D hot spot volumes (e.g., formed as the union thereof).

[0086] In one embodiment, a particular patient index is an overall average voxel intensity, calculated as the overall average value of voxel intensities located within the combined hot spot volume.

[0087] In another aspect, the present invention is a method for automatically determining the prognosis for a subject having prostate cancer from one or more medical images of the subject [e.g., one or more PSMA PET images (PET images obtained in response to administering a PSMA-targeted compound to the subject) and / or one or more anatomical (e.g., CT) images], the method comprising: (a) receiving, by a processor of a computing device, and / or accessing, one or more images of the subject; and (b) automatically determining, by the processor, a quantitative assessment of one or more prostate cancer lesions (e.g., metastatic prostate cancer lesions) from the one or more images [e.g., the quantitative assessment comprises one or more members selected from the group consisting of: (i) a molecular imaging TNM (miTNM) lesion type classification for local (T), pelvic lymph node (N), and / or extrapelvic (M) disease (e.g., miT, miN, miMa (lymph), miMb (bone), miMc (other)), (ii) an indication of the lesion location (e.g., prostate, ilium, pelvic bone, thorax, etc.), (iii) a standardized physiological uptake value (SUV) (e.g., SUV max , SUV peak , SUV mean ), (iv) the total lesion volume, (v) the change in lesion volume (e.g., individual lesions and / or total lesions), and (vi) the calculated PSMA (aPSMA) score] (e.g., using one or more of the methods described herein); and (c) automatically determining, from the quantitative assessment in (b), the prognosis for the subject, the prognosis comprising one or more of the following for the subject: (I) the expected survival rate (e.g., in months), (II) the expected time to disease progression, (III) the expected time to radiological progression, (IV) the risk of synchronous (simultaneous) metastases, and (V) the risk of future (metachronous) metastases.

[0088] In one embodiment, the quantitative assessment of one or more prostate cancer lesions determined in step (b) comprises one or more of the following, namely: (A) total tumor volume, (B) change in tumor volume, (C) total SUV, and (D) PSMA score, and the prognosis for the subject determined in step (c) comprises one or more of the following, namely: (E) expected survival rate (e.g., number of months), (F) time to progression, and (G) time to radiographic progression.

[0089] In one embodiment, the quantitative assessment of one or more prostate cancer lesions determined in step (b) comprises one or more characteristics of PSMA expression in the prostate, and the prognosis for the subject determined in step (c) comprises the risk of synchronous (synchrony) metastases and / or the risk of future (metachronous) metastases.

[0090] In another aspect, the present invention is a method for automatically determining a response to treatment for a subject suffering from prostate cancer from a plurality of medical images of the subject [e.g., one or more PSMA PET images (PET images obtained in response to administering a PSMA targeting compound to the subject) and / or one or more anatomical (e.g., CT) images], comprising: (a) receiving and / or accessing, by a processor of a computing device, a plurality of images of the subject, wherein at least a first image of the plurality of images is obtained prior to administration of the treatment and at least a second image of the plurality of images is obtained subsequent to (e.g., after a period of) administration of the treatment; (b) automatically determining, by the processor, a quantitative assessment of one or more prostate cancer lesions (e.g., metastatic prostate cancer lesions) from the images [e.g., the quantitative assessment comprises one or more members selected from the group consisting of: (i) molecular imaging TNM (miTNM) lesion type classification for local (T), pelvic lymph node (N), and / or extrapelvic (M) disease (e.g., miT, miN, miMa (lymph), miMb (bone), miMc (other)), (ii) indication of lesion location (e.g., prostate, ilium, pelvic bone, thorax, etc.), (iii) standardized uptake value (SUV) (e.g., SUV max , SUV peak , SUV mean ), (iv) total lesion volume, (v) change in lesion volume (e.g., individual lesions and / or total lesions), and (vi) calculated PSMA (aPSMA) score] (e.g., using one or more of the methods described herein) (e.g., the quantitative assessment comprises Response Evaluation Criteria In PSMA-Imaging (RECIP) criteria and / or PSMA PET Progression (PPP) criteria); and (c) automatically determining, from the quantitative assessment in (b), whether (e.g., yes / no) the subject is responding to the treatment and / or the extent to which (e.g., numerically or categorically) the subject is responding to the treatment.

[0091] In another aspect, the present invention is a method for automatically identifying whether a subject suffering from prostate cancer (e.g., metastatic prostate cancer) is likely to benefit from a particular treatment for prostate cancer using a plurality of medical images of the subject [e.g., one or more PSMA PET images (PET images obtained in response to administering a PSMA-targeted compound to the subject) and / or one or more anatomical (e.g., CT) images], the method comprising: (a) receiving and / or accessing, by a processor of a computing device, a plurality of images of the subject; and (b) automatically determining, by the processor, a quantitative assessment of one or more prostate cancer lesions (e.g., metastatic prostate cancer lesions) from the images [e.g., the quantitative assessment comprises one or more members selected from the group consisting of (i) a molecular imaging TNM (miTNM) lesion type classification for local (T), pelvic lymph node (N), and / or extra-pelvic (M) disease (e.g., miT, miN, miMa (lymph), miMb (bone), miMc (other)), (ii) an indication of the lesion location (e.g., prostate, ilium, pelvic bone, thorax, etc.), (iii) a standardized physiological uptake value (SUV) (e.g., SUV max , SUV peak , SUV mean ), (iv) the total lesion volume, (v) changes in the lesion volume (e.g., individual lesions and / or total lesions), and (vi) a calculated PSMA (aPSMA) score] (e.g., using one or more of the methods described herein) (e.g., the quantitative assessment comprises Response Evaluation Criteria In PSMA-Imaging (RECIP) criteria and / or PSMA PET Progression (PPP) criteria); and (c) automatically determining, from the quantitative assessment in (b), whether the subject is likely to benefit from a particular treatment for prostate cancer [e.g., determining an efficacy score for one or more particular treatments and / or classes of treatments for the subject, e.g., a particular radioligand therapy, e.g., lutetium vipivotide tetraxetan (Pluvicto®)]. The method is directed to a subject.

[0092] In another aspect, the present invention is a system for automatically processing a 3D image of a subject and determining a value of one or more patient indices (plural indices) that measure the (e.g., overall) disease burden and / or risk associated with the subject, the system comprising a processor of a computing device and a memory having instructions stored thereon, the instructions, when executed by the processor, causing the processor to: (a) receive a 3D functional image of the subject acquired using a functional imaging modality; (b) compartmentalize a plurality of 3D hot spot volumes within the 3D functional image, each 3D hot spot volume corresponding to a local region of elevated intensity relative to its surroundings and representing a potential cancerous lesion within the subject, thereby obtaining a set of 3D hot spot volumes; (c) calculate a value of a specific one of one or more individual hot spot quantification measurements for each individual 3D hot spot volume of the set; (d) determine a value of one or more patient indices (plural indices), at least a portion of each of the patient indices being associated with one or more specific individual hot spot quantification measurements and being a function of at least a portion (e.g., substantially all, e.g., a specific subset) of the values of one or more specific individual hot spot quantification measurements calculated with respect to the set of 3D hot spot volumes.

[0093] In certain embodiments, the system has one or more features and / or instructions that cause the processor to perform one or more steps described herein (e.g., in the paragraphs above, e.g., paragraphs

[0004] -

[0031] ).

[0094] In another aspect, the present invention is a system for automated analysis of time - series medical images of a subject [e.g., 3 - D images, e.g., nuclear medicine images (e.g., bone scan (scintigraphy), PET, and / or SPECT), e.g., anatomical images (e.g., CT, X - ray, MRI), e.g., combined nuclear medicine and anatomical images (e.g., overlaid)], comprising a processor of a computing device and a memory having instructions stored thereon, the instructions, when executed by the processor, causing the processor to: (a) receive and / or access time - series medical images of the subject; (b) identify a plurality of hot spots within each of the medical images, and the processor to determine one, two, or all three of the following, namely: (i) a change in the number of identified lesions; (ii) a change in the overall volume of the identified lesions (e.g., a change in the sum value of the volumes of each identified lesion); and (iii) a change in the PSMA (e.g., lesion index) - weighted total volume (e.g., the sum value of the product of the lesion index and the lesion volume for all lesions within the region of interest) [e.g., the changes identified in step (b) are used to: (1) identify the disease status [e.g., progression, regression, or no change]; (2) make treatment management decisions [e.g., active surveillance, prostatectomy, anti - androgen therapy, prednisone, radiation, radiotherapy, radiation PSMA therapy, or chemotherapy]; or (3) identify treatment effectiveness (e.g., when the subject has started or is continuing treatment with a drug or other therapy following an initial set of images in the time - series medical images)] [e.g., step (b) includes the step of using a machine - learning module / model]].

[0095] In another aspect, the present invention is a system for analyzing a plurality of medical images of a subject (e.g., for assessing a disease state and / or progression within the subject), the system comprising a processor of a computing device and a memory having instructions stored thereon, the instructions, when executed by the processor, causing the processor to: (a) receive and / or access a plurality of medical images of the subject and cause the processor to obtain a plurality of 3D hot spot maps, each corresponding to a respective (plurality of) specific medical image, and identify one or more hot spots (e.g., representing potential underlying physical lesions within the subject) within the specific medical image; (b) for each of a specific one (medical image) of the plurality of medical images, use a machine learning module [e.g., a deep learning network (e.g., a convolutional neural network (CNN))] to determine a corresponding 3D anatomical partitioning map that identifies a set of organ regions within the specific medical image [e.g., representing one or more of soft tissue and / or bone structures within the subject (e.g., cervical vertebrae, thoracic vertebrae, lumbar vertebrae, left and right iliac bones, sacrum, and coccyx, left rib and left scapula, right rib and right scapula, left thigh, right thigh, skull, brain, and mandible)], thereby generating a plurality of 3D anatomical partitioning maps; (c) (i) use the plurality of 3D hot spot maps and (ii) the plurality of 3D anatomical partitioning maps to determine an identification of one or more lesion correspondences, each (lesion correspondence) being determined (e.g., by the processor) to identify two or more corresponding hot spots within different medical images and represent the same underlying physical lesion within the subject; (d) based on the plurality of 3D hot spot maps and the identification of one or more lesion correspondences, determine a value of one or more metrics {e.g., one or more hot spot quantification metrics and / or changes therein [e.g., quantifying changes in properties such as the volume of individual hot spots and / or the underlying physical lesions they represent, radiopharmaceutical uptake rate, shape, etc. (e.g., over time / between a plurality of medical images)], e.g., a patient index (e.g., measuring the overall disease burden and / or state and / or risk associated with the subject) and / or changes therein, e.g., a value for classifying the patient (e.g.,Belonging to and / or having a particular disease state, progression, category, etc., e.g., prognostic measurement values [e.g., indicating and / or quantifying the likelihood of one or more clinical outcomes (e.g., disease state, progression, likely survival rate, treatment efficacy, and equivalents) (e.g., overall survival rate)], e.g., predictive measurement values (e.g., indicating the predicted response to therapy and / or other clinical outcomes)} to be determined, a system comprising a memory is targeted.

[0096] In certain embodiments, the system has one or more features and / or instructions that cause the processor to perform one or more steps described herein (e.g., in the above paragraphs, e.g., paragraphs

[0034] -

[0048] ).

[0097] In another aspect, the present invention is a system for analyzing a plurality of medical images of a subject, the system comprising a processor of a computing device and a memory having instructions stored thereon, the instructions, when executed by the processor, causing the processor to: (a) obtain (e.g., receive, access, and / or generate) a first 3D hot spot map for the subject; (b) obtain (e.g., receive, access, and / or generate) a first 3D anatomical compartmentalization map associated with the first 3D hot spot map; (c) obtain (e.g., receive, access, and / or generate) a second 3D hot spot map for the subject; (d) obtain (e.g., receive, access, and / or generate) a second 3D anatomical compartmentalization map associated with the second 3D hot spot map; (e) use / based on the first 3D anatomical compartmentalization map and the second 3D anatomical compartmentalization map, determine a registration field (e.g., a 3D registration field and / or point-by-point registration); (f) use the registration field to register the first 3D hot spot map and the second 3D hot spot map, thereby generating a co-registered pair of 3D hot spot maps; (g) use the co-registered pair of 3D hot spot maps to determine the identification of one or more lesion correspondences; and (h) store and / or provide the identification of one or more lesion correspondences for display and / or further processing.

[0098] In another aspect, the present invention is a system for analyzing a plurality of medical images of a subject (e.g., for assessing a disease state and / or progression within the subject), comprising a processor of a computing device and a memory having instructions stored thereon, the instructions, when executed by the processor, causing the processor to: (a) receive and / or access a plurality of medical images of the subject; (b) for each particular one (medical image) of the plurality of medical images, use a machine learning module [e.g., a deep learning network (e.g., a convolutional neural network (CNN))] to determine a corresponding 3D anatomical compartmentalization map that identifies a set of organ regions within the particular medical image [e.g., representing one or more of soft tissue and / or bone structures within the subject (e.g., cervical vertebrae, thoracic vertebrae, lumbar vertebrae, left and right iliac bones, sacrum, and coccyx, left ribs and left scapula, right ribs and right scapula, left thigh, right thigh, skull, brain, and mandible)], thereby generating a plurality of 3D anatomical compartmentalization maps; (c) use the plurality of 3D anatomical compartmentalization maps to determine one or more alignment fields (e.g., a full 3D alignment field, e.g., point - by - point alignment), apply the one or more alignment fields, and align the plurality of medical images, thereby generating a plurality of aligned medical images; (d) for each particular one of the plurality of aligned medical images, determine a corresponding aligned 3D hot - spot map that identifies one or more hot - spots (e.g., representing potential underlying physical lesions within the subject) within the particular aligned medical image, thereby generating a plurality of aligned 3D hot - spot maps; (e) use the plurality of 3D aligned hot - spot maps to determine one or more lesion correspondence identifications, each (lesion correspondence) being determined (e.g., by the processor) to identify two or more corresponding hot - spots in different medical images and represent the same underlying physical lesion within the subject; (e) based on the plurality of 3D hot - spot maps and the one or more lesion correspondence identifications, values of one or more measurements {e.g.,One or more hot spot quantification measurement values and / or changes therein [e.g., quantifying changes in properties such as the volume of individual hot spots and / or the underlying physical lesions they represent (e.g., over time / between multiple medical images), radiopharmaceutical uptake rate, shape, etc.], e.g., patient indices (e.g., measuring the overall disease burden and / or status and / or risk for a subject) and / or changes therein, e.g., values for classifying patients (e.g., belonging to and / or having a particular disease state, progression, category, etc.), e.g., prognostic measurement values [e.g., indicating and / or quantifying one or more clinical outcomes (e.g., disease state, progression, likely survival rate, treatment effectiveness, and the like) (e.g., overall survival rate)], e.g., predictive measurement values (e.g., indicating the predicted response to therapy and / or other clinical outcomes)}, a system comprising a memory for determining.

[0099] In another aspect, the present invention is a system for analyzing a plurality of medical images of a subject, comprising a processor of a computing device and a memory having instructions stored thereon, the instructions, when executed by the processor, causing the processor to: (a) acquire (e.g., receive, access, and / or generate) a first 3D anatomical image (e.g., CT, X-ray, MRI, etc.) and a first 3D functional image [e.g., nuclear medicine image (e.g., PET, SPECT, etc.)] of the subject; (b) acquire (e.g., receive, access, and / or generate) a second 3D anatomical image and a second 3D functional image of the subject; (c) acquire (e.g., receive, access, and / or generate) a first 3D anatomical segmentation map based on (e.g., using) the first 3D anatomical image; (d) acquire (e.g., receive, access, and / or generate) a second 3D anatomical segmentation map based on (e.g., using) the second 3D anatomical image; (e) determine a registration field (e.g., 3D registration field and / or point-by-point registration) using and / or based on the first 3D anatomical segmentation map and the second 3D anatomical segmentation map; (f) register (align) the second 3D functional image with the first 3D functional image using the registration field, thereby generating a registered version of the second 3D functional image; (g) acquire a first 3D hot spot map associated with the first functional image; (h) determine a second 3D hot spot map using the registered version of the second 3D functional image, the second 3D hot spot map thereby being aligned with the first 3D hot spot map; (i) determine the identification of one or more lesion correspondences using the first 3D hot spot map and the second 3D hot spot map aligned therewith; (j) store and / or provide the identification of one or more lesion correspondences for display and / or further processing, a memory.The system is targeted.

[0100] In another aspect, the present invention is a system for automated or semi-automated whole-body evaluation of a subject suffering from metastatic prostate cancer [e.g., metastatic castration-resistant prostate cancer (mCRPC) or metastatic hormone-sensitive prostate cancer (mHSPC)] for assessing disease progression and / or treatment efficacy, the system comprising a processor of a computing device and a memory having instructions stored thereon, the instructions, when executed by the processor, causing the processor to: (a) receive a first prostate-specific membrane antigen (PSMA)-targeted positron emission tomography (PET) image (first PSMA-PET image) of the subject and a first 3D anatomical image of the subject [e.g., a computed tomography (CT) image, e.g., a magnetic resonance image (MRI)], the first 3D anatomical image of the subject being acquired simultaneously with, immediately after, or immediately before (e.g., on the same date as) the first PSMA-PET image such that the first 3D anatomical image and the first PSMA PET image correspond to a first date, the images depicting a sufficiently large area of the subject's body to cover the area of the body to which metastatic prostate cancer has spread (e.g., a whole-body image or full-body image covering multiple organs) {e.g., the PSMA-PET image is PYLARIFY®, F-18 piflufolastat F 18 PSMA (i.e., 2-(3-{1-carboxy-5-[(6-[18F]fluoropyridine-3-carbonyl)amino]-pentyl}ureido)-pentanedioic acid, also known as [18F]F-DCFPyL), or Ga-68PSMA-11, or other radiopharmaceutical imaging agents that are radiolabeled prostate specific membrane antigen inhibitors, (b) receive a second PSMA-PET image of the subject and a second 3D anatomical image of the subject, both acquired subsequent to a first date on a second date, (c) automatically determine a registration field (e.g., a full 3D registration field, e.g., point-by-point registration) using landmarks automatically identified within the first and second 3D anatomical images (e.g., an identified region representing one or more of the cervical vertebrae, thoracic vertebrae, lumbar vertebrae, left and right lumbar bones, sacrum, and coccyx, left ribs and left scapula, right ribs and right scapula, left thigh, right thigh, skull, brain, and mandible), and cause the processor to align the first and second PSMA-PET images using the determined registration field [e.g., either before or after segmentation of the CT and / or PSMA-PET images to identify organ and / or bone boundaries, and either before or after automatic hot spot (e.g., lesion) detection from the PSMA-PET images], (d) use the first and second PSMA-PET images thus aligned to automatically detect (e.g., stage and / or quantify) changes in the disease (e.g., its progression or remission) from the first date to the second date [e.g., one or both of (i) and (ii) below, i.e., (i) a change in the number of lesions {e.g., one or more new lesions (e.g., organ-specific lesions), or elimination of one or more lesions (e.g., organ-specific)}, and (ii) a change in tumor size {e.g., an increase in tumor size (PSMA-VOL increase / decrease), e.g., total tumor size, or a decrease in tumor size (PSMA-VOL decrease)} {e.g., a change in the volume of each of one or more specific lesions, or a change in the overall volume of a specific type of lesion (e.g., organ-specific tumor), or a change in the total volume of the identified lesions} of (i) and (ii) above, and / or cause such to be identified (e.g., tagged, labeled)], and a memory.

[0101] In one embodiment, the system has one or more features and / or instructions that cause a processor to perform one or more steps described herein (e.g., in the above paragraphs, e.g., paragraphs

[0057] -

[0060] ).

[0102] In another aspect, the present invention is a system for quantifying and reporting disease (e.g., tumor) burden in a patient having and / or at risk of cancer, the system comprising a processor of a computing device and a memory having instructions stored thereon, the instructions, when executed by the processor, causing the processor to: (a) obtain a medical image of the patient; (b) detect one or more (e.g., a plurality of) hot spots in the medical image, each hot spot corresponding to (e.g., being, or comprising) a particular 3D volume in the medical image [e.g., a 3D hot spot volume, e.g., voxels of the 3D hot spot volume having increased intensity relative to their surroundings (e.g., and / or otherwise indicating an increased radiopharmaceutical uptake rate)] and representing a potential underlying physical lesion in the subject; (c) for each particular lesion class of a plurality of lesion classes representing specific tissue regions and / or lesion subtypes, identify a corresponding subset of one or more hot spots as belonging to the particular lesion class (e.g., based on a determination made by the processor that the hot spot is located within a particular tissue region and / or represents an underlying physical lesion belonging to a particular lesion subtype represented by the particular lesion class), and based on the corresponding subset of hot spots, determine a value of one or more patient indices (plural) that quantify the disease (e.g., tumor) burden within and / or associated with the particular lesion class; (d) cause a graphical representation (e.g., a summary table listing each lesion class and the calculated patient index value for each lesion class) of the patient index values calculated for each of the plurality of lesion classes to be displayed, thereby providing the user with a graphical report that outlines the tumor burden within a specific tissue region and / or associated with a specific lesion subtype.

[0103] In one embodiment, the system has one or more features and / or instructions that cause a processor to perform one or more steps described herein (e.g., in the above paragraphs, e.g., paragraphs

[0062] -

[0067] ).

[0104] In one embodiment, the present invention is a system for characterizing and reporting individual lesions detected based on an imaging assessment of a patient having and / or at risk of cancer, comprising a processor of a computing device and a memory having instructions stored thereon, the instructions, when executed by the processor, causing the processor to: (a) acquire a medical image of the patient; (b) detect a set of one or more (e.g., multiple) hot spots within the medical image, each hot spot in the set corresponding to (e.g., being, or comprising) a particular 3D volume within the medical image [e.g., a 3D hot spot volume, e.g., voxels of the 3D hot spot volume having an elevated intensity relative to their surroundings (e.g., and / or otherwise indicating an increased radiopharmaceutical uptake rate)] and representing a potential underlying physical lesion within the subject; (c) assign one or more lesion class labels to each of one or more hot spots in the set, each lesion class label class representing a specific tissue region and / or lesion subtype and identifying the hot spot as representing a potential lesion located within the specific tissue region and / or belonging to the lesion subtype; (d) calculate, for each of a particular one of one or more individual hot spot quantification measurements, the value of the particular individual hot spot quantification measurement for each individual hot spot in the set; (e) cause a graphical representation [e.g., a summary table (e.g., a scrollable summary table) listing, for each of at least some particular hot spots in the set, the identification of the particular hot spot (e.g., a row in a table, optionally, an alphanumeric identification such as a number identifying the particular hot spot) along with one or more lesion class labels assigned to the particular hot spot and the value of one or more individual hot spot quantification measurements calculated for the particular hot spot] to be displayed.

[0105] In one embodiment, the system has one or more features and / or instructions that cause a processor to perform one or more steps described herein (e.g., in the above paragraphs, e.g., paragraphs

[0069] -

[0072] ).

[0106] In another aspect, the present invention is a system for quantifying and reporting over time the progression and / or risk of a disease (e.g., a tumor) in a patient having and / or at risk of cancer, the system comprising a processor of a computing device and a memory having instructions stored thereon, the instructions, when executed by the processor, causing the processor to: (a) obtain a plurality of medical images of the patient, each medical image representing a scan of the patient (e.g., a longitudinal dataset) obtained at a particular time; (b) for each particular one of the plurality of medical images, detect a corresponding set of one or more (e.g., a plurality of) hot spots within the particular medical image, each hot spot corresponding to a particular 3D volume within the medical image [e.g., a 3D hot spot volume, e.g., the voxels of the 3D hot spot volume having an elevated intensity relative to their surroundings (e.g., and / or otherwise indicating an increased radiopharmaceutical uptake rate)] and representing a potential underlying physical lesion within the subject; (c) for each particular one of one or more (e.g., overall) patient indices that measure (e.g., quantify) the overall disease (e.g., tumor) burden within the patient at a particular time, determine a value of the particular (e.g., overall) patient index for each particular medical image based on the corresponding set of hot spots detected for the particular medical image, thereby determining a set of values that track over time the change in disease burden as measured by the particular patient index value for each particular one of one or more patient indices; (d) cause a graphical representation of a set of values for at least a portion (e.g., a particular one, a particular subset) of one or more of the patient index values to be displayed, thereby communicating a measure of the progression of the disease over time for the patient.

[0107] In one embodiment, the system has one or more features and / or instructions that cause a processor to perform one or more steps described herein (e.g., in the above paragraphs, e.g., paragraphs

[0074] -

[0076] ).

[0108] In another aspect, the invention is a system for automatically processing a 3D image of a subject and determining a value of one or more patient indices (plural indices) that measure the (e.g., overall) disease burden and / or risk associated with the subject, the system comprising a processor of a computing device and a memory having instructions stored thereon, the instructions, when executed by the processor, causing the processor to: (a) receive a 3D functional image of the subject acquired using a functional imaging modality; (b) segment a plurality of 3D hot spot volumes within the 3D functional image, each 3D hot spot volume corresponding to a local region of elevated intensity relative to its surroundings and representing a potential cancerous lesion within the subject, thereby obtaining a set of 3D hot spot volumes; (c) calculate, for each of one or more specific individual hot spot quantification measurements, a value of the specific individual hot spot quantification measurement for each individual 3D hot spot volume of the set, wherein for a particular individual 3D hot spot volume, each hot spot quantification measurement is a specific function of the properties (e.g., intensity, volume, etc.) of the particular 3D hot spot volume and of the intensity and / or number of individual voxels within the particular 3D hot spot volume (e.g., calculated as such); (d) determine a value of one or more patient indices (plural indices), at least some of each of the patient indices being associated with one or more specific individual hot spot quantification measurements and being calculated using a (e.g., the same) specific function involving the intensity and / or number of voxels within a combined hot spot volume that comprises at least a portion (e.g., substantially all, e.g., a particular subset) of the set of 3D hot spot volumes (e.g., formed as its union).

[0109] In one embodiment, a particular patient index is the overall average voxel intensity, calculated as the overall average value of the voxel intensities located within the combined hot spot volume.

[0110] In another aspect, the present invention is a system for automatically determining a prognosis for a subject having prostate cancer from one or more medical images of the subject [e.g., one or more PSMA PET images (PET images obtained in response to administering a PSMA-targeted compound to the subject) and / or one or more anatomical (e.g., CT) images], the system comprising a processor of a computing device and a memory having instructions stored thereon, the instructions, when executed by the processor, causing the processor to: (a) receive and / or access one or more images of the subject, and (b) automatically determine a quantitative assessment of one or more prostate cancer lesions (e.g., metastatic prostate cancer lesions) from the one or more images [e.g., the quantitative assessment may include: (i) a molecular imaging TNM (miTNM) lesion type classification for local (T), pelvic lymph node (N), and / or extrapelvic (M) disease (e.g., miT, miN, miMa (lymph), miMb (bone), miMc (other)), (ii) an indication of the lesion location (e.g., prostate, ilium, pelvic bone, thorax, etc.), (iii) a standardized uptake value (SUV) (e.g., SUV max , SUV peak , SUV mean)、(iv) total lesion volume, (v) change in lesion volume (e.g., individual lesions and / or total lesions), and (vi) one or more members selected from the group consisting of the calculated PSMA (aPSMA) score](e.g., using one or more of the methods described herein), (c) automatically determine the prognosis for the subject from the quantitative assessment in (b), and the prognosis comprises one or more of the following for the subject, namely, (I) expected survival rate (e.g., number of months), (II) expected time to disease progression, (III) expected time to radiological progression, (IV) risk of synchronous (simultaneous) metastases, and (V) risk of future (metachronous) metastases, and a memory.

[0111] In certain embodiments, the quantitative assessment of one or more prostate cancer lesions determined in step (b) comprises one or more of the following, namely, (A) total tumor volume, (B) change in tumor volume, (C) total SUV, and (D) PSMA score, and the prognosis for the subject determined in step (c) comprises one or more of the following, namely, (E) expected survival rate (e.g., number of months), (F) time to progression, and (G) time to radiological progression.

[0112] In certain embodiments, the quantitative assessment of one or more prostate cancer lesions determined in step (b) comprises one or more characteristics of PSMA expression in the prostate, and the prognosis for the subject determined in step (c) comprises the risk of synchronous (simultaneous) metastases and / or the risk of future (metachronous) metastases.

[0113] In another aspect, the present invention is a system for automatically determining a response to treatment for a subject suffering from prostate cancer from a plurality of medical images of the subject [e.g., one or more PSMA PET images (PET images obtained in response to administering a PSMA targeting compound to the subject) and / or one or more anatomical (e.g., CT) images], comprising a processor of a computing device and a memory having instructions stored thereon, the instructions, when executed by the processor, causing the processor to: (a) receive and / or access a plurality of images of the subject by the processor of the computing device, at least a first image of the plurality of images being obtained prior to administration of the treatment and at least a second image of the plurality of images being obtained subsequent to (e.g., after a period of) administration of the treatment; (b) automatically determine a quantitative assessment of one or more prostate cancer lesions (e.g., metastatic prostate cancer lesions) from the images [e.g., the quantitative assessment comprises one or more members selected from the group consisting of: (i) a molecular imaging TNM (miTNM) lesion type classification for local (T), pelvic lymph node (N), and / or extrapelvic (M) disease (e.g., miT, miN, miMa (lymph), miMb (bone), miMc (other)); (ii) an indication of the lesion location (e.g., prostate, ilium, pelvic bone, thorax, etc.); (iii) a standardized physiologic uptake value (SUV) (e.g., SUV max , SUV peak , SUV mean ); (iv) a total lesion volume; (v) a change in lesion volume (e.g., individual lesions and / or total lesions); and (vi) a calculated PSMA (aPSMA) score] (e.g., using one or more of the methods described herein) (e.g., the quantitative assessment comprises Response Evaluation Criteria In PSMA-Imaging (RECIP) criteria and / or PSMA PET Progression (PPP) criteria); (c) automatically determine from the quantitative assessment in (b) whether the subject is responding to the treatment (e.g., yes / no) and / or the degree to which the subject is responding to the treatment (e.g., a numerical value or classification). The present invention is directed to a system comprising the memory.

[0114] In another aspect, the present invention provides a system for automatically identifying whether a subject suffering from prostate cancer (e.g., metastatic prostate cancer) is likely to benefit from a particular treatment for prostate cancer using a plurality of medical images of the subject [e.g., one or more PSMA PET images (PET images obtained in response to administering a PSMA-targeted compound to the subject) and / or one or more anatomical (e.g., CT) images]. The system includes a processor of a computing device and a memory having instructions stored thereon. When the instructions are executed by the processor, the processor is caused to: (a) receive and / or access a plurality of images of the subject; (b) automatically determine a quantitative assessment of one or more prostate cancer lesions (e.g., metastatic prostate cancer lesions) from the images [e.g., the quantitative assessment may include: (i) a molecular imaging TNM (miTNM) lesion type classification for local (T), pelvic lymph node (N), and / or extrapelvic (M) disease (e.g., miT, miN, miMa (lymph), miMb (bone), miMc (other)); (ii) an indication of the lesion location (e.g., prostate, ilium, pelvic bone, rib cage, etc.); (iii) a standardized uptake value (SUV) (e.g., SUV max 、SUV peak 、SUV mean)、(iv) total lesion volume, (v) changes in lesion volume (e.g., individual lesions and / or total lesions), and (vi) one or more members selected from the group consisting of the calculated PSMA (aPSMA) score (e.g., using one or more of the methods described herein) (e.g., quantitative assessment comprises Response Evaluation Criteria in PSMA-Imaging (RECIP) criteria and / or PSMA Prostate Progress (PPP) criteria), and (c) automatically determining from the quantitative assessment in (b) whether the subject is likely to benefit from a particular treatment for prostate cancer [e.g., determining an efficacy score for the subject with respect to one or more particular treatments and / or classes of treatments, e.g., a particular radioligand therapy, e.g., lutetium vipivotide tetraxetan (Pluvicto®)], a system comprising a memory is targeted.

[0115] In another aspect, the invention is a therapeutic agent for use in the treatment of a subject having and / or at risk of a particular disease (e.g., prostate cancer (e.g., metastatic castration-resistant prostate cancer)) (e.g., via multiple cycles of a therapeutic agent), wherein the subject has been administered and imaged during a first cycle of the therapeutic agent (e.g., before, and / or during, and / or after the first cycle of the therapeutic agent), and (ii) has been identified as a responder to the therapeutic agent using the methods described in the above paragraphs of this specification, e.g., paragraphs

[0003] -

[0052] etc. (e.g., the subject has been identified / classified as a responder based on a value of one or more risk indices determined using the methods described in the above paragraphs of this specification, e.g., paragraphs

[0003] -

[0052] etc.), the therapeutic agent is targeted.

[0116] In another aspect, the present invention is a second (e.g., second - choice) therapeutic agent for use in the treatment of a subject having and / or at risk of having a particular disease (e.g., prostate cancer (e.g., metastatic castration - resistant prostate cancer)), wherein the subject has (i) been administered a cycle of a first therapeutic agent and imaged (e.g., before, and / or during, and / or after a cycle of the first therapeutic agent), and (ii) has been identified as a non - responder to the first therapeutic agent using the methods described in the paragraphs above, e.g., paragraphs

[0003] -

[0052] herein (e.g., the subject has been identified / classified as a non - responder based on a value of one or more risk indices determined using the methods described in the paragraphs above, e.g., paragraphs

[0003] -

[0052] herein), (e.g., thereby potentially transitioning the subject to a more effective therapy), and the therapeutic agent is directed to the subject.

[0117] Features of embodiments described with respect to one aspect of the invention may be applicable to another aspect of the invention. **Brief Description of the Drawings**

[0118] The foregoing and other objects, aspects, features, and advantages of the present disclosure will become more apparent and better understood by referring to the following description in conjunction with the accompanying drawings.

[0119]

Figure 1A

[0120]

Figure 1B

[0121]

Figure 2

[0122]

Figure 3

[0123]

Figure 4A

[0124]

Figure 4B

[0125]

Figure 4C

[0126]

Figure 5

[0127]

Figure 6A

[0128]

Figure 6B

[0129]

Figure 6C

[0130]

Figure 7

[0131]

Figure 8

[0132]

Figure 9A

[0133]

Figure 9B

[0134]

Figure 9C

[0135]

Figure 10

[0136]

Figure 11A

[0137]

Figure 11B-1

Figure 11B-2

[0138]

Figure 12

[0139]

Figure 13

[0140]

Figure 14

[0141] The features and advantages of the present disclosure will become more apparent from the following detailed description taken in conjunction with the drawings, in which like reference symbols identify corresponding elements throughout. In the drawings, like reference numbers generally indicate identical, functionally similar, and / or structurally similar elements.

DETAILED DESCRIPTION OF THE INVENTION

[0142] Specific Definitions For easier understanding of the present disclosure, certain terms are first defined below. Additional definitions regarding the following terms and other terms are set forth throughout the specification.

[0143] A, An: As used herein, the articles “a” and “an” are used to refer to one or more than one (i.e., at least one) of the grammatical objects of the articles. By way of example, “an element” means one element or more than one element. Accordingly, in this specification and the appended claims, the singular forms “a,” “an,” and “the” include plural references unless the context clearly dictates otherwise. Thus, for example, a reference to a pharmaceutical composition comprising “an agent” includes a reference to two or more agents.

[0144] About, Approximately: As used in this application, the terms “about” and “approximately” are used as synonyms. Any number used in this application is meant to cover any normal variation understood by one of ordinary skill in the art, whether or not the term “about” or “approximately” is present. In certain embodiments, the term “about” or “approximately” refers to a range of values that fall within 25%, 20%, 19%, 18%, 17%, 16%, 15%, 14%, 13%, 12%, 11%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, or less of the recited reference value, in either direction (greater than or less than the recited reference value), unless otherwise stated or otherwise apparent from the context (except in cases where such a number would be greater than 100% of a possible value).

[0145] First, second, etc.: Any reference in this specification to an element using the notations "first", "second", etc. is to be understood as not limiting the quantity or order of those elements unless a limitation is explicitly stated. Rather, these notations may be used in this specification as a convenient way to distinguish between two or more elements or instances of an element. Thus, a reference to first and second elements does not mean that only two elements may be employed or that the first element must precede the second element in a particular manner. Additionally, unless stated otherwise, a set of elements may comprise one or more elements.

[0146] Image: As used herein, an "image", e.g., a 3D image of a subject, includes any visual representation such as a photograph, video frame, streaming video, and any electronic, digital, or mathematical analog thereof, such as a photograph (e.g., a digital image), video frame, or streaming video, that is displayed or stored in memory (e.g., a digital image can be displayed for visual inspection, but need not be). In certain embodiments, any device described herein includes a display for displaying an image or any other result produced by a processor. Any method described herein includes, in certain embodiments, the step of displaying an image or any other result produced via the method. In certain embodiments, an image is a 3D image that conveys information that varies with a position within a 3D volume. Such an image may be represented digitally, for example, as a 3D matrix (e.g., an N×M×L matrix), where each voxel of the 3D image is represented by an element of the 3D matrix. Other representations are also contemplated and included, e.g., the 3D matrix may be reformed as a vector by stitching each row or column end-to-end (e.g., a 1×K size vector, where K is the total number of voxels). Examples of images include, for example, bone scan images (also referred to as scintigraphic images), computed tomography (CT) images, magnetic resonance images (MRI), optical images (e.g., bright field microscopy images, fluorescence images, reflection or transmission images, etc.), positron emission tomography (PET) images, single photon emission computed tomography (SPECT) images, ultrasound images, X-ray images, and medical images such as equivalents. In certain embodiments, a medical image is or includes a nuclear medicine image made from radiation emitted from within the subject being imaged. In certain embodiments, a medical image is or includes an anatomical image (e.g., a 3D anatomical image) that conveys information regarding the location and extent of anatomical structures such as internal organs, bones, soft tissue, and blood vessels within the subject. Examples of anatomical images include, but are not limited to, X-ray images, CT images, MRI, and ultrasound images.In one embodiment, the medical image is, or includes, a functional image (e.g., a 3D functional image) that conveys information related to physiological activity within a specific organ and / or tissue, such as metabolism, blood flow, local chemical composition, absorption rate, etc. Examples of functional images include, but are not limited to, nuclear medicine images such as PET images, SPECT images, and other functional imaging modalities such as functional MRI (fMRI), which measure small changes in blood flow for use in assessing brain activity.

[0147] Map: As used herein, the term "map" is understood to mean a visual display or any data representation interpretable for a visual display, which contains spatially correlated information. For example, a three-dimensional map of a given volume may include a dataset of values of a given quantity that vary in three spatial dimensions throughout the volume. The three-dimensional map may be displayed in two dimensions (e.g., on a two-dimensional screen or on a two-dimensional print).

[0148] Compartmentalized map: As used herein, the term "compartmentalized map" refers to a computer representation that identifies one or more 2D or 3D regions determined by partitioning an image. In one embodiment, the compartmentalized map discriminately identifies a plurality of different (e.g., partitioned) regions such that they can be individually and discriminately accessed and operated on, and / or used to act on, for example, one or more images.

[0149] 3D, three-dimensional: As used herein, "3D" or "three-dimensional" with reference to an "image" means conveying information about three dimensions. A 3D image may be rendered as a dataset in three dimensions and / or displayed as a collection of two-dimensional representations or as a three-dimensional representation. In one embodiment, a 3D image is represented as voxel (e.g., volume pixel) data.

[0150] Whole body: As used herein, the terms "whole body" and "entire body", used (synonymously) in the context of compartmentalization and other modes of identifying regions within an image of a subject, refer to an approach that assesses a majority (e.g., greater than 50%) of the graphical representation of the body of the subject within a 3D anatomical image and identifies the target tissue regions of interest. In some embodiments, whole body and entire body compartmentalization refers to, at least, the identification of target tissue regions within the entire torso of the subject. In some embodiments, portions of the extremities are also included, along with the head of the subject.

[0151] Radionuclide: As used herein, "radionuclide" refers to a moiety that includes a radioisotope of at least one element. Exemplary suitable radionuclides include, but are not limited to, those described herein. In some embodiments, the radionuclide is one used in positron emission tomography (PET). In some embodiments, the radionuclide is one used in single photon emission computed tomography (SPECT). In some embodiments, a non-limiting list of radionuclides is 99m Tc, 111 In, 64 Cu, 67 Ga, 68 Ga, 186 Re, 188 Re, 153 Sm, 177 Lu, 67 Cu, 123 I, 124 I, 125 I, 126 I, 131 I, 11 C, 13 N, 15 O, 18 F, 153 Sm, 166 Ho, 177 Lu, 149 Pm, 90 Y, 213 Bi, 103 Pd, 109 Pd, 159 Gd, 140 La, 198 Au, 199 Au, 169 Yb,175 Yb, 165 Dy, 166 Dy, 105 Rh, 111 Ag, 89 Zr, 225 Ac, 82 Rb, 75 Br, 76 Br, 77 Br, 80 Br, 80m Br, 82 Br, 83 Br, 211 At, and 192 includes Ir.

[0152] Radiopharmaceutical: As used herein, the term "radiopharmaceutical" refers to a compound that contains a radionuclide. In certain embodiments, the radiopharmaceutical is used for diagnostic and / or therapeutic purposes. In certain embodiments, the radiopharmaceutical includes a small molecule labeled with one or more radionuclides, an antibody labeled with one or more radionuclides, and a light source binding portion of an antibody labeled with one or more radionuclides.

[0153] Machine learning module: Certain embodiments described herein utilize (e.g., include) software instructions that include one or more machine learning modules, also referred to herein as artificial intelligence software. As used herein, the term "machine learning module" refers to a computer-implemented process (e.g., function) that implements one or more specific machine learning algorithms to determine one or more output values for a given input (such as an image (e.g., a 2D image, e.g., a 3D image), a dataset, and the like). For example, a machine learning module may receive, as input, a 3D image of a subject (e.g., a CT image, e.g., an MRI), and for each voxel of the image, determine a value representing the likelihood that the voxel is within a region of the 3D image that corresponds to the representation of a particular organ or tissue of the subject. In certain embodiments, two or more machine learning modules may be combined and implemented as a single module and / or a single software application. In certain embodiments, two or more machine learning modules may also be implemented separately, e.g., as separate software applications. The machine learning module may be software and / or hardware. For example, the machine learning module may be implemented entirely as software, or certain functions of the CNN module may be performed via special hardware (e.g., via an application specific integrated circuit (ASIC)).

[0154] Subject: As used herein, "subject" means a human or other mammal (e.g., rodents (mice, rats, hamsters), pigs, cats, dogs, horses, primates, rabbits, and the like).

[0155] Administer: As used herein, to "administer" an agent means to introduce a substance (e.g., a contrast agent) into a subject. Generally, any route of administration may be utilized, including, for example, parenteral (e.g., intravenous), oral, topical, subcutaneous, peritoneal, intra-arterial, inhalation, vaginal, rectal, nasal, introduction into the cerebrospinal fluid, or infusion into a body compartment.

[0156] Tissue: As used herein, the term "tissue" refers to bone (osseous tissue) as well as soft tissue. Detailed Description

[0157] The systems, architectures, devices, methods, and processes of the claimed invention are contemplated to include variations and adaptations developed using information from the embodiments described herein. Conformance and / or modification of the systems, architectures, devices, methods, and processes described herein may be implemented as considered by this description.

[0158] Throughout the description, when articles, devices, systems, and architectures are described as having, including, or comprising specific components, or when processes and methods are described as having, including, or comprising specific steps, in addition, there exist articles, devices, systems, and architectures of the invention consisting essentially of, or consisting of, the recited components, and there exist processes and methods according to the invention consisting essentially of, or consisting of, the recited process steps.

[0159] It should be understood that the order of steps or the order for performing certain actions is not important as long as the invention remains operable. Also, two or more steps or actions may be performed simultaneously.

[0160] Any mention of a publication herein, for example, in the Background section, is not an admission that the publication serves as prior art with respect to any of the claims presented herein. The Background section is presented for clarity purposes and is not intended as an explanation of the prior art with respect to any claim.

[0161] The document is incorporated herein by reference as described. If any differences exist in the meaning of specific terms, the meaning provided in the above-defined section shall control.

[0162] The headings are provided for the convenience of the reader and are not intended to limit the scope of the subject matter described herein.

[0163] A. Nuclear Medicine Images Nuclear medicine images can be acquired using nuclear medicine imaging modalities such as bone scan imaging (also referred to as scintigraphy), positron emission tomography (PET) imaging, and single photon emission computed tomography (SPECT) imaging.

[0164] In certain embodiments, nuclear medicine images are acquired using a contrast agent comprising a radiopharmaceutical. The nuclear medicine images are acquired following administration of the radiopharmaceutical to a patient (e.g., a human subject) and can provide information regarding the distribution of the radiopharmaceutical within the patient.

[0165] Nuclear imaging techniques detect radiation emitted from the radionuclides of radiopharmaceuticals and form images. The distribution of a specific radiopharmaceutical within a patient can be affected and / or determined by biological mechanisms such as blood flow or perfusion, as well as by specific enzyme or receptor binding interactions. Different radiopharmaceuticals utilize different biological mechanisms and / or specific enzyme or receptor binding interactions and can thus be designed to selectively concentrate within specific types of tissue and / or regions within a patient when administered. More radiation is emitted from regions within a patient that have a higher concentration of the radiopharmaceutical than other regions, such that these regions appear brighter in the nuclear medicine image. Thus, intensity variations within the nuclear medicine image can be used to map the distribution of the radiopharmaceutical within the patient. The mapped distribution of the radiopharmaceutical within the patient can be used, for example, to infer the presence of cancerous tissue within various regions of the patient's body. In certain embodiments, the intensity of a voxel in a nuclear medicine image, such as a PET image, represents a standardized uptake value (SUV) (e.g., calibrated with respect to the radiopharmaceutical dose injected and / or the patient's body weight).

[0166] For example, in response to administration to a patient, technetium 99m methylene diphosphonate ( 99m Tc MDP) selectively accumulates within the skeletal regions of a patient, particularly at sites associated with abnormal bone formation associated with malignant bone lesions. The selective concentration of the radiopharmaceutical at these sites creates identifiable hot spots, i.e., regions of high intensity, within the nuclear medicine image. Thus, the presence of malignant bone lesions associated with metastatic prostate cancer can be inferred by identifying such hot spots within a whole body scan of the patient. In certain embodiments, by detecting and evaluating the characteristics of the hot spots, etc., to the patient 99mAnalyzing the intensity variations in a whole-body scan obtained following administration of Tc MDP can be used to calculate a risk index that correlates with the overall survival rate and disease state, progression, treatment effectiveness, and other prognostic measures indicative of the patient, such as equivalents. In certain embodiments, other radiopharmaceuticals can also be used in a manner similar to 99m Tc MDP.

[0167] In certain embodiments, the particular radiopharmaceutical used depends on the particular nuclear medicine imaging modality used. For example, 18 sodium fluoride (NaF) can also accumulate within bone lesions in a manner similar to 99m Tc MDP and can be used in conjunction with PET imaging. In certain embodiments, PET imaging can also utilize a radioactive form of vitamin choline that is readily absorbed by prostate cancer cells.

[0168] In certain embodiments, a radiopharmaceutical that selectively binds to a particular protein or receptor of interest, particularly one whose expression is increased within cancerous tissue, may be used. Such proteins or receptors of interest include, but are not limited to, tumor antigens such as CEA, which is expressed within colorectal cancer, Her2 / neu, which is expressed within multiple cancers, BRCA1 and BRCA2, which are expressed within breast and ovarian cancers, and TRP-1 and TRP-2, which are expressed within melanoma.

[0169] For example, human prostate-specific membrane antigen (PSMA) is upregulated in prostate cancer, including metastatic disease. PSMA is expressed by virtually all prostate cancers, and its expression is further increased in poorly differentiated, metastatic, and hormone-resistant cancers. Thus, a radiopharmaceutical comprising a PSMA binder (e.g., a compound having high affinity for PSMA) labeled with one or more radionuclides can be used to obtain a nuclear medicine image of a patient, from which the presence and / or status of prostate cancer in various regions of the patient (e.g., including, but not limited to, the skeletal region) can be assessed. In certain embodiments, the nuclear medicine image obtained using a PSMA binder is used to identify the presence of cancerous tissue within the prostate when the disease is in a localized state. In certain embodiments, the nuclear medicine image obtained using a radiopharmaceutical comprising a PSMA binder is used to identify the presence of cancerous tissue in various regions, including not only the prostate but also other organ and tissue regions such as the lungs, lymph nodes, and bones, that may be relevant when the disease is metastatic.

[0170] In particular, upon administration to a patient, the radionuclide-labeled PSMA binders selectively accumulate within cancerous tissue based on their affinity for PSMA. 99mIn a manner similar to that described above with respect to Tc MDP, the selective accumulation of a radionuclide-labeled PSMA binder at a specific site within a patient creates a detectable hot spot in the nuclear medicine image. As the PSMA binder accumulates within various cancerous tissues and regions of the body that express PSMA, localized cancer within the patient's prostate and / or metastatic cancer within various regions of the patient's body can be detected and / or evaluated. Various measurements that indicate and / or quantify the severity (e.g., likelihood of malignancy) of an individual lesion, the overall disease burden and risk for the patient, and the like, can be calculated based on an automated analysis of intensity variations within the nuclear medicine image obtained following administration of the PSMA binder radiopharmaceutical to the patient. These disease burden and / or risk measurements may be used to stage the disease and to perform assessments regarding patient overall survival and other prognostic measurements that indicate disease state, progression, and treatment effectiveness.

[0171] Various radionuclide-labeled PSMA binders may be used to detect and evaluate prostate cancer as a radiopharmaceutical contrast agent for nuclear medicine imaging. In certain embodiments, the particular radionuclide-labeled PSMA binder used depends on factors such as a particular imaging modality (e.g., PET, e.g., SPECT) and the particular region (e.g., organ) of the patient to be imaged. For example, one radionuclide-labeled PSMA binder may be suitable for PET imaging while another is suitable for SPECT imaging. For example, one radionuclide-labeled PSMA binder facilitates the step of imaging the patient's prostate and is primarily used when the disease is localized, while another facilitates the step of imaging organs and regions throughout the patient's body and is useful for evaluating metastatic prostate cancer.

[0172] Several exemplary PSMA binders and their radionuclide-labeled versions are described in further detail in Section H of this specification, as well as in U.S. Pat. Nos. 8,778,305, 8,211,401, and 8,962,799, and U.S. Patent Publication No. US2021 / 0032206A1, the contents of each of which are incorporated herein by reference in their entireties.

[0173] B. Image Segmentation in Nuclear Medicine Imaging Nuclear medicine images are functional images. Functional images convey information related to physiological activities within specific organs and / or tissues, such as metabolism, blood flow, local chemical composition, and / or absorption rate. In certain embodiments, nuclear medicine images are obtained and / or analyzed in combination with anatomical images, such as computed tomography (CT) images. Anatomical images provide information regarding the location and extent of anatomical structures, such as internal organs, bones, soft tissues, and blood vessels, within a subject. Examples of anatomical images include, but are not limited to, X-ray images, CT images, magnetic resonance images, and ultrasound images.

[0174] Thus, in certain embodiments, anatomical images can be analyzed together with nuclear medicine images to provide the anatomical context for the functional information conveyed by the latter. For example, nuclear medicine images, such as PET and SPECT, convey the three-dimensional distribution of radiopharmaceuticals within a subject, but adding the anatomical context from an anatomical imaging modality, such as CT imaging, enables determination of the specific organs, soft tissue regions, bones, etc., in which the radiopharmaceutical has accumulated.

[0175] For example, functional images may be aligned with anatomical images such that locations within each image that correspond to the same physical location and thus correspond to each other can be identified. For example, coordinates and / or pixels / voxels within the functional and anatomical images may be defined with respect to a common coordinate system, or a mapping (i.e., a functional relationship) may be established between voxels in the anatomical image and voxels in the functional image. In this way, one or more voxels in the anatomical image and one or more voxels in the functional image that represent the same physical location or volume can be identified as corresponding to each other.

[0176] For example, FIG. 1 shows axial slices of a 3D CT image 102 and a 3D PET image 104 together with a fused image 106 in which slices of the 3D CT image are displayed in grayscale and the PET image is displayed as a translucent overlay. By alignment between the CT image and the PET image, the location of hot spots in the PET image, which indicate accumulated radiopharmaceutical, and thus potential lesions, can be identified in the corresponding CT image and visualized in an anatomical context, e.g., within a specific location within the pelvic region (e.g., within the prostate). FIG. 1B shows another PET / CT fusion showing cross-sectional and sagittal slices.

[0177] In some embodiments, the pairs to be aligned are composite images such as PET / CT or SPECT / CT. In some embodiments, the anatomical image (e.g., a 3D anatomical image such as a CT image) and the functional image (e.g., a 3D functional image such as a PET or SPECT image) are obtained using separate anatomical and functional imaging modalities, respectively. In some embodiments, the anatomical image (e.g., a 3D anatomical image such as a CT image) and the functional image (e.g., a 3D functional image such as a PET or SPECT image) are obtained using a single multi-modal imaging system. The functional and anatomical images may be obtained, for example, via two scans using a single multimode imaging system where, for example, a CT scan is first performed and then a PET scan is performed a second time, during which the subject remains in a substantially fixed position.

[0178] In some embodiments, the 3D boundaries of a particular tissue region of interest can be accurately identified by analyzing a 3D anatomical image. For example, automated segmentation of the 3D anatomical image can be performed to segment the 3D boundaries of a particular organ, organ sub-region and soft tissue region, as well as regions such as bone. In some embodiments, organs such as the prostate, bladder, liver, aorta (e.g., a portion of the aorta such as the thoracic aorta), and parotid gland are segmented. In some embodiments, one or more particular bones are segmented. In some embodiments, the entire skeleton is segmented.

[0179] In one embodiment, the automated segmentation of 3D anatomical images may be performed using one or more machine learning modules trained to receive a 3D anatomical image and / or a portion thereof as input, segment one or more specific regions of interest, and produce a 3D segmentation map as output. For example, multiple machine learning modules implementing a convolutional neural network (CNN), as described in PCT Publication No. WO / 2020 / 144134, titled "Systems and Methods for Platform Agnostic Whole Body Segmentation" and published on July 16, 2020, the content of which is incorporated herein by reference in its entirety, may be used to segment 3D anatomical images such as CT images of a subject's whole body, thereby generating a 3D segmentation map that identifies multiple target tissue regions across the subject's body.

[0180] In one embodiment, for example, to segment an organ where a functional image is considered to provide additional useful information that facilitates segmentation, the machine learning module may receive both the anatomical image and the functional image as input, for example, as two different channels of the input (similar to multiple color channels in a color RGB image), and use these two inputs to determine the anatomical segmentation. This multi-channel approach is described in more detail in U.S. Patent Publication No. US2021 / 0334974A1, titled "Systems and Methods for Deep-Learning-Based Segmentation of Composite Images" and published on October 28, 2021, the content of which is incorporated herein by reference in its entirety.

[0181] In one embodiment, as shown in FIG. 2, an anatomical image 204 (e.g., a 3D anatomical image such as a CT image) and a functional image 206 (e.g., a 3D functional image such as a PET or SPECT image) may be aligned (e.g., co - registered) with each other, such as in a composite image 202 such as a PET / CT image. The anatomical image 204 may be segmented 208 to generate a segmentation map 210 (e.g., a 3D segmentation map) that discriminatively identifies one or more tissue regions and / or regions of interest, such as one or more specific organs and / or bones. The segmentation map 210, which is generated from the anatomical image 204, is aligned with the anatomical image 204, which in turn is aligned with the functional image 206. Thus, the boundaries of specific regions, such as specific organs and / or bones, identified via the segmentation map 210 (e.g., segmentation masks) are transferred and / or overlaid 212 onto the functional image 206 for the purpose of determining useful indices that serve as measurements and / or predictors of cancer status, progression, and response to treatment, and can identify volumes within the functional image 206. The segmentation map and mask may also be displayed as a graphical representation, for example, overlaid on a medical image to guide physicians and other medical practitioners.

[0182] C. Lesion Detection and Characterization In one embodiment, the approaches described herein include techniques for detecting and characterizing lesions within a subject via (e.g., automated) analysis of medical images such as nuclear medicine images. As described herein, in one embodiment, hot spots are highly intense (e.g., contiguous) localized regions within an image, such as a 3D functional image, relative to their surroundings, and may indicate potential cancerous lesions present within the subject.

[0183] Various approaches can be used to detect, segment, and classify hot spots. In some embodiments, hot spots are detected and segmented using analytical methods such as filtering techniques, including but not limited to the Difference of Gaussians (DoG) filter and the Laplacian of Gaussians (LoG) filter. In some embodiments, hot spots are segmented using a machine learning module that receives, as input, a 3D functional image such as a PET image and generates, as output, a hot spot segmentation map ("hot spot map") that differentiates the boundaries of the identified hot spots from the background. In some embodiments, each segmented hot spot within the hot spot map is individually identifiable (e.g., individually labeled). In some embodiments, the machine learning module used to segment hot spots may take, as input, in addition to the 3D functional image, one or both of a 3D anatomical image (e.g., a CT image) and a 3D anatomical segmentation map. The 3D anatomical segmentation map may be generated via automated segmentation of the 3D anatomical image (e.g., as described herein).

[0184] In certain embodiments, the compartmentalized hot spots may be classified according to the anatomical regions in which they are located. For example, in certain embodiments, the location of each individual compartmentalized hot spot within a hot spot map (representing and identifying the compartmentalized hot spots) is compared to the 3D boundaries of the compartmentalized tissue regions such as various organs and bones within a 3D anatomical compartmentalization map, and may be labeled according to their location, for example, based on proximity and / or overlap with a particular organ. In certain embodiments, a machine learning module may be used to classify the hot spots. For example, in certain embodiments, the machine learning module, as output, not only labels and makes individually distinguishable (e.g., mutually distinguishable) the compartmentalized hot spots therein, but also may generate a hot spot map labeled to correspond to, for example, one of bone, lymph, or prostate lesions. In certain embodiments, one or more machine learning modules may be combined with mutual and analytical compartmentalization (e.g., thresholding) techniques to perform various tasks in parallel and in sequence to generate a final labeled hot spot map.

[0185] Various approaches for performing detailed compartmentalization of 3D anatomical images and identification of hotspots representing lesions within 3D functional images, which can be used in conjunction with the various approaches described herein, are described in PCT Publication No. WO / 2020 / 144134, entitled "Systems and Methods for Platform Agnostic Whole Body Segmentation", published on July 16, 2020, U.S. Patent Publication No. US2021 / 0334974A1, entitled "Systems and Methods for Deep-Learning-Based Segmentation of Composite Images", published on October 28, 2021, and PCT Publication No. WO / 2022 / 008374, entitled "Systems and Methods for Artificial Intelligence-Based Image Analysis for Detection and Characterization of Lesions", published on January 13, 2022 (the contents of each of which are incorporated herein by reference in their entirety).

[0186] FIG. 3 shows an exemplary process 300 for segmenting and classifying hot spots based on an exemplary approach, which is further described in PCT Publication No. WO / 2022 / 008374, entitled "Systems and Methods for Artificial Intelligence-Based Image Analysis for Detection and Characterization of Lesions", published on January 13, 2022. The approach illustrated in FIG. 3 uses two machine learning modules, each of which receives, as input, a 3D functional image 306, a 3D anatomical image 304, and a 3D anatomical segmentation map 310. Machine learning module 312a is a binary classifier that generates a single-class hot spot map 320a by labeling voxels as either hot spots or background (non-hot spots). Machine learning module 312b performs multi-class segmentation and generates a multi-class hot spot map 320b, in which hot spots are segmented and labeled as belonging to one of three classes: prostate, lymph, or bone. In particular, classifying hot spots in this way (e.g., as opposed to directly comparing the hot spot locations with the segmented boundaries from the segmentation map 310) obviates the need to segment an area. For example, in one embodiment, machine learning module 312b may classify hot spots as belonging to the prostate, lymph, or bone without the need to identify and segment the prostate region from the 3D anatomical image 304 (e.g., in one embodiment, the 3D anatomical segmentation map 310 does not include a prostate region). In one embodiment, hot spot maps 320a and 320b are merged (e.g., based on overlap), for example, by transferring labels to hot spot segmentation identified within single-class hot spot map 320a from multi-class hot spot map 320b.While not desiring to be bound by any particular theory, this approach is thought to combine improved segmentation and hot spot detection from the single-class machine learning module 312a with classification results from the multi-class machine learning module 312b. In certain embodiments, the hot spot regions identified via this final merged hot spot map are further refined using analysis techniques such as adaptive thresholding techniques, as described in PCT Publication No. WO / 2022 / 008374, published on January 13, 2022, entitled "Systems and Methods for Artificial Intelligence-Based Image Analysis for Detection and Characterization of Lesions".

[0187] In certain embodiments, once detected and segmented, hot spots may be identified and labeled according to the specific anatomical (e.g., tissue) regions in which they are located and / or the specific lesion subtypes they are likely to represent. For example, in certain embodiments, hot spots may be assigned an anatomical location that identifies them as representing a location with one of a set of tissue regions as listed in Table 1 below. In certain embodiments, the list of tissue regions may include those in Table 1 as well as the gluteus maximus (e.g., left and right) and the gallbladder. In certain embodiments, hot spots are assigned to and / or labeled as belonging to a particular tissue region based on machine learning classification and / or via comparison of the various tissue volumes identified via the location and / or overlap of their 3D hot spot volumes and masks within the anatomical segmentation map. In certain embodiments, the prostate gland is not segmented. For example, as described above, in certain embodiments, the machine learning module 312b may classify hot spots as belonging to the prostate, lymph, or bone without the need for the prostate region to be identified and segmented from the 3D anatomical image 304. [Table 1]

[0188] In certain embodiments, additionally or alternatively, hot spots may be classified as belonging to one or more lesion subtypes. In certain embodiments, lesion subtype classification may be performed by comparing the hot spot location with the classes of anatomical regions. For example, in certain embodiments, a miTNM classification scheme may be used, and hot spots are labeled as belonging to one of three classes, namely miT, miN, or miM, based on whether they represent lesions located within the prostate (miT), pelvic lymph nodes (miN), or distant metastases (miM). In certain embodiments, a five-class version of the miTNM scheme may be used, and distant metastases are further divided into three subclasses, namely miMb for bone metastases, miMa for lymph node metastases, and miMc for other soft tissue metastases.

[0189] For example, in certain embodiments, hot spots located within the prostate are labeled as belonging to class "T" or "miT", for example, representing a local tumor. In certain embodiments, hot spots that are outside the prostate but located within the pelvic region are labeled as class "N" or "miN". In certain embodiments, for example, as described in U.S. Application No. 17 / 959,357, filed Oct. 4, 2022, titled "Systems and Methods for Automated Identification and Classification of Lesions in Local Lymph and Distant Metastases", and published Apr. 13, 2023, as U.S. 2023 / 0115732A1, the contents of which are incorporated herein by reference in their entirety, a pelvic anatomy map may be aligned for the purpose of identifying pelvic lymph node lesions and may identify the boundaries of the pelvic region and / or various sub-regions therein. The pelvic anatomy map may include, for example, the boundaries of the pelvic region and / or a planar reference (e.g., a plane passing through the bifurcation of the aorta), against which the hot spot locations can be compared (e.g., such that a hot spot located outside the pelvic region and / or above the planar reference passing through the bifurcation of the aorta is labeled as "M" or "miM", for example, as a distant metastasis). In certain embodiments, distant metastases may be classified as lymph (miMa), bone (miMb), or visceral (miMc) based on a comparison of the hot spot locations with an anatomical compartmentalization map. For example, one or more hot spots located within the bone (e.g., outside the pelvic region) may be labeled as distant metastases, one or more hot spots located within a compartmentalized organ or subset of organs (e.g., brain, lung, liver, spleen, kidney) may be labeled as visceral (miMc) distant metastases, and the remaining hot spots located outside the pelvic region may be labeled as distant lymph metastases (miMa).

[0190] Additionally, or alternatively, in certain embodiments, hot spots may be assigned an miTNM class based on a determination of their location within a particular anatomical region, e.g., based on a table such as Table 2 where each column corresponds to a particular miTNM label (the first row indicates a particular miTNM class) and rows 2 and onwards include the particular anatomical regions associated with each miTNM class. In certain embodiments, hot spots may be assigned as being located within the specific tissue regions listed in Table 2 based on a comparison of the hot spot locations to an anatomical compartmentalization map, enabling automated miTNM class assignment.

Table 2

[0191] In certain embodiments, hot spots may be further classified in terms of their anatomical location and / or lesion subtype. For example, in certain embodiments, hot spots identified as being located within pelvic lymph (miN) may be identified as belonging to a particular pelvic lymph node sub-region such as one of the left / right internal iliac, left or right external iliac, left or right common iliac, left or right obturator lymph nodes, presacral region, or other pelvic regions. In certain embodiments, distant lymph node metastases (miMa) may be classified as retroperitoneal (RP), supra-diaphragmatic (SD), or other extra-pelvic (OE). Approaches for local (miN) and distant (miMa) lymph node metastasis classification may include registration of pelvic anatomy images and / or identification of various systemic landmarks, which are described in further detail in U.S. Application No. 17 / 959,357, filed Oct. 4, 2022, titled "Systems and Methods for Automated Identification and Classification of Lesions in Local Lymph and Distant Metastases," and published Apr. 13, 2023, as U.S. 2023 / 0115732A1, the content of which is incorporated herein by reference in its entirety.

[0192] D. Individual Hot Spot Quantification Measurement Values In certain embodiments, detected, e.g., identified and compartmentalized, hot spots may be characterized via various individual hot spot quantification measurement values. In particular, for a particular individual hot spot, the individual hot spot quantification measurement values may be used in a manner that indicates the size of the underlying (e.g., potential) physical lesion represented by the particular hot spot and / or the level of radiopharmaceutical uptake rate therein, to quantify measurements of the size (e.g., 3D volume) and / or intensity of the particular hot spot. Thus, the individual hot spot quantification measurement values can convey, for example, to a physician or radiologist, the likelihood that a hot spot appearing in an image represents a true underlying physical lesion and / or the likelihood or level of its malignancy (e.g., enabling differentiation between a benign and a malignant lesion).

[0193] In certain embodiments, the image compartmentalization, lesion detection, and characterization techniques described herein are used to determine a corresponding set of hot spots for each one or more medical images. As described herein, the image compartmentalization technique may be used to determine a 3D hot spot volume that represents and / or indicates a particular 3D volume, i.e., the volume of a potential underlying physical lesion within the subject (e.g., 3D location and extent), for each hot spot detected within a particular image. Each 3D hot spot volume, and thus, a set of image voxels, each having a particular intensity value, is provided.

[0194] Once determined, the set of 3D hot spot volumes may be used to calculate one or more hot spot quantification measurement values for each individual hot spot. The individual hot spot quantification measurement values may be calculated, for example, according to various methods and formulas described herein below. In the following description, the variable L is used to refer to a set of detected hot spots associated with a particular image, where L = {1, 2, …, l, …, NL} represents the set of N L (i.e., N L is the number of hot spots) hot spots, and the variable l indicates the l-th hot spot. As described herein, each hot spot corresponds to a specific 3D hot spot volume within the image, and R l represents the 3D hot spot volume of the l-th hot spot.

[0195] The hot spot quantification measurement values may be presented to the user via a graphical user interface (GUI) and / or a generated report (e.g., automatically or semi-automatically). As described in more detail herein, the individual hot spot quantification measurement values may each include a hot spot intensity measurement value and a hot spot volume measurement value (e.g., lesion volume) that quantify the intensity and size of a specific hot spot and / or the underlying lesion it represents. The hot spot intensity and size may in turn indicate the level and size of the radiopharmaceutical uptake rate within the underlying physical lesion within the subject, respectively.

[0196] Hot spot intensity measurement value In certain embodiments, the hot spot quantification measurement values are, or include, individual hot spot intensity measurement values that quantify the intensity of the individual 3D hot spot volumes. The hot spot intensity measurement values may be calculated based on the individual voxel intensities within the identified hot spot volumes. For example, for a particular hot spot, the value of the hot spot intensity measurement value may be calculated as a function of at least a portion (e.g., a specific subset, e.g., all) of the voxel intensities of that hot spot. The hot spot intensity measurement values may include, but are not limited to, measurements such as maximum hot spot intensity, average hot spot intensity, and peak hot spot intensity, and equivalents thereof. Similar to the voxel intensities in nuclear medicine images, in certain embodiments, the hot spot intensity measurement values may represent SUV values (e.g., in that unit).

[0197] In one embodiment, the value of a particular hot spot intensity measurement is calculated (e.g., as a function thereof) based only on the voxel intensity of the target hot spot, rather than based on the intensity of other image voxels outside the 3D volume of the target hot spot, for example, with respect to the target hot spot.

[0198] For example, the hot spot intensity measurement is calculated as the maximum voxel intensity (e.g., SUV or uptake rate) within the 3D hot spot volume, the maximum hot spot intensity (e.g., SUV) or "SUV" max may be. In one embodiment, the maximum hot spot intensity may be calculated according to the following equations (1a), (1b), or (1c).

Chemical formula

[0199] In one embodiment, the hot spot intensity measurement may be the average hot spot intensity (e.g., SUV) or "SUV" mean and may be calculated as the average value (e.g., SUV or uptake rate) across all voxel intensities within the 3D hot spot volume. In one embodiment, the average hot spot intensity may be calculated according to the following equations (2a), (2b), or (2c).

Chemical formula

[0200] In one embodiment, the hot spot intensity measurement value may be the peak hot spot intensity (e.g., SUV) or " peak SUV", and may be calculated as an average value over the intensity of the voxels (e.g., SUV or uptake rate), the midpoint of which is located within a specific distance (e.g., within 5 mm) of the midpoint of the hot spot voxel (e.g., predefined), and the maximum intensity (e.g., max SUV) is located within the hot spot and may, therefore, be calculated according to the following equations (3a)-(3c).

Chemical Formula

[0201] Lesion index measurement value In one embodiment, the hot spot intensity measurement value is an individual lesion index value that maps the intensity of the voxels within a specific 3D hot spot volume to a value on a standardized scale. Such lesion index values are described in more detail in PCT / EP2020 / 050132, filed Jan. 6, 2020, and PCT / EP2021 / 068337, filed Jul. 2, 2021, the contents of each of which are hereby incorporated by reference in their entirety. The calculation of the lesion index value may include the calculation of reference intensity values within a specific reference tissue region such as the aortic portion (also referred to as the blood pool) and / or the liver.

[0202] For example, in one particular implementation, the first blood pool reference intensity value is determined based on the measured intensity (e.g., average SUV) within the aortic region, and the second liver reference intensity value is determined based on the measured intensity (e.g., average SUV) within the liver region. For example, as described in more detail in PCT / EP2021 / 068337, filed July 2, 2021, the content of which is hereby incorporated by reference in its entirety, the calculation of the reference intensity identifies a reference volume (e.g., the aorta or a part thereof, e.g., the liver volume) within a functional image such as a PET or SPECT image, and contracts and / or expands a reference volume, e.g., to avoid including voxels on the edge of the reference volume, and selects a subset of the reference voxel intensities based on a modeling approach, e.g., to account for abnormal tissue features such as cysts and lesions within the liver. In certain embodiments, the third reference intensity value may be determined either as a multiple (e.g., two times) of the liver reference intensity value or based on the intensity of another reference tissue region such as the parotid gland.

[0203] In certain embodiments, the hot spot intensity may be compared to one or more reference intensity values, and a lesion index may be determined as a value on a standardized scale, which facilitates comparison across different images. For example, FIG. 4C illustrates an approach for assigning lesion index values ranging from 0 to 3 to a hot spot. In the approach shown in FIG. 4C, the blood pool (aorta) intensity value is assigned a lesion index of 1, the liver intensity value is assigned a lesion index of 2, and a value that is two times the liver intensity is assigned a lesion index of 3. The lesion index for a particular hot spot can be determined by first calculating a value of an initial hot spot intensity measurement for the particular hot spot, e.g., Q mean (l) or SUV mean , and comparing the value of the initial hot spot intensity measurement to the reference intensity value. For example, the value of the initial hot spot intensity measurement falls into four ranges, namely, [0, SUV blood , (SUV blood , SUVliver , (SUV liver , 2×SUV liver , and 2×SUV liver (e.g., (2×SUV liver , ∞)) among those exceeding can fall within one of them. The lesion index value can then be calculated for a particular hot spot based on (i) the value of the initial hot spot intensity measurement and (ii) a linear interpolation according to a specific range within which the value of the initial hot spot intensity measurement falls, as illustrated in FIG. 4C. The filled dots and empty dots on the horizontal (SUV) and vertical (LI) axes illustrate exemplary values of the initial hot spot intensity measurement and the resulting lesion index values, respectively. In certain embodiments, if an SUV reference value for either the liver or the aorta cannot be calculated, or if the aorta value is higher than the liver value, the lesion index will not be calculated and will be displayed as "-".

[0204] The lesion index value according to the mapping scheme, described above and illustrated in FIG. 4C, may be calculated, for example, as shown in Equation (4) below.

Chemical formula

[0205] Hot spot / lesion volume In certain embodiments, the hot spot quantification measurement value may be a volume measurement such as lesion volume Q vol which provides a measurement of the size (e.g., volume) of the underlying physical lesion represented by the hot spot. The lesion volume may be calculated, in certain embodiments, as shown in Equations (5a) and (5b) below.

Chemical formula

[0206] E. Aggregating Hot Spot Measurements In certain embodiments, the systems and methods described herein calculate patient index values that quantify the disease burden and / or risk for a particular subject. The values of various patient indices may be calculated using the values of the individual hot spot quantification measurements, e.g., as a function thereof. In particular, in certain embodiments, a particular patient index value aggregates the values of a plurality of individual hot spot quantification measurements calculated with respect to the entire set of hot spots detected for a patient and / or with respect to a particular subset of hot spots, e.g., associated with a particular tissue region and / or lesion subtype. In certain embodiments, a particular patient index is associated with one or more specific individual hot spot quantification measurements and is calculated using the value(s) of the specific individual hot spot quantification measurement(s) calculated for at least a portion of each of the individual 3D hot spot volumes within the set.

[0207] Overall Patient Index For example, in certain embodiments, a particular patient index aggregates the values of one or more specific individual hot spot quantification measurements calculated across substantially the entire set of 3D hot spot volumes detected for a patient at a particular point in time, e.g., providing an overall measurement of the total disease burden for the subject at a particular point in time, and may be an overall patient index.

[0208] In one embodiment, a particular patient index may be associated with a single specific individual hot spot quantification measurement value, or may be calculated as a function of substantially all values of that specific individual hot spot quantification measurement value with respect to a set of 3D hot spot volumes. Such a patient index can be regarded as having a functional form.

Chemical formula

[0209] The function f (p) can be various functions, which preferably aggregate (combine) the entire set of values of a specific individual hot spot quantification measurement value Q (m) . For example, the function f (p) can be a total value, an average value, a median value, a mode value, a maximum value, etc. Different specific functions can be used for f (m) depending on the specific individual hot spot quantification measurement value Q (p) being aggregated. Therefore, various individual hot spot quantification measurement values (e.g., average intensity, median intensity, mode intensity, peak intensity, individual lesion index, volume) can be combined in various ways, for example, by obtaining the overall total value, average value, median value, mode value, etc. over substantially all values calculated with respect to the set of 3D hot spot volumes.

[0210] For example, in one embodiment, the overall patient index may be the overall intensity maximum value, which is calculated as the maximum value over all individual hot spot maximum intensity values, as shown in the following equation (7a) or (7b).

Chemical formula

Chemical formula

[0211] In one embodiment, a particular patient index value may be calculated, for example, as the sum of the average intensity values of substantially all individual hot spots, as shown in the following equations (8a) and (8b), for example, as the total value of the average intensity values.

Chemical formula

[0212] In one embodiment, the overall patient index is calculated, for example, as the total value over all individual hot spot volumes, thereby providing a measure of the total lesion volume, which is the total lesion volume. The total lesion volume may be calculated, for example, as shown in the following equations (9a) and / or (9b).

Chemical formula

[0213] In one embodiment, the overall patient index may be calculated as a function of the intensity, volume, and / or number of voxels within the entire set of (e.g., direct) hot spots (e.g., as a function of all hot spot voxels within the union of all 3D hot spot volumes, e.g., not necessarily as a function of individual hot spot quantification measurements). For example, in one embodiment, the patient index may be an overall average value, e.g., as shown in the following equations (10a) and (10b), (i.e., by summing the intensities of all individual hot spot voxels for the entire set of hot spots L and dividing by the total number of hot spot voxels (for the entire set L)).

Chemical Formula

[0214] In one embodiment, a particular patient index may be calculated using, for example, two or more specific individual hot spot quantification measurements, as follows.

Chemical Formula

[0215] For example, a volume-weighted measurement of intensity may be calculated using both a measurement of hot spot intensity and a measurement of hot spot volume. For example, the weighted total volume may be calculated at the patient level by calculating the product of the lesion index calculated for each individual hot spot and the volume of the hot spot for each individual hot spot. The sum over substantially all weighted volumes may then be calculated, for example, according to the following equation, to determine the total score, where Q LI (l) and Q vol (l) are the individual lesion index and volume values, respectively, for the i-th 3D hot spot volume.

Chemical Formula

[0216] Other measurements of intensity may also be used, for example, as described above, to weight the hot spot volume or to calculate a version of another metric. In certain embodiments, in addition to, or alternatively, the patient index is calculated by multiplying the total lesion volume (e.g., calculated in equation (9a) or (9b)) by the total SUV mean (e.g., calculated in equation (10a) or (10b)) and providing an assessment that combines intensity and volume.

[0217] In certain embodiments, the patient index is, or comprises, the total lesion count calculated as the total number of substantially all detected hot spots (e.g., N L ).

[0218] Region and Lesion Subtype Stratified Patient Index In certain embodiments, in addition to, or alternatively, multiple values of a particular patient index may be calculated, each value being associated with and calculated for a particular subset of the 3D hot spot volume (e.g., as opposed to the set L of substantially all hot spots).

[0219] In particular, in certain embodiments, the 3D hot spot volumes within the set may be arranged / assigned into one or more subsets according to, for example, the particular tissue region in which they are located, or according to subtypes based on a classification scheme such as the miTNM classification. Approaches for grouping hot spots according to tissue region and / or according to an anatomical classification such as miTNM are described in more detail in PCT / EP2020 / 050132, filed Jan. 6, 2020, and PCT / EP2021 / 068337, filed Jul. 2, 2021, the contents of each of which are incorporated herein by reference in their entirety.

[0220] Thus, the value of the patient index as described herein may be calculated for one or more specific tissue regions such as the skeletal region, prostate, or lymphatic region. In certain embodiments, the lymphatic region may be further stratified in a granularity manner, for example, using the approach as described in PCT / EP22 / 77505, filed on October 4, 2022, and published as WO2023 / 057411 on April 13, 2023 (the content of which is incorporated herein by reference in its entirety). Additionally, or alternatively, in certain embodiments, each 3D hot spot volume may be assigned a specific miTNM subtype and grouped into subsets according to the miTNM classification, and values of various patient indices may be calculated for each miTNM classification.

[0221] For example, when hot spots are assigned specific lesion subtypes according to the miTNM staging system, the miTNM class-specific version of the overall patient index described above is used. For example, in certain embodiments, hot spots are identified as local tumor (T), pelvic lymph nodes (N), or distant metastasis (M) (e.g., automatically, based on their location), and assigned labels such as miT, miN, and miM, respectively, to identify three subsets. In certain embodiments, distant metastases may be further subdivided, as appropriate, depending on whether the lesion appears in other sites such as the distant lymph node region (a), bone (b), or another organ (c) (e.g., determined by the hot spot location). Thus, a hot spot may be assigned one of five lesion (e.g., miTNM) classes (e.g., miT, miN, miMa, miMb, miMc). Therefore, each hot spot may be assigned to a specific subset S such that, for example, the value of the patient index P(S) can be calculated for each subset S of hot spots in the image. For example, the following equations (13a-d) can be used to calculate the patient index value for a specific subset of hot spots.

Chemical formula

[0222] In one embodiment, the lesion count may be calculated as the number of substantially all detected hot spots within a specific subset S (e.g., N S ).

[0223] Scaled patient index value In one embodiment, various patient index values may be scaled according to, for example, the physical characteristics of the subject (e.g., weight, height, BMI, etc.) and / or the volume of tissue regions (e.g., the volume of the total skeletal region, prostate volume, total lymph volume, etc.) determined by analyzing the image of the subject (e.g., 3D anatomical image), as appropriate.

[0224] Report the patient index value Turning to FIG. 4A, the patient index value calculated as described herein may be displayed within a report (e.g., an automatically generated report), such as a portion of an electronic document or graphical user interface, for review and verification / approval by a user, for example, in a chart, graph, table, etc.

[0225] In particular, as shown in FIG. 4A, the report 400 generated as described herein may include an overview of the patient index value 402 that quantifies the disease burden within the patient, for example, grouping hot spot subsets according to lesion type (e.g., miTNM classification) and displaying one or more calculated patient index values for that subtype for each lesion type. For example, the overview portion 402 of the report 400 may display patient index values for five subsets of hot spots labeled as miT, miN, miMa (lymph), miMb (bone), and miMc (other) based on the miTNM staging system. For each lesion subtype, the overview table 402 displays the number of detected hot spots belonging to that subtype (e.g., within a particular subset), the maximum SUV (SUV max ), the average SUV (SUV mean ), the total volume, and a quantity referred to as the "aPSMA score". For each lesion subtype S, the values for SUV max , SUV mean , the total volume, and the aPSMA score may be calculated, respectively, as described above, for example, according to equations (13a), (13b), (13c), and (13d). In FIG. 4A, the term "aPSMA score" is used to reflect the use of a PSMA binder such as [18F]DCFPyL for imaging.

[0226] The summary table 402 in FIG. 4A also includes alphanumeric codes (e.g., miTx, miN1a, miM0a, miM1b, miM0c, shown from top to bottom) that characterize the severity, number, and location of lesions in various regions according to the systemic miTNM staging system as described in Siefert et al., “Second Version of the Prostate Cancer Molecular Imaging Standardized Evaluation Framework Including Response Evaluation for Clinical Trials (PROMISE V2),” Eur Urol. 2023 May;83(5):405-412. doi: 10.1016 / j.eururo.2023.02.002 for each lesion type. The notation miTx for the miT (local tumor) subtype uses "x" as a substitute for the various alphanumeric codes used in the miTNM system to indicate, for example, whether the local tumor is unifocal or multifocal, confined to an organ, or invading other adjacent structures such as the seminal vesicles, external sphincter, rectum, bladder, levator ani, pelvic wall, etc., and whether it represents a local recurrence after radical prostatectomy. In certain embodiments, such granular information may not be calculated, for example, due to certain imaging parameters and / or specific anatomically segmented structures. In certain embodiments, additional granular numerical values (e.g., miT2, miT3, miT4) and alphanumeric (e.g., miT2u, miT2m, miT3a, miT3b, miT4, miTr) codings may be calculated (e.g., automatically, based on automated anatomical segmentation) and displayed. In certain embodiments, such codings may be calculated but not displayed in the report (e.g., intentionally) for report simplification / readability purposes (e.g., to avoid an excessive burden on physicians or radiologists), such as in a report like 400. When the level of detail in information such as detailed miTNM (or other staging system) coding information displayed in a high-level report may be limited (e.g., intentionally), the systems and methods described herein may include features for providing additional detail.For example, when providing a report such as report 400 via a graphical user interface, the user may be provided with options to view additional coding information, for example, by clicking (or, for example, tapping within a touch screen device) or by hovering a mouse over a portion of report 400. For example, a click or touch interaction may be used to expand summary table 402 to enable a larger view where additional coding information may be presented, or a click on a specific code such as "miTx" may be used to bring additional information to the foreground (e.g., via a pop-up).

[0227] The generated reports such as report 400 may also include reference values (e.g., SUV uptake rate) 404 determined with respect to a blood pool (e.g., calculated from the aortic region or a part thereof) that quantifies the physiological uptake rate within a patient and various reference organs such as the liver, disease stage codes 406 such as alphanumeric codes based on the miTNM scheme, or information such as other schemes. In certain embodiments, the disease stage expression 406 includes an indication of the specific staging criteria used. For example, as shown in FIG. 4A, the disease stage expression 406 includes the text "miTNM" and, together with the miTNM staging criteria, indicates the use of a specific code determined through the analysis of a specific scan on which report 400 is based.

[0228] The report may additionally or alternatively include a hot spot table 410 that provides, for each hot spot, information such as lesion subtype, lesion location (e.g., a specific tissue volume in which the lesion is located), and values of various individual hot spot quantification measurements as described herein, along with a list of each identified individual hot spot.

[0229] As shown in FIG. 4A, the report may be generated from a single imaging session (e.g., functional and anatomical images such as PET / CT or SPECT / CT images) as appropriate and used to provide a snapshot of the patient's disease at a specific time.

[0230] In certain embodiments, as described in further detail herein, multiple images taken over time may be used to track disease evolution over time. Such information may also be included, for example, within a report or a portion thereof, as shown in FIG. 4B.

[0231] F. Lesion Tracking across Medical Images In certain embodiments, among other things, the image analysis and decision support tools of the present disclosure provide systems and methods for tracking lesions in a patient and assessing disease progression and / or treatment response via the analysis of nuclear medicine images. In particular, in certain embodiments, the approach described herein may be used to analyze longitudinal image data, i.e., a series of medical images collected over time (e.g., two or more images).

[0232] The lesion tracking techniques described herein may be used in conjunction with various medical image types and / or imaging modalities. For example, the medical image may be, or include, an anatomical image. An anatomical image conveys anatomical information about the structures / morphological structures within the subject's body and is acquired using anatomical imaging modalities such as CT, MRI, ultrasound, etc.

[0233] In particular, with respect to tracking lesions across time-series medical images, as described herein, the lesion tracking approach of the present disclosure may, in addition or alternatively, be used to identify lesion correspondence between medical images (e.g., of the same subject) obtained using different contrast agents (e.g., different radiopharmaceuticals), their dosages, image reconstruction techniques, available equipment such as different cameras, combinations thereof, etc.

[0234] Referring to FIG. 5, in certain embodiments, the approach herein may be used when a patient undergoes an initial baseline scan followed by, for example, a follow-up scan (e.g., at a later time) to evaluate, for example, the response to treatment and / or to track the disease.

[0235] In certain embodiments, the medical images analyzed via the approaches described herein are, or include, nuclear medicine images, such as three-dimensional (3D) images, such as bone scan (scintigraphy) images, PET images, and / or SPECT images. In certain embodiments, the nuclear medicine images are complemented (e.g., overlaid) with anatomical images, such as computed tomography (CT) images, x-rays, or MRIs.

[0236] Following an initial baseline scan of a patient, medical images resulting from the scan, such as PET / CT images, are acquired 502 and analyzed, as described herein (e.g., in Sections B and C), to detect and segment 504 hot spots and identify image regions indicative of underlying cancerous lesions in the subject.

[0237] The identified hot spots may be analyzed to calculate 506 values of various individual hot spot quantification measurements and / or patient index measurements, as described herein. As described herein, hot spot quantification measurements may include, for example, intensity measurements (e.g., peak value, average value, median value, etc., of the intensity within a particular hot spot), size measurements (e.g., hot spot volume), and lesion index values that combine both size and intensity and may, for example, give an overall severity of a particular underlying lesion. In certain embodiments, the intensity of one or more reference organs, such as the liver, aorta, parotid gland, etc., may be used to scale the hot spot intensity, enabling calculation of a lesion index value on a standardized scale.

[0238] The individual hot spot quantification measurements may be combined / aggregated to provide an overall picture of the overall risk / disease severity for the entire patient and / or for specific anatomical regions (e.g., prostate, skeletal burden, lymph) and / or tumor classifications (e.g., various classes of lesions according to the miTNM classification or other schemes). For example, the volume of the hot spots may be totaled and / or otherwise aggregated across the entire patient (e.g., or selected regions) to calculate the total lesion volume for a particular patient.

[0239] The values of the hot spot quantification measurements and / or patient-level risk measurements (patient indices) may be used, for example, to provide an initial assessment of the patient and / or may be stored and / or provided for further processing.

[0240] Again, referring to FIG. 5, after a period of time (e.g., following a period of treatment), one or more follow-up images (image at time 2) are acquired 522 as discussed above, hot spots are identified 524, and quantification / risk measurements are calculated 526. The change in one or more of the measurements between the initial image and the image at time 2 is calculated. For example, (i) the change in the number of (automatically and / or semi-automatically) identified lesions may be identified, and / or (ii) the change in the overall volume of the (automatically and / or semi-automatically) identified lesions (e.g., the change in the sum of the volumes of each identified lesion) may be calculated, and / or (iii) the change in the PSMA (e.g., lesion index) weighted total volume (e.g., the sum of the products of the lesion index and the lesion volume for all lesions within the region of interest) may be calculated. Other measurements indicating change may also or alternatively be automatically determined. Similarly, additional follow-up images may be acquired and analyzed in this manner at later time points, e.g., time 3, time 4, etc. This longitudinal dataset for lesion tracking may be used by a healthcare provider, for example, to determine treatment efficacy.

[0241] For example, in one embodiment, the hot spot map is retained with the patient record, and each follow-up map is compared to the baseline map (or previous follow-up map) to identify corresponding (identical) lesions, e.g., identify new lesions, and / or generate lesion-by-lesion longitudinal data to enable tracking of volume, intensity, lesion index score, or other parameters for each lesion. Thus, the methods described herein provide semi-automated and / or automated analysis of medical image data taken over time, creating a longitudinal dataset that provides the reality of how a patient's risk and / or disease evolves over time during surveillance and / or in response to treatment.

[0242] In one embodiment, the methods described herein can be used to calculate metrics that can be used to classify patient disease for treatment / decision-making purposes and / or to stratify groups for clinical trial data collection and analysis. For example, in one embodiment, a change in one or more metrics can be used to classify a patient as belonging to one of three categories, namely, (i) response / partial response, characterized by a PSMA-volume decrease of 30% or more and a decrease in the number of lesions, as shown in FIG. 6A, stable disease (FIG. 6B), characterized by a PSMA-volume decrease of more than 30% but accompanied by the appearance of new lesions, and (ii) progressive disease (FIG. 6C), characterized by an increase in PSMA-volume of 20% or more and the appearance of one or more new lesions, e.g., according to the RECIP classification.

[0243] Aligning a plurality of medical images Turning to FIG. 7, in one embodiment, two or more different medical images may be acquired 702 from the same subject, for example, at different times (e.g., in a time series). Each particular medical image may have a particular hot spot map associated therewith that identifies one or more hot spots within the particular medical image. In one embodiment, the medical images and the associated hot spot maps may be analyzed to identify corresponding hot spots within two or more medical images that are determined to represent the same underlying lesion. In this way, the presence (e.g., appearance and / or disappearance) and / or characteristics of the size / volume of the lesion, radiopharmaceutical uptake rate, etc. may be compared between a plurality of different medical images.

[0244] In one embodiment, the plurality of medical images may be, or may comprise, time series medical images acquired with respect to the same particular subject, with each medical image being acquired, for example, at a different time. Additionally or alternatively, the plurality of medical images may comprise medical images acquired using different contrast agents (e.g., different radiopharmaceuticals), their dosages, image reconstruction techniques, acquisition devices such as different cameras, combinations thereof, etc.

[0245] In one embodiment, a plurality of hot spot maps may be acquired 704. Each hot spot map is associated with a particular medical image and identifies one or more hot spots therein. A hot spot is a region of interest (ROI) identified within a particular medical image and / or its sub-images (e.g., in the case of a composite image) as representing a potential underlying physical lesion within the subject. The hot spot map may identify, for example, a hot spot volume (e.g., a 3D volume) that is determined via partitioning of a 3D image.

[0246] In one embodiment, a hot spot is identified and / or partitioned within a 3D functional image as, for example, a localized region of increased intensity.

[0247] In certain embodiments, the hot spot map may be generated via manual and / or automated detection and / or compartmentalization or combinations thereof. Manual and / or semi-automated approaches may include, for example, receiving user input via an image analysis graphical user interface (GUI). The user may review the rendering of one or more medical images and / or sub-images thereof, with or without various computer-generated annotations such as organ compartmentalizations displayed in combination, and perform actions such as selecting regions to include in and / or exclude from the hot spot map. In certain embodiments, automated hot spot identification and compartmentalization is performed prior to user review to generate a preliminary hot spot map, which is then reviewed by the user to, for example, generate a final hot spot map.

[0248] In certain embodiments, hot spots are classified (e.g., assigned a label) as belonging to a particular anatomical region (e.g., bone, lymph, pelvis, prostate, viscera (e.g., soft tissue organs (other than prostate, lymph), e.g., liver, kidney, spleen, lung, and brain)), and / or to a lesion category such as those of the miTNM classification scheme.

[0249] In certain embodiments, each medical image is compartmentalized to identify a set of organ regions therein and generate a corresponding anatomical compartmentalization map 706. The anatomical compartmentalization map identifies a set of organ regions within a particular medical image, and each member of the set corresponding to a particular organ includes various soft tissue and / or bone regions. As described herein, anatomical compartmentalization may be performed using a machine learning module. The machine learning module may receive an anatomical image as input, analyze the anatomical image, and generate an anatomical compartmentalization map.

[0250] In certain embodiments, the anatomical partitioning maps determined from each medical image may be used to perform image registration. In particular, at least a portion of the set of identified organ regions (e.g., corresponding to one or more of the cervical vertebrae, thoracic vertebrae, lumbar vertebrae, left and right iliac bones, sacrum, and coccyx, left ribs and left scapula, right ribs and right scapula, left thigh, right thigh, skull, brain, and mandible) may be used to determine one or more registration fields that co-register two or more anatomical partitioning maps. Once determined, the one or more registration fields can be used to co-register the medical images from which the anatomical partitioning maps and / or their corresponding hotspot maps 708 were determined.

[0251] For example, referring to FIG. 8, this approach may be used to co-register the first and second medical images and / or their corresponding hotspot maps. In process 800, the first and second medical images are composite images, each comprising an anatomical and a functional image pair (802a / 802b and 804a / 804b).

[0252] The first hotspot map 814 may be generated and / or is generated by detecting and / or partitioning the hotspot 812 within the first functional image 802b that identifies the first set of hotspots within the first medical image. The second hotspot map 824 may be generated and / or is generated by detecting and / or partitioning the hotspot 822 within the second functional image 804b that identifies the second set of hotspots within the second medical image.

[0253] The first anatomical image 802a may be segmented, for example, using a machine learning module, i.e., an anatomical segmentation module, to determine a first anatomical segmentation map 834 that identifies a set of one or more organ regions within the first medical image (i.e., within the first anatomical image and / or the first functional image). The second anatomical image 804a may be segmented, for example, using the anatomical segmentation module, to determine a second anatomical segmentation map 844 that identifies a set of one or more organ regions within the second medical image (i.e., within the second anatomical image and / or the second functional image).

[0254] Full Field Image Registration In some embodiments, the first 834 and second 844 anatomical segmentation maps are used to determine one or more registration fields. The registration field may be calculated (e.g., implemented) based on an affine transformation. For example, in some embodiments, one or more particular subsets of the identified set of organ regions are used as landmarks to register the first and second anatomical segmentation maps. In particular, each particular subset of the identified organ regions may be used to determine a corresponding registration field that aligns the particular subset within the first anatomical segmentation map with the same particular subset within the second anatomical segmentation map. This process may be performed for multiple subsets of the identified organ regions to determine multiple registration fields 850, which can then be combined to generate a final overall registration field that is used to perform the final image registration.

[0255] For example, each subset may comprise an organ region corresponding to a location within a specific anatomical region or portion of the body of interest. For example, as shown in FIGS. 9A and 9B, a first left pelvic region alignment field is determined using a subset of the organ region corresponding to the left pelvic bone of the subject (FIG. 9A), and a second right pelvic region alignment field may be determined using a subset of the organ region corresponding to the right pelvic bone of the subject (FIG. 9B). As shown in FIG. 9C, these two (left and right pelvic region) alignment fields may be combined, for example, via a distance-weighted per-voxel average, whereby each voxel of the final alignment field is calculated as the weighted average of the values for that voxel within the left and right pelvic region alignment fields. For each voxel, the weights for the left and right voxel values used in the average may be determined based on the distance from that voxel to the identification of the left and right pelvic bones, respectively. An example of this alignment approach is described in more detail in PCT / EP22 / 77505, filed Oct. 4, 2022 (published as WO2023 / 057411 on Apr. 13, 2023), with respect to the portion of the image centered on the pelvic region. The approach may then be extended throughout the body of the subject (e.g., each organ subset is associated with a specific part of the body such as the head, neck, chest, abdomen, pelvic region, left side, right side, front, back, etc. and combinations thereof (e.g., left pelvic region, right pelvic region, left anterior chest, right anterior chest, etc.)) to determine a plurality of local alignment fields, each using a specific organ region subset as a landmark, which are then combined (e.g., via a distance-weighted average) to create a final overall alignment field.

[0256] As shown in FIG. 10, this approach may be used to perform accurate whole-body image alignment. For example, FIG. 10 shows a first PET / CT composite image obtained via a first scan and, first, a second PET / CT composite image obtained via a second scan (upper row). Each CT scan shows the identified organ regions of an overlaid anatomical compartmentalization map (colored portions). The lower row of FIG. 10 shows, again, the first PET / CT image together with a transformed version of the second PET / CT image, which is here aligned to the first image via a weighted per-segment affine alignment approach as described herein.

[0257] FIG. 11A shows a schematic view of a second image aligned to a first image, illustrating voxel changes. FIG. 11B illustrates a schematic view of an alignment field with vectors for a subset of voxels. As illustrated in FIG. 11B, in some embodiments, the alignment field comprises reference values for positions (e.g., voxels) in a first image relative to corresponding points (e.g., voxels) in a second image (target voxels in the second image are darkened in FIG. 11B). In some embodiments, an inverted alignment field may be determined. The inverted alignment field comprises reference values for positions (e.g., voxels) in a second image relative to positions (e.g., voxels) in a first image. In some embodiments, an inverted reference field is first generated for each affine alignment. The inverted fields may then be weighted together in the same manner as the affine alignment to generate an inverted alignment field for the whole body.

[0258] In one embodiment, although not wishing to be bound by any particular theory, the first scan resides within a certain space (e.g., within world coordinates), and the second scan resides within a different space. A registration field from the first image space to the second image space is generated by finding a registration that best aligns the organ compartmentalization from the second scan to the organ compartmentalization in the first scan (e.g., via finding local optima in an optimization problem). The registration field can then be applied to any image (e.g., PET, CT, organ compartmentalization, hot spot map) that resides within the same space as the second scan to register it to the space of the first scan.

[0259] Point - by - point registration In addition or alternatively, in one embodiment, the approach described herein can be used to generate a point - by - point registration 850. In one embodiment, point - by - point registration can be used, for example, for triangulation between two PET / CT image stacks taken at two different points in time. In one embodiment, as described herein, the point - by - point registration approach uses "anchor points", which are single - point correspondences as opposed to corresponding masks that identify corresponding 3D tissue regions (e.g., pelvic bone), as described above.

[0260] In one embodiment, the point - by - point registration approach utilizes an anatomical compartmentalization map determined for two different images, e.g., PET / CT images taken at two different points in time for the same patient, to identify a set of anchor points. For example, the set of anchor points can be, or include, the following points: the centroid of all left - side ribs, the centroid of all right - side ribs, the centroid of the left iliac bone, the centroid of the right iliac bone, and the centroid of the thoracic vertebrae. For example, for a particular medical image obtained at a particular point in time, the anatomical compartmentalization map can be used to determine coordinates for each anchor point within a particular set of anchor points. Anchor point coordinates can, therefore, be determined, for example, for each of a plurality of medical images within a time - series of medical images.

[0261] In one embodiment, the point-by-point alignment approach determines a transformation operation, such as a translation, that matches corresponding anchor points between two images. For example, in one embodiment, the set of anchor points may include N anchor points. Coordinate values (e.g., (x, y, z) coordinates in three dimensions) may be calculated for each of the N anchor points in the first and second images, which will be aligned with each other. For each anchor point i in the set, an individual anchor point translation that aligns its location in the first image with its location in the second image

Chemical formula

Chemical formula

[0262] For example, for a particular selected point and the set of N anchor points, a weighted translation

Chemical formula

Chemical formula

Chemical formula

[0263] Looking again at FIGS. 7 and 8, the alignment field and / or per-point alignment 850 determined as described herein may be used to transform the second and / or first hot spot maps 824 and / or 814, respectively, and align them 708, 852 with each other. In this way, the set of hot spots identified in different (e.g., first and second) medical images can be aligned, enabling corresponding hot spots representing the same physical lesion to be accurately identified 710, 854.

[0264] In certain embodiments, additionally or alternatively, the alignment field and / or per-point alignment may be determined as described herein, for example, before the second hot spot map is generated, and used to align the second medical image with the first medical image (e.g., collected at a previous time). The aligned version of the second medical image may be used to generate the second hot spot map, which will be aligned with the first hot spot map generated from the first medical image by virtue of being generated from the aligned version of the second medical image.

[0265] Identifying corresponding hot spots Looking at FIG. 12, in certain embodiments, corresponding hot spots can be identified by calculating one or more lesion correspondence metrics that quantify the proximity and / or similarity between two or more hot spots identified, for example, in different medical images. Exemplary metrics include, but are not limited to, the following.

[0266] Hot spot overlap: In certain embodiments, hot spots within the first and second images that overlap (subsequent to alignment) may be identified as corresponding hot spots for inclusion within a lesion correspondence. In certain embodiments, a relative fraction (percentage) of volume overlap may be calculated and compared to one or more overlap thresholds. Hot spots having an overlap fraction that exceeds a particular threshold (e.g., greater than 20 percent, greater than 30 percent, greater than 40 percent, greater than 50 percent, greater than 70 percent) may be identified as lesion correspondences, as illustrated, for example, in panel A of FIG. 12.

[0267] Hot spot distance: In certain embodiments, as shown, for example, in panel B of FIG. 12, hot spot distance may be calculated as the distance between two points, such as the center of mass (COM) of each hot spot. Hot spot pairs that are separated by a hot spot distance that is less than a particular distance threshold, such as less than 10 mm, less than 20 mm, less than 30 mm, less than 40 mm, less than 50 mm, etc., may be identified as belonging to a lesion correspondence. In certain embodiments, multiple distance thresholds may be used, for example, with respect to different regions. For example, in certain embodiments, a larger threshold (e.g., 50 mm) may be used in the rib / thoracic region to account for respiratory motion, and smaller distance thresholds (e.g., 10 mm, 20 mm, etc.) may be used elsewhere.

[0268] Type / location match: In certain embodiments, each hot spot may be assigned a lesion classification (e.g., miTNM classification) and / or a location (e.g., pelvis, bone, lymph). In certain embodiments, it may be required that a hot spot have a matching lesion classification and / or assigned location in order to be identified as a corresponding hot spot within a lesion correspondence.

[0269] In this way, hot spots appearing in different images can be matched with each other 854 and identified as representing the same underlying physical lesion. The correspondence between such matching hot spots can be encoded via lesion correspondence, which identifies corresponding hot spots in two or more different medical images (e.g., the first and second images can be generated). The lesion correspondence can be bidirectional.

[0270] Lesion tracking measurement In certain embodiments, the systems and methods described herein can be used to calculate measurements 712 that provide for classifying a patient's disease for treatment / decision-making purposes and / or stratifying groups for clinical trial data collection and analysis 714. As described herein, such measurements can include, for example, a total lesion volume calculated as a sum of hot spot volumes across an entire subject, and / or a change thereof, and the number and / or absence of newly identified lesions (or reduction in the number of total lesions), and other measurements, such as various hot spot quantifications and / or patient indices / measurements described herein, for example, in Sections D and E. In certain embodiments, such measurements can be reported, for example, in tabular format or, for example, as a series of graphs or traces within a graph, as shown in FIG. 4B. In certain embodiments, values of normal (non-cancerous) physiological uptake rates may also be displayed, as shown in FIG. 4B.

[0271] In one embodiment, the approach described herein for identifying corresponding hotspots may be used, for example, to match other target regions (e.g., corresponding to other physical features of the subject) identified in different images, such as those collected at different times, from different subjects, with different tracers. Such an approach may be employed to align and identify corresponding target regions identified in different images and to assess the presence, progression, status, response to treatment, etc. of various conditions, such as muscle, ligament, tendon injuries, aneurysms, diagnosis of cancer, assessment of cognitive activity (e.g., via fMRI), and the like, which are not necessarily limited to cancer.

[0272] G. Informing Clinical Decision Making and Treatment Evaluation In one embodiment, measurements calculated based on the analysis of images as described herein may, in turn, be used to determine the values of various measurements indicative of, for example, disease state, progression, prognosis, subject response to therapy, and / or prediction of likely subject response to one or more specific therapies, and / or to stratify subjects accordingly.

[0273] In one embodiment, these measurements may be endpoints in themselves (e.g., measuring the degree of patient function, sensation, or viability), and / or may be correlated therewith, and may be used, for example, in the context of population analysis in clinical trials to assess treatment efficacy, either alone and / or in combination with other markers such as prostate-specific antigen (PSA).

[0274] In certain embodiments, endpoints that can be determined and / or correlated with patient measurements and / or classifications described herein include, but are not limited to, overall survival (OS), radiologic progression-free survival (rPFS), various symptom endpoints (e.g., patient-reported outcomes), disease-free survival (DFS), event-free survival (EFS), objective response rate (ORR), complete response (CR) / partial response (PR) / stable disease (SD) / progressive disease (PD), progression-free survival (PFS), time to progression (TTP), time to radiologic progression.

[0275] In certain embodiments, the various measurements described herein and / or endpoint values determined therefrom can be used to guide treatment decisions. For example, the approaches described herein may be used to identify whether a subject is a responder to a particular treatment, provide an opportunity to discontinue an ineffective treatment, adjust a dosage, or switch to a new therapy at an early stage.

[0276] Thus, among other things, the image analysis and decision support tools described herein may be used to determine prognostic information, measure response to therapy, stratify patients for radioligand therapy, and / or provide predictive information for other therapies.

[0277] For example, in certain embodiments, measurements calculated from the images such as the miTNM classification of individual lesions and / or the overall disease stage (e.g., as shown in FIG. 4A), expression scores, PRIMARY scores, tumor volume (e.g., total tumor volume for a patient and / or stratified by lesion class), presence and / or count of new lesions, etc., as described herein, may be used to calculate a specific response classification. For example, the lesion tracking tool described herein may identify new lesions and quantify increases in tumor size, changes in aPSMA scores (e.g., lesion index scores and / or intensity-weighted total volume as described herein), which may in turn be used to evaluate prostate cancer progression criteria such as the PSMA PET Progression (PPP) score (see, e.g., Fanti et al., "Proposal of Systemic Therapy Response Assessment Criteria in time of PSMA PET / CT imaging: PSMA PET Progression (PPP)," J. Nucl. Med., 2019 https: / / doi.org / 10.2967 / jnumed.119.233817), RECIP criteria scores, and equivalents.

[0278] In certain embodiments, patient index quantification values at single and / or multiple time points may be used as inputs into a prognostic model to determine prognostic measures that indicate and / or quantify the likelihood of a specific clinical event, disease recurrence, or progression in a patient (e.g., having or at risk of prostate cancer). Prognostic measures may include overall survival (OS), radiologic progression-free survival (rPFS), various symptom endpoints (e.g., patient-reported outcomes), disease-free survival (DFS), event-free survival (EFS), objective response rate (ORR), complete response (CR) / partial response (PR) / stable disease (SD) / progressive disease (PD), progression-free survival (PFS), time to progression (TTP), time to radiologic progression.

[0279] The prognostic model may be a statistical model such as regression, and may include additional clinical variable inputs such as patient physical characteristics such as race / ethnicity, prostate-specific antigen (PSA) level and / or velocity, hemoglobin level, lactate dehydrogenase level, albumin level, clinical T stage, biopsy Gleason score, and positive core percentage score. In certain embodiments, the prognostic model compares a calculated value such as a patient index to one or more thresholds, classifies the patients, and / or places them into a "bucket" such as one of a set of ranges such as an OS value. In certain embodiments, the prognostic model may be a machine learning model. For example, various individual hot spot quantification measurements and / or aggregated patient-level indices may be features input into a machine learning model that generates, as output, a predicted value for one or more of the prognostic endpoints described herein. Such a machine learning model may be, for example, an artificial neural network (ANN). The machine learning model may also include clinical variables as inputs (i.e., features).

[0280] For example, in certain embodiments, quantitative measurements of disease burden from a single time point are used to calculate values of overall tumor volume, overall measure of intensity such as total SUV mean / max / peak and patient-level measurements such as an aPSMA score (e.g., intensity-weighted total volume). These measurements are used as inputs into the prognostic model and may generate, as output, one or more of expected survival rate (e.g., in months), time to progression (TTP), and time to radiological progression.

[0281] In certain embodiments, quantitative data over multiple time points such as change in total lesion volume, SUV, aPSMA score, and measurements over time of lesion changes (e.g., number of new lesions, number of disappeared lesions, number of lesions followed) are used as inputs into the prognostic model and may generate, as output, one or more of expected survival rate (e.g., in months), time to progression, and time to radiological progression.

[0282] In certain embodiments, in addition to, or alternatively, for example, the characteristics of PSMA expression in the prostate (and / or other tissue regions, which may be identified, for example, via the anatomical compartmentalization techniques described herein) can be used as an input into a prognostic model. For example, the spatial intensity pattern within a particular tissue region (e.g., from the intensity of a functional image such as a PET or SPECT image) can be used, alone and / or together with the quantitative measurements and clinical variables described herein, as an input into a machine learning module to generate predictions such as the risk of synchronous metastases, the risk of future (metachronous) metastases, etc. For example, data from the lesion tracking techniques described herein can be used as an input to improve prediction techniques such as those described in U.S. Patent No. 11,564,621, the content of which is incorporated herein by reference in its entirety. In certain embodiments, the intensity pattern can be used, for example, for each image of the subject at a particular time point, to determine a score such as a PRIMARY score or a similar score as described in Siefert et al., “Second Version of the Prostate Cancer Molecular Imaging Standardized Evaluation Framework Including Response Evaluation for Clinical Trials (PROMISE V2),” Eur Urol. 2023 May; 83(5):405-412. doi:10.1016 / j.eururo.2023.02.002. Such automatically calculated intensity scores can be included, for example, within a patient report such as that shown in FIG. 4A.

[0283] In one embodiment, the approach described herein may be used to generate a model for categorizing patient responses to therapy. For example, the lesion tracking techniques described herein may be used to determine inputs such as changes in tumor volume, intensity, appearance / disappearance of lesions, etc. These inputs may be used, via one or more response models, to determine whether a patient responds to treatment (e.g., a yes / no classification) and / or the extent to which a patient responds to treatment (e.g., a numerical value). As described herein, such an approach may leverage existing response criteria such as RECIP and PPP, which currently rely on variable and time-consuming manual radiologist readings, and may thus be improved by the present technique, improving accuracy, robustness (e.g., uniformity across different operators, imaging sites, etc.), and speed of patient staging and response assessment to therapy.

[0284] In one embodiment, the approach described herein may be used to assess the likelihood that a patient is suffering from favorable benefits and / or unfavorable effects (e.g., that may be associated with high cost and / or harmful side effects) from a particular treatment. For example, software may be used to provide an indication of whether a patient is likely to benefit from a particular radioligand therapy. Thus, the approach described herein may provide an indication of whether a patient is likely to benefit from a radioligand therapy (e.g., Pluvicto TM) meets the significant unmet needs in [description], and can help physicians, especially in the late stages of the disease, navigate among multiple and increasing therapies. For example, for a set of possible treatments (e.g., abiraterone, enzalutamide, apalutamide, darolutamide, sipuleucel-T, Ra223, docetaxel, cabazitaxel, pembrolizumab, olaparib, rucaparib, 177Lu-PSMA617, etc.), the prediction model receives, as input, various imaging measurements described herein, and as output, generates a score indicating the likelihood that a patient will respond positively to a treatment (or a specific therapy class, e.g., androgen biosynthesis inhibitors (e.g., abiraterone), androgen receptor inhibitors (e.g., enzalutamide, apalutamide, darolutamide, etc. treatment classes), cellular immunotherapy (e.g., sipuleucel-T), internal radiation therapy (e.g., Ra223), anti-neoplastic drugs (e.g., docetaxel, cabazitaxel), immune checkpoint inhibitors (pembrolizumab), PARP inhibitors (e.g., olaparib, rucaparib), PSMA binders (e.g., radioligand therapy, e.g., with Lu177) for each.

[0285] H. Contrast agent As described herein, various radionuclide-labeled PSMA binders may be used as radiopharmaceutical contrast agents for nuclear medicine imaging to detect and evaluate prostate cancer. In certain embodiments, one radionuclide-labeled PSMA binder is suitable for PET imaging, while others are suitable for SPECT imaging.

[0286] i. PET imaging radionuclide-labeled PSMA binder In certain embodiments, the radionuclide-labeled PSMA binder is a radionuclide-labeled PSMA binder suitable for PET imaging.

[0287] In certain embodiments, the radionuclide-labeled PSMA binder is [18F]DCFPyL (also referred to as PyL TM and also referred to as DCFPyL-18F), i.e., [Chemical formula] or a pharmaceutically acceptable salt thereof.

[0288] In certain embodiments, the radionuclide-labeled PSMA binder is [18F]DCFBC, i.e.,

Chem.

[0289] In certain embodiments, the radionuclide-labeled PSMA binder is 68 Ga-PSMA-HBED-CC ( 68 also referred to as Ga-PSMA-11), i.e.,

Chem.

[0290] In certain embodiments, the radionuclide-labeled PSMA binder is PSMA-617, i.e.,

Chem.

[0291] In certain embodiments, the radionuclide-labeled PSMA binder is PSMA-I&T, i.e.,

Chem.

[0292] In certain embodiments, the radionuclide-labeled PSMA binder is PSMA-1007, namely,

Chemical Structure

[0293] In certain embodiments, the radionuclide-labeled PSMA binder is 18F-JK-PSMA-7, namely,

Chemical Structure

[0294] ii. SPECT imaging radionuclide-labeled PSMA binder In certain embodiments, the radionuclide-labeled PSMA binder is a radionuclide-labeled PSMA binder suitable for SPECT imaging.

[0295] In certain embodiments, the radionuclide-labeled PSMA binder is 1404 (also referred to as MIP-1404), namely,

Chemical Structure

[0296] In certain embodiments, the radionuclide-labeled PSMA binder is 1405 (also referred to as MIP-1405), namely,

Chemical Structure

[0297] In one embodiment, the radionuclide-labeled PSMA binder is 1427 (also referred to as MIP-1427), i.e., [Chemical formula] or a pharmaceutically acceptable salt thereof.

[0298] In one embodiment, the radionuclide-labeled PSMA binder is 1428 (also referred to as MIP-1428), i.e., [Chemical formula] or a pharmaceutically acceptable salt thereof.

[0299] In one embodiment, the PSMA binder is labeled with a radionuclide by chelating it with a radioactive isotope of a metal [e.g., a radioactive isotope of technetium (Tc) (e.g., technetium-99m ( 99m Tc)), e.g., a radioactive isotope of rhenium (Re) (e.g., rhenium-188 ( 188 Re), e.g., rhenium-186 ( 186 Re)), e.g., a radioactive isotope of yttrium (Y) (e.g., 90 Y), e.g., a radioactive isotope of lutetium (Lu) (e.g., 177 Lu), e.g., a radioactive isotope of gallium (Ga) (e.g., 68 Ga, e.g., 67 Ga), e.g., a radioactive isotope of indium (e.g., 111 In), e.g., a radioactive isotope of copper (Cu) (e.g., 67 Cu)].

[0300] In one embodiment, 1404 is labeled with a radionuclide (e.g., chelated with a radioactive isotope of a metal). In one embodiment, the radionuclide-labeled PSMA binder is 99m 1404 labeled (e.g., chelated) with Tc, 99m Tc-MIP-1404, i.e., [Chemical formula] or a pharmaceutically acceptable salt thereof. In certain embodiments, 1404 is chelated to other metal radioisotopes [e.g., radioisotopes of rhenium (Re) (e.g., rhenium-188 ( 188 Re), e.g., rhenium-186 ( 186 Re)), e.g., radioisotopes of yttrium (Y) (e.g., 90 Y), e.g., radioisotopes of lutetium (Lu) (e.g., 177 Lu), e.g., radioisotopes of gallium (Ga) (e.g., 68 Ga, e.g., 67 Ga), e.g., radioisotopes of indium (e.g., 111 In), e.g., radioisotopes of copper (Cu) (e.g., 67 Cu)] and forms a compound having a structure similar to that shown above for 99m Tc-MIP-1404, with other metal radioisotopes used in place of 99m Tc.

[0301] In certain embodiments, 1405 is labeled with a radionuclide (e.g., chelated to a radioisotope of a metal). In certain embodiments, the radionuclide-labeled PSMA binder is 99m 1405 labeled (e.g., chelated) with 99m Tc, i.e., [Chemical formula] or a pharmaceutically acceptable salt thereof. In certain embodiments, 1405 is chelated to other metal radioisotopes [e.g., radioisotopes of rhenium (Re) (e.g., rhenium-188 ( 188 Re), e.g., rhenium-186 ( 186 Re)), e.g., radioisotopes of yttrium (Y) (e.g., 90 Y), e.g., radioisotopes of lutetium (Lu) (e.g., 177Lu), for example, a radioisotope of gallium (Ga) (e.g., 68 Ga, for example, 67 Ga), for example, a radioisotope of indium (e.g., 111 In), for example, a radioisotope of copper (Cu) (e.g., 67 Cu)] chelated thereto, 99m accompanied by other metal radioisotopes used in place of 99m Tc-MIP-1405, may form a compound having a structure similar to the structure shown above.

[0302] In certain embodiments, 1427 is labeled (e.g., chelated thereto) with a radioisotope of a metal and is a compound according to the following chemical formula, namely,

Chemical formula

[0303] In certain embodiments, 1428 is labeled (e.g., chelated thereto) with a radioisotope of a metal and is a compound according to the following chemical formula, namely,

Chemical formula

[0304] In certain embodiments, the radionuclide-labeled PSMA binder is PSMA I&S, i.e.,

Chemical formula

[0305] I. Computer Systems and Network Environments Certain embodiments described herein utilize computer algorithms in the form of software instructions, executed by a computer processor. In certain embodiments, the software instructions include a machine learning module, also referred to herein as artificial intelligence software. As used herein, a machine learning module refers to a computer-implemented process (e.g., a software function) that implements one or more specific machine learning techniques, such as artificial neural networks (ANNs), such as convolutional neural networks (CNNs), such as recurrent neural networks, such as regression neural networks such as long short-term memory (LSTM) or bidirectional long short-term memory (Bi-LSTM), random forests, decision trees, support vector machines, and equivalents, to determine one or more output values for a given input.

[0306] In one embodiment, a machine learning module implementing machine learning techniques is trained using a dataset that includes, for example, categories of data (e.g., CT images, MRI images, PET images, SPECT images) described herein. Such training may be used to determine various parameters of a machine learning algorithm implemented by the machine learning module, such as weights, associated with layers within a neural network. In one embodiment, once the machine learning module is trained to perform a specific task, such as, for example, partitioning an anatomical region, partitioning and / or classifying hot spots, or determining values related to prognosis, treatment response, and / or predictive metrics, the determined parameter values are fixed and the (e.g., invariant, static) machine learning module processes new data (e.g., different from the training data) and is used to perform its trained task without further updating its parameters (e.g., without the machine learning module receiving feedback and / or updates). In one embodiment, the machine learning module may receive feedback, for example, based on a user review of accuracy, and such feedback may be used as additional training data to dynamically update the machine learning module. In one embodiment, two or more machine learning modules may be combined and implemented as a single module and / or a single software application. In one embodiment, two or more machine learning modules may also be implemented separately, for example, as separate software applications. The machine learning module may be software and / or hardware. For example, the machine learning module may be implemented entirely as software, or certain functions of the ANN module may be implemented via special hardware (e.g., via an application specific integrated circuit (ASIC)).

[0307] As shown in FIG. 13, an implementation of a network environment 1300 for use in providing the systems, methods, and architectures described herein is shown and described. In general overview, referring now to FIG. 13, a block diagram of an exemplary cloud computing environment 1300 is shown and described. The cloud computing environment 1300 may include one or more resource providers 1302a, 1302b, 1302c (collectively, 1302). Each resource provider 1302 may include computing resources. In some implementations, the computing resources may include any hardware and / or software used to process data. For example, the computing resources may include hardware and / or software capable of executing algorithms, computer programs, and / or computer applications. In some implementations, exemplary computing resources may include application servers and / or databases with storage and read capabilities. Each resource provider 1302 may be connected to any other resource provider 1302 within the cloud computing environment 1300. In some implementations, the resource providers 1302 may be connected via a computer network 1308. Each resource provider 1302 may be connected via the computer network 1308 to one or more computing devices 1304a, 1304b, 1304c (collectively, 1304).

[0308] The cloud computing environment 1300 may include a resource manager 1306. The resource manager 1306 may be connected to a resource provider 1302 and a computing device 1304 via a computer network 1308. In some implementations, the resource manager 1306 may facilitate the provisioning of computing resources from one or more resource providers 1302 to one or more computing devices 1304. The resource manager 1306 may receive requests for computing resources from a particular computing device 1304. The resource manager 1306 may identify one or more resource providers 1302 that are capable of providing the computing resources requested by the computing device 1304. The resource manager 1306 may select a resource provider 1302 to provide the computing resources. The resource manager 1306 may facilitate the connection between the resource provider 1302 and a particular computing device 1304. In some implementations, the resource manager 1306 may establish a connection between a particular resource provider 1302 and a particular computing device 1304. In some implementations, the resource manager 1306 may redirect a particular computing device 1304 to a particular resource provider 1302 with the requested computing resources.

[0309] Figure 14 shows an example of a computing device 1400 and a mobile computing device 1450 that may be used to implement the techniques described in this disclosure. The computing device 1400 is intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The mobile computing device 1450 is intended to represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, and other similar computing devices. The components shown here, their connections and relationships, and their functions are intended only as examples and are not intended to be limiting.

[0310] The computing device 1400 includes a processor 1402, a memory 1404, a storage device 1406, a high-speed interface 1408 that is connected to the memory 1404 and a plurality of high-speed expansion ports 1410, and a low-speed interface 1412 that is connected to a low-speed expansion port 1414 and the storage device 1406. Each of the processor 1402, the memory 1404, the storage device 1406, the high-speed interface 1408, the high-speed expansion ports 1410, and the low-speed interface 1412 is interconnected using various buses and may be mounted on a common motherboard or in other manners as appropriate. The processor 1402 can process instructions for execution within the computing device 1400, including instructions stored in the memory 1404 or on the storage device 1406 so as to display graphical information for a GUI on an external input / output device such as a display 1416 coupled to the high-speed interface 1408. In other implementations, multiple processors and / or multiple buses may be used as appropriate, along with multiple memories and types of memory. Also, multiple computing devices may be connected, and each device may provide a portion of the necessary operations (e.g., as a server bank, a blade server group, or a multi-processor system). Thus, when the term is used herein and multiple functions are described as being performed by a "processor", this encompasses embodiments in which the multiple functions are performed by any number (one or more) of processors of any number (one or more) of computing devices. Further, when a function is described as being performed by a "processor", this encompasses embodiments in which the function is performed by any number (one or more) of processors of any number (one or more) of computing devices (e.g., within a distributed computing system).

[0311] Memory 1404 stores information within computing device 1400. In some implementations, memory 1404 is a plurality of volatile memory units or a plurality of units. In some implementations, memory 1404 is a non-volatile memory unit or a plurality of units. Memory 1404 may also be another form of computer-readable medium, such as a magnetic or optical disk.

[0312] Storage device 1406 can provide a mass storage device for computing device 1400. In some implementations, storage device 1406 is a computer-readable medium, or contains a computer-readable medium, such as a floppy (registered trademark) disk device, a hard disk device, an optical disk device, or a tape device, a flash memory or other similar solid-state memory device, or an array of devices including a device within a storage area network or other configuration. Instructions can be stored within an information carrier. When executed by one or more processing devices (e.g., processor 1402), the instructions implement one or more methods such as those described above. Instructions can also be stored by one or more storage devices such as a computer or machine-readable medium (e.g., memory 1404, storage device 1406, or memory on processor 1402).

[0313] While the high-speed interface 1408 manages the bandwidth-intensive operations for the computing device 1400, the low-speed interface 1412 manages the lower-bandwidth-intensive operations. Such a distribution of functions is merely an example. In some implementations, the high-speed interface 1408 is coupled to the memory 1404, the display 1416 (e.g., through a graphics processor or accelerator), and the high-speed expansion port 1410 into which various expansion cards (not shown) can be plugged. In this implementation, the low-speed interface 1412 is coupled to the storage device 1406 and the low-speed expansion port 1414. The low-speed expansion port 1414, which may include various communication ports (e.g., USB, Bluetooth®, Ethernet®, wireless Ethernet®), may be coupled to one or more input / output devices such as a keyboard, a pointing device, a scanner, or a networking device such as a switch or router (e.g., through a network adapter).

[0314] As shown in the figure, the computing device 1400 may be implemented in several different forms. For example, this may be implemented as a standard server 1420 or multiple times among such a group of servers. Additionally, this may be implemented within a personal computer such as a laptop computer 1422. This may also be implemented as part of a rack server system 1424. Alternatively, components from the computing device 1400 may be combined with other components within a mobile device (not shown) such as a mobile computing device 1450. Each of such devices may contain one or more of the computing device 1400 and the mobile computing device 1450, and the entire system may be composed of multiple computing devices that communicate with each other.

[0315] The mobile computing device 1450 includes, among other components, a processor 1452, a memory 1464, input / output devices such as a display 1454, a communication interface 1466, and a transceiver 1468. The mobile computing device 1450 may also be provided with a storage device such as a microdrive or other device to provide additional storage. Each of the processor 1452, the memory 1464, the display 1454, the communication interface 1466, and the transceiver 1468 is interconnected using various buses, and some of the components may be mounted on a common motherboard or in other suitable manners as appropriate.

[0316] The processor 1452 can execute instructions within the mobile computing device 1450, including instructions stored in the memory 1464. The processor 1452 may be implemented as a chipset of chips including separate and multiple analog and digital processors. The processor 1452 may provide coordination of other components of the mobile computing device 1450, such as, for example, a user interface, applications launched by the mobile computing device 1450, and control of wireless communication by the mobile computing device 1450.

[0317] Processor 1452 may communicate with a user through a display interface 1456 coupled to a control interface 1458 and a display 1454. The display 1454 may be, for example, a TFT (Thin Film Transistor Liquid Crystal Display) display, or an OLED (Organic Light Emitting Diode) display, or other suitable display technology. The display interface 1456 may comprise appropriate circuitry for driving the display 1454 to present graphical and other information to the user. The control interface 1458 may receive commands from the user and convert them for submission to the processor 1452. Additionally, an external interface 1462 may provide communication with the processor 1452 to enable short-range communication of the mobile computing device 1450 with other devices. The external interface 1462 may provide, for example, wired communication in some implementations, or wireless communication in other implementations, and multiple interfaces may also be used.

[0318] Memory 1464 stores information within mobile computing device 1450. Memory 1464 can be implemented as one or more than one of a computer-readable medium or media, a volatile memory unit or units, or a non-volatile memory unit or units. Extended memory 1474 may also be provided and may be connected to mobile computing device 1450 through an extended interface 1472, which may include, for example, a SIMM (Single In-line Memory Module) card interface. Extended memory 1474 may provide additional storage space for mobile computing device 1450 or may store applications or other information for mobile computing device 1450. Specifically, extended memory 1474 may include instructions for performing or complementing the processes described above and may also include secure information. Thus, for example, extended memory 1474 may be provided as a security module for mobile computing device 1450 and may be programmed with instructions that enable secure use of mobile computing device 1450. In addition, secure applications may be provided via the SIMM card in addition to additional information such as placing identification information on the SIMM card in a non-hackable manner.

[0319] The memory may include, for example, flash memory and / or NVRAM memory (non-volatile random access memory) as discussed below. In some implementations, instructions are stored within an information carrier. When executed by one or more processing devices (e.g., processor 1452), the instructions implement one or more methods such as those described above. The instructions may also be stored by one or more storage devices such as one or more computer or machine-readable media (e.g., memory 1464, extended memory 1474, or memory on processor 1452). In some implementations, the instructions can be received within a propagated signal, for example, via transceiver 1468 or external interface 1462.

[0320] Mobile computing device 1450 may communicate wirelessly through communication interface 1466, which may include digital signal processing circuitry if necessary. Communication interface 1466 may provide communication under various modes or protocols, such as, among others, GSM (registered trademark) voice calls (pan-European digital cellular telephone system), SMS (Short Message Service), EMS (Enhanced Messaging Service), or MMS messaging (Multimedia Messaging Service), CDMA (Code Division Multiple Access), TDMA (Time Division Multiple Access), PDC (Personal Digital Cellular), WCDMA (registered trademark) (Wideband Code Division Multiple Access), CDMA2000, or GPRS (General Packet Radio Service). Such communication may occur, for example, using radio frequencies through transceiver 1468. Additionally, short-range communication may occur using Bluetooth (registered trademark), Wi-Fi TM , or other such transceivers (not shown). Additionally, a GPS (Global Positioning System) receiver module 1470 may provide additional navigation and location-related wireless data to mobile computing device 1450, which may be used as appropriate by applications running on mobile computing device 1450.

[0321] Mobile computing device 1450 may also communicate audibly using an audio codec 1460 that can receive oral information from a user and convert it into usable digital information. The audio codec 1460 may similarly generate audible sounds for the user, for example, through a speaker or the like within a handset of the mobile computing device 1450. Such sounds may include sounds from a voice call, may include recorded sounds (such as voice messages, music files, etc.), and may also include sounds generated by an application operating on the mobile computing device 1450.

[0322] As shown in the figure, mobile computing device 1450 may be implemented in several different forms. For example, it may be implemented as a mobile phone 1480. It may also be implemented as part of a smartphone 1482, a personal digital assistant, or other similar mobile device.

[0323] The various implementations of the systems and techniques described herein can be realized in digital electronic circuitry, integrated circuitry, specially designed ASICs (application specific integrated circuits), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementations in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor coupled to receive data and instructions from, and to transmit data and instructions to, a memory system, at least one input device, and at least one output device. These various implementations may be either special purpose or general purpose.

[0324] These computer programs (also known as programs, software, software applications, or code) include machine instructions for a programmable processor and can be implemented in high-level procedural and / or object-oriented programming languages and / or in assembly / machine language. As used herein, the terms "machine-readable medium" and "computer-readable medium" refer to any computer program product, apparatus, and / or device (e.g., magnetic disks, optical disks, memory, programmable logic device (PLD)) used to provide machine instructions and / or data to a programmable processor, including a machine-readable medium that receives the machine instructions as a machine-readable signal. The term "machine-readable signal" refers to any signal used to provide machine instructions and / or data to a programmable processor.

[0325] To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or trackball) by which the user can provide input to the computer. Other kinds of devices can also be used to provide for interaction with a user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback), and input received from the user can be in any form, including acoustic, speech, or tactile input.

[0326] The systems and techniques described herein can be implemented within a computing system that includes back-end components (such as a data server), or middleware components (such as an application server), or front-end components (such as a client computer having a graphical user interface, or a web browser through which a user can implement and interact with the systems and techniques described herein), or any combination of such back-end, middleware, or front-end components. The components of the system can be interconnected by digital data communication in any form or medium (such as a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0327] A computing system can include a client and a server. The client and server are generally remote from each other and typically interact through a communication network. The relationship between the client and the server arises from computer programs that are launched on separate computers and have a client / server relationship with each other.

[0328] In some implementations, the various modules described herein can be separated, combined, or incorporated into single or combined modules. The modules depicted in the figures are not intended to limit the systems described herein to the software architecture shown therein.

[0329] Elements of different implementations described herein can be combined to form other implementations not specifically recited above. Elements can be excluded from the processes, computer programs, databases, etc. described herein without adversely affecting their operation. Additionally, the logical flows depicted in the figures do not require a particular order or sequential order to achieve the desired result. Various distinct elements may be combined into one or more individual elements to perform the functions described herein.

[0330] Throughout the description, where an apparatus and system are described as having, including, or comprising specific components, or where a process and method are described as having, including, or comprising specific steps, it is contemplated that there exist apparatus and systems of the invention that consist essentially of, or consist of, the recited components, and that there exist processes and methods according to the invention that consist essentially of, or consist of, the recited process steps.

[0331] It should be understood that the order of steps or the order for performing certain actions is not important as long as the invention remains operable. Also, two or more steps or actions may be performed simultaneously.

[0332] Although the invention has been specifically shown and described with reference to specific preferred embodiments, it should be understood by those skilled in the art that various changes in form and detail can be made therein without departing from the spirit and scope of the invention as defined by the appended claims. (Item 1) A method for automatically processing a 3D image of a subject and determining a value of one or more patient indices (plural indices) for measuring a (e.g., overall) disease burden and / or risk associated with the subject, the method comprising: (a) A step of receiving, by a processor of a computing device, the 3D functional image of the object obtained using a functional imaging modality; (b) A step of partitioning, by the processor, a plurality of 3D hot spot volumes within the 3D functional image, wherein each 3D hot spot volume corresponds to a local region of increased intensity relative to its surroundings and represents a potential cancerous lesion within the object, thereby obtaining a set of 3D hot spot volumes; (c) A step of calculating, by the processor, a value of the specific individual hot spot quantification measurement for each individual 3D hot spot volume of the set for each specific one of one or more individual hot spot quantification measurements; (d) A step of determining, by the processor, a value of the one or more patient indices, wherein at least some of the patient indices are each associated with one or more specific individual hot spot quantification measurements and are a function of at least some (e.g., substantially all, e.g., a specific subset) of the values of the one or more specific individual hot spot quantification measurements calculated with respect to the set of 3D hot spot volumes; A method comprising the above steps. (Item 2) The method according to Item 1, wherein at least one specific patient index of the one or more patient index values is associated with a single specific individual hot spot quantification measurement and is calculated as a function (e.g., mean value, median value, mode value, sum value, etc.) of substantially all (e.g., all, e.g., excluding only statistical outliers) of the values of the specific individual hot spot quantification measurement calculated with respect to the set of 3D hot spot volumes. (Item 3) The above single specific individual hot spot quantification measurement value is the method according to item 2, which is an individual hot spot intensity measurement value for quantifying the intensity within the 3D hot spot volume (calculated for each individual 3D hot spot volume, for example, as a function of the intensity of the voxels within the above 3D hot spot volume). (Item 4) The above individual hot spot intensity measurement value is the method according to item 3, which is the average hot spot intensity (calculated for each individual 3D hot spot volume, for example, as the average value of the intensities of the voxels within the above 3D hot spot volume). (Item 5) The above specific patient index is calculated as the sum of substantially all the values of the above individual hot spot intensity measurement values calculated for the set of the above 3D hot spot volumes, according to any one of items 3-4. (Item 6) The above single specific individual hot spot quantification measurement value is the method according to item 6, which is the lesion volume (calculated for a specific 3D hot spot volume, for example, as the sum of the volumes of each individual voxel within the above specific 3D hot spot volume). (Item 7) The value of the above specific patient index is calculated as the sum of substantially all the above lesion volume values calculated for the set of the above 3D hot spot volumes (for example, such that the above specific patient index value provides a measurement of the total lesion volume within the above object), according to the method described in item 6. (Item 8) A specific one of the above one or more overall patient indices (plural indices) is associated with two or more specific individual hot spot quantification measurement values, and is calculated as a function (for example, a weighted sum, a weighted average, etc.) of substantially all the values of the two or more specific individual hot spot quantification measurement values calculated for the set of the above 3D hot spot volumes, according to any one of the above items. (Item 9) The above two or more specific individual hot spot quantification measurement values are the method according to item 8, comprising (i) an individual hot spot intensity measurement value and (ii) a lesion volume. (Item 10) The above individual hot spot intensity measurement values are the method according to item 9, which are individual lesion indices that map the values of the above hot spot intensity to values on a standardized scale. (Item 11) The above specific patient index (value) is For each of the individual 3D hot spot volumes of substantially all of the above 3D hot spot volumes, weighting the value of the above lesion volume by the value of the above individual hot spot intensity measurement value (for example, calculating the product of the above lesion volume value and the value of the above individual hot spot intensity measurement value), thereby calculating a plurality of intensity-weighted lesion volumes; Calculating the total value of substantially all of the above intensity-weighted lesion volumes as the value of the above specific patient index; The method according to item 9 or 10, which is calculated as the total value of the intensity-weighted lesion (for example, hot spot) volume. (Item 12) The above one or more specific individual hot spot quantification measurement values are the method according to any one of the above items, comprising one or more individual hot spot intensity measurement values that quantify the intensity within the 3D hot spot volume (for example, calculated for the individual 3D hot spot volume as a function of the intensity of the voxels of the above 3D hot spot volume). (Item 13) The above one or more specific individual hot spot quantification measurement values are (For example, calculated for a specific 3D hot spot volume as the average value of the intensities of the voxels within the above specific 3D hot spot volume) an average hot spot intensity, and (For example, calculated for a specific 3D hot spot volume as the maximum value of the intensities of the voxels within the above specific 3D hot spot volume) a maximum hot spot intensity, (e.g., calculated for a particular 3D hot spot volume as the median intensity of the voxels within the 3D hot spot volume) central hot spot intensity and The method according to item 12, comprising one or more members selected from the group consisting of (Item 14) The one or more individual hot spot intensity measurements include the peak intensity of the 3D hot spot volume [e.g., for a particular 3D hot spot volume, the value of the peak intensity is (i) identifying the maximum intensity voxel within the particular 3D hot spot volume; and (ii) identifying the voxels within a sub-region centered on the maximum intensity voxel (e.g., comprising voxels within a particular threshold distance thereof) and within the particular 3D hot spot; and (iii) calculating the average value of the intensities of the voxels within the sub-region as the corresponding peak intensity calculated by) The method according to item 12 or 13. (Item 15) The one or more individual hot spot intensity measurements include an individual lesion index that maps the value of the hot spot intensity to a value on a standardized scale, according to any one of items 12 - 14. (Item 16) identifying, by the processor, in the 3D functional image, one or more 3D reference volumes, each corresponding to a particular reference tissue region; and determining, by the processor, one or more reference intensity values, each associated with a particular 3D reference volume of the one or more 3D reference volumes and corresponding to a measured value of intensity within the particular 3D reference volume; and In step (c), for each 3D hot spot volume in the set, The step of determining corresponding values (e.g., average hot spot intensity, central hot spot intensity, maximum hot spot intensity, etc.) of specific individual hot spot intensity measurement values by the above processor, and The step of determining corresponding values of the individual lesion indices by the above processor based on the corresponding values of the specific individual hot spot intensity measurement values and the corresponding values of the one or more reference intensity values above it The method according to item 15, comprising: (Item 17) The step of mapping each of the one or more reference intensity values above it to corresponding reference index values on a scale, and For each 3D hot spot volume, using the reference intensity value and the corresponding reference index value to determine the corresponding value of the individual lesion index, and interpolating the corresponding individual lesion index value onto the above scale based on the corresponding value of the specific individual hot spot intensity measurement value The method according to item 16, comprising: (Item 18) The method according to any one of items 16 or 17, wherein the above reference tissue region comprises one or more members selected from the group consisting of the liver, aorta, and parotid gland. (Item 19) The first reference intensity value is (i) a blood reference intensity value associated with a reference volume corresponding to an aortic portion and (ii) mapped to a first reference index value, The second reference intensity value is (i) a liver reference intensity value associated with a reference volume corresponding to the liver and (ii) mapped to a second reference index value, The above second reference intensity value is greater than the above first reference intensity value, and the above second reference index value is greater than the above first reference index value, The method according to any one of items 16 - 18. (Item 20) The above reference intensity value has a maximum reference intensity value mapped to the maximum reference index value, and is a 3D hot spot volume. For the 3D hot spot volume where the corresponding value of the specific individual hot spot intensity measurement value exceeds the maximum reference intensity value, an individual lesion index value equal to the maximum reference index value is assigned. The method according to any one of items 16 - 19. (Item 21) In the set of the above 3D hot spot volumes, identifying one or more subsets each associated with a specific tissue region and / or lesion classification; For each specific subset of the one or more subsets, using the values of the individual hot spot quantification measurement values calculated for the 3D hot spot volumes within the specific subset to calculate the corresponding values of one or more specific patient indices (plural indices); The method according to any one of the above items, including. (Item 22) Each of the one or more subsets is associated with a specific one of the one or more tissue regions. The method according to item 21 includes identifying a subset of the 3D hot spot volumes located within the volume of interest corresponding to the specific tissue region for each specific tissue region. (Item 23) The one or more tissue regions include one or more members selected from the group consisting of a skeletal region, a lymph region, and a prostate region, each having one or more bones of the subject. The method according to item 22. (Item 24) Each of the one or more subsets is associated with a particular one of one or more lesion subtypes [e.g., according to a lesion classification scheme (e.g., miTNM classification)], and the method includes, for each 3D hot spot volume, determining the corresponding lesion subtype and assigning the 3D hot spot volume to the one or more subsets according to their corresponding lesion subtypes. The method according to any one of items 21-23. (Item 25) Using at least a portion of the value(s) of the one or more patient indices as an input to a prognostic model (e.g., a statistical model such as regression, e.g., a classification model by which a patient is assigned to a particular class based on a comparison of the one or more patient index values to one or more thresholds, e.g., a machine learning model that receives the value(s) of the one or more patient indices as an input) to generate, as an output, an expected value and / or range (e.g., a class) (e.g., in terms of time, e.g., representing expected survival rate, time to progression, time to radiological progression, etc. in months) indicating a value with a high likelihood of a particular patient outcome. The method according to any one of the above items. (Item 26) Using at least a portion of the value(s) of the one or more patient indices as an output, one or more treatment options (e.g., abiraterone, enzalutamide, apalutamide, darolutamide, sipuleucel-T, Ra223, docetaxel, cabazitaxel, pembrolizumab, olaparib, rucaparib, 177For each of Lu-PSMA-617, etc. and / or classes of therapeutic agents [e.g., androgen biosynthesis inhibitors (e.g., abiraterone), androgen receptor inhibitors (e.g., enzalutamide, apalutamide, darolutamide), cellular immunotherapy (e.g., sipuleucel-T), internal radiotherapy (Ra223), anti-neoplastic agents (e.g., docetaxel, cabazitaxel), immune checkpoint inhibitors (pembrolizumab), PARP inhibitors (e.g., olaparib, rucaparib), PSMA binders], a step of using as an input to a prediction model (e.g., a statistical model such as regression, e.g., a classification model by which a patient is assigned to a specific class based on comparison of the patient to one or more patient index values and one or more thresholds, e.g., a machine learning model that receives as input the value of one or more patient indices), wherein the efficacy score for a particular treatment option and / or therapeutic class indicates a prediction of whether the patient will benefit from the particular treatment and / or therapeutic class, the method according to any one of the above items. (Item 27) A step of generating (e.g., automatically) a report [e.g., an electronic document, e.g., within a graphical user interface (e.g., for verification / approval by a user)] comprising at least a portion of the value of the one or more patient index(es), the method according to any one of the above items. (Item 28) Step (b) uses one or more machine learning modules [e.g., one or more neural networks (e.g., one or more convolutional neural networks)] to A step of detecting a plurality of hot spots, wherein each of at least a portion of the plurality of 3D hot spot volumes corresponds to a particular detected hot spot and is created by partitioning the particular detected hot spot, and A step of partitioning at least a portion of the plurality of 3D hot spot volumes, and Classifying at least a part of the above-mentioned 3D hot spot volume (for example, determining the likelihood that each 3D hot spot volume represents a lower layer cancerous lesion) step and A method according to any one of the above items, comprising a step of performing one or more functions selected from the group consisting of. (Item 29) The method according to any one of the above items, wherein the 3D functional image includes a PET or SPECT image acquired following administration of an agent to the subject. (Item 30) The method according to item 29, wherein the agent comprises a PSMA binder. (Item 31) The agent is 18 F, the method according to item 29 or 30. (Item 32) The method according to item 30 or 31, wherein the agent comprises [18F]DCFPyL. (Item 33) The method according to item 30, wherein the agent comprises PSMA-11. (Item 34) The agent is 99m Tc, 68 Ga, 177 Lu, 225 Ac, 111 In, 123 I, 124 I, and 131 I, the method according to item 30, comprising one or more members selected from the group consisting of. (Item 35) A method for automated analysis of time-series medical images of a subject [for example, three-dimensional images, for example, nuclear medicine images (for example, bone scans (scintigraphy), PET, and / or SPECT), for example, anatomical images (for example, CT, X-ray, MRI), for example, combined nuclear medicine and anatomical images (for example, overlaid)], the method comprising: (a) Receiving and / or accessing the time-series medical images of the subject by a processor of a computing device; and (b) The processor identifies a plurality of hot spots within each of the medical images, and the processor determines one, two, or all three of the following: (i) a change in the number of identified lesions, (ii) a change in the overall volume of the identified lesions (e.g., a change in the sum value of the volumes of each identified lesion), and (iii) a change in the PSMA (e.g., lesion index)-weighted total volume (e.g., the sum value of the product of the lesion index and the lesion volume for all lesions within the region of interest) [e.g., the changes identified in step (b) are used to (1) identify the disease status (e.g., progression, regression, or no change), (2) make a treatment management decision (e.g., active surveillance, prostatectomy, anti-androgen therapy, prednisone, radiation, radiotherapy, radiation PSMA therapy, or chemotherapy), or (3) identify treatment effectiveness (e.g., when the subject has started or is continuing treatment with a drug or other therapy following an initial set of images in the time-series medical images)] [e.g., step (b) includes the step of using a machine learning module / model] and A method comprising. (Item 36) A method for analyzing a plurality of medical images of a subject (e.g., for assessing a disease state and / or degree of progression within the subject), the method comprising: (a) The processor of a computing device receives and / or accesses the plurality of medical images of the subject, and the processor obtains a plurality of 3D hot spot maps, each corresponding to a particular one of the (plurality of) medical images, and identifies one or more hot spots (e.g., representing potential underlying physical lesions within the subject) within the particular medical image; and (b) For each particular one (medical image) of the plurality of medical images, by the processor, using a machine learning module [e.g., a deep learning network (e.g., a convolutional neural network (CNN))], a corresponding 3D anatomical partitioning map is determined that identifies a set of organ regions within the particular medical image [e.g., representing one or more of the soft tissues and / or bone structures within the subject (e.g., cervical vertebrae, thoracic vertebrae, lumbar vertebrae, left and right lumbar bones, sacrum, and coccyx, left ribs and left scapula, right ribs and right scapula, left thigh, right thigh, skull, brain, and mandible)], thereby generating a plurality of 3D anatomical partitioning maps; (c) A step by the processor of determining the identification of one or more lesion correspondences using (i) the plurality of 3D hot spot maps and (ii) the plurality of 3D anatomical partitioning maps, where each (lesion correspondence) is determined to identify two or more corresponding hot spots in different medical images and represent the same underlying physical lesion within the subject (e.g., by the processor); (d) A step by the processor of determining the value of one or more measurement values {e.g., one or more hot spot quantification measurement values and / or changes therein [e.g., quantifying changes in properties such as the volume of individual hot spots and / or the underlying physical lesions they represent (e.g., over time / between multiple medical images), radiopharmaceutical uptake rate, shape, etc.], e.g., patient indices (e.g., measuring the overall disease burden and / or status and / or risk for the subject) and / or changes therein, e.g., values for classifying the patient (e.g., belonging to and / or having a particular disease state, progression, category, etc.), e.g., prognostic measurement values [e.g., indicating and / or quantifying the likelihood of one or more clinical outcomes (e.g., disease state, progression, likely survival rate, treatment efficacy, and the like) (e.g., overall survival rate)], e.g., predictive measurement values (e.g., indicating the predicted response to therapy and / or other clinical outcomes)} based on the plurality of 3D hot spot maps and the identification of one or more lesion correspondences; A method comprising (Item 37) The method according to item 36, wherein the plurality of medical images comprises one or more anatomical images (e.g., CT, X-ray, MRI, ultrasound, etc.). (Item 38) The method according to any one of items 36-37, wherein the plurality of medical images comprises one or more nuclear medicine images [e.g., bone scan (scintigraphy) (e.g., obtained following administration of a radiopharmaceutical such as 99mTc-MDP to the subject), PET (e.g., [18F]DCFPyL, [68Ga]PSMA-11, [18F]PSMA-1007, rhPSMA-7.3(18F), [18F]-JK-PSMA-7, etc., obtained following administration of a radiopharmaceutical to the subject), or SPECT (e.g., obtained following administration of a radiopharmaceutical such as a 99mTc-labeled PSMA binder to the subject)]. (Item 39) The method according to any one of items 36-38, wherein the plurality of medical images comprises one or more composite images (e.g., overlaid / co-registered with each other, e.g., obtained with respect to the subject at approximately the same time) (e.g., one or more PET / CT images), each of which comprises an anatomical and a nuclear medicine pair. (Item 40) The method according to any one of items 36-39, wherein the plurality of medical images is or comprises time-series medical images, and each medical image in the time series is associated with a different specific time and is obtained at that time. (Item 41) The time-series medical images include a first medical image obtained before administering a specific therapeutic agent [e.g., a PSMA binder (e.g., PSMA-617, e.g., PSMA I&T), e.g., a radiopharmaceutical, e.g., a radionuclide-labeled PSMA binder (e.g., 177Lu-PSMA-617, e.g., 177Lu-PSMA I&T)] (e.g., one or more cycles thereof) to the subject, and a second medical image obtained after administering the specific therapeutic agent (e.g., one or more cycles thereof) to the subject, and the method according to item 40. (Item 42) The method according to item 41, including the step of classifying the subject as a responder and / or non-responder to the specific therapeutic agent based on one or more measured values determined in step (d). (Item 43) Step (a) includes generating each hot spot map by partitioning at least a part of the corresponding medical image (e.g., a sub-image thereof such as a nuclear medicine image) (e.g., automatically), e.g., using a second hot spot partitioning machine learning module [e.g., the hot spot partitioning machine module includes a deep learning network (e.g., a convolutional neural network (CNN))], and the method according to any one of items 36-42. (Item 44) Each hot spot map includes one or more labels for identifying one or more assigned anatomical regions and / or lesion subtypes (e.g., miTNM classification labels) for at least a part of the hot spots identified therein, and the method according to any one of items 36-43. (Item 45) The plurality of hot spot maps include (i) a first hot spot map corresponding to the first medical image (e.g., identifying a first set of one or more hot spots therein), and (ii) a second hot spot map corresponding to the second medical image (e.g., identifying a second set of one or more hot spots therein). The plurality of 3D anatomical partitioning maps includes (i) a first 3D anatomical partitioning map that identifies a set of organ regions in the first medical image, and (ii) a second 3D anatomical partitioning map that identifies a set of organ regions in the second medical image. Step (c) includes using the first 3D anatomical partitioning map and the second 3D anatomical partitioning map to align (i) the first hot spot map and (ii) the second hot spot map (e.g., using a set of organ regions and / or one or more subsets thereof as landmarks within the first and second 3D anatomical partitioning maps to determine one or more alignment fields (e.g., a full 3D alignment field, e.g., point-by-point alignment), and using the one or more determined alignment fields to co-align the first and second hot spot maps). The method according to any one of items 36 - 44. (Item 46) Step (c) determining, for a group of two or more hot spots, each being a member of a different hot spot map and identified in different medical images, a value of one or more lesion correspondence metrics (e.g., volume overlap, e.g., centroid distance, e.g., lesion type match); determining, based on the value of the one or more lesion correspondence metrics, the two or more hot spots of the group that represent the same underlying physical lesion, thereby including the two or more hot spots of the group within one of the one or more lesion correspondences; The method according to any one of items 36 - 45, including. (Item 47) Step (d) includes the step of determining one, two, or all three of (i), (ii), and (iii) below, namely, (i) the change in the number of identified lesions, (ii) the change in the overall volume of the identified lesions (e.g., the change in the total value of the volumes of each identified lesion), and (iii) the change in the PSMA (e.g., lesion index) weighted total volume (e.g., the total value of the product of the lesion index and the lesion volume for all lesions in the region of interest) [e.g., the above changes identified in step (b) are used to (1) identify the disease status [e.g., progression, regression, or no change], (2) make treatment management decisions [e.g., active surveillance, prostatectomy, anti-androgen therapy, prednisone, radiation, radiotherapy, radiation PSMA therapy, or chemotherapy], or (3) identify treatment effectiveness (e.g., when the subject has started or continued treatment with a drug or other therapy following an initial set of images in the time series medical images)]. The method according to any one of items 36 - 46. (Item 48) The step of determining the value of one or more prognostic measurement values indicating the disease state / progression and / or treatment (e.g., based on the value of the one or more measurement values in step (d)), such as the step of determining the overall survival rate (OS) (e.g., the predicted number of months) expected for the subject. The method according to any one of items 36 - 47. (Item 49) The value of the one or more measurement values (e.g., the change in tumor volume, SUV mean 、SUV max(e.g., time, e.g., expected survival rate, time to progression, time to radiological progression, etc., expressed in months), an expected value and / or a range (e.g., class) indicating a value highly likely to be a particular patient outcome as output, such as the PSMA score, the number of new lesions, the number of disappeared lesions, the total number of tracked lesions), and using them as inputs to a prognostic model (e.g., a statistical model such as regression, e.g., a classification model by which a patient is assigned to a particular class based on the comparison of one or more patient index values above and one or more thresholds, e.g., a machine learning model that receives as input the value of one or more patient indices above), the method according to any one of items 36 - 48, including the step of generating. (Item 50) The value of the above - mentioned one or more measured values (e.g., change in tumor volume, SUV mean SUV max (e.g., binary classification), and using them as inputs to a response model (e.g., a statistical model such as regression, e.g., a classification model by which a patient is assigned to a particular class based on the comparison of one or more patient index values above and one or more thresholds, e.g., a machine learning model that receives as input the value of one or more patient indices above) to generate a classification indicating the patient response to treatment, such as the PSMA score, the number of new lesions, the number of disappeared lesions, the total number of tracked lesions), the method according to any one of items 36 - 49, including the step of using. (Item 51) The value of the above - mentioned one or more measured values (e.g., change in tumor volume, SUV mean SUV max, PSMA score, number of new lesions, number of disappeared lesions, total number of tracked lesions), as output, one or more treatment options (e.g., abiraterone, enzalutamide, apalutamide, darolutamide, Sipuleucel-T, Ra223, docetaxel, cabazitaxel, pembrolizumab, olaparib, rucaparib, 177Lu-PSMA-617, etc.) and / or classes of therapeutic agents [e.g., androgen biosynthesis inhibitors (e.g., abiraterone), androgen receptor inhibitors (e.g., enzalutamide, apalutamide, darolutamide), cellular immunotherapy (e.g., Sipuleucel-T), internal radiation therapy (Ra223), antineoplastic agents (e.g., docetaxel, cabazitaxel), immune checkpoint inhibitors (pembrolizumab), PARP inhibitors (e.g., olaparib, rucaparib), PSMA binders], for each, a prediction model (e.g., a statistical model such as regression, e.g., a classification model by which a patient is assigned to a specific class based on comparison of one or more patient index values and one or more thresholds, e.g., a machine learning model that receives as input the value of one or more patient indices) that generates an efficacy score, including the step of using as input to the prediction model, and the efficacy score for a specific treatment option and / or therapeutic class indicates a prediction of whether the patient will benefit from the specific treatment and / or therapeutic class, the method according to any one of items 36-50. (Item 52) A method for analyzing a plurality of medical images of a subject, the method comprising: (a) obtaining (e.g., receiving, and / or accessing, and / or generating) a first 3D hot spot map for the subject by a processor of a computing device; (b) obtaining (e.g., receiving, and / or accessing, and / or generating) a first 3D anatomical compartmentalization map associated with the first 3D hot spot map by the processor; (c) A step of obtaining (e.g., receiving, and / or accessing, and / or generating) a second 3D hot spot map for the object by the processor (d) A step of obtaining (e.g., receiving, and / or accessing, and / or generating) a second 3D anatomical partitioning map associated with the second 3D hot spot map by the processor (e) A step of determining a registration field (e.g., a full 3D registration field, e.g., point - by - point registration) using and / or based on the first 3D anatomical partitioning map and the second 3D anatomical partitioning map by the processor (f) A step of registering the first 3D hot spot map and the second 3D hot spot map using the determined registration field by the processor, thereby generating a co - registered pair of 3D hot spot maps (g) A step of determining the identification of one or more lesion correspondences using the co - registered pair of 3D hot spot maps by the processor (h) A step of storing and / or providing the identification of one or more lesion correspondences for display and / or further processing by the processor A method comprising (Item 53) A method for analyzing a plurality of medical images of an object (e.g., evaluating a disease state and / or progression within the object), the method comprising: (a) A step of receiving and / or accessing the plurality of medical images of the object by a processor of a computing device (b) For each particular one (medical image) of the plurality of medical images, the processor uses a machine learning module [e.g., a deep learning network (e.g., a convolutional neural network (CNN))] to determine a corresponding 3D anatomical partitioning map that identifies a set of organ regions within the particular medical image [e.g., representing one or more of the soft tissues and / or bone structures within the subject (e.g., cervical vertebrae, thoracic vertebrae, lumbar vertebrae, left and right lumbar bones, sacrum, and coccyx, left ribs and left scapula, right ribs and right scapula, left thigh, right thigh, skull, brain, and mandible)], thereby generating a plurality of 3D anatomical partitioning maps. (c) The processor uses the plurality of 3D anatomical partitioning maps to determine one or more alignment fields (e.g., a full 3D alignment field, e.g., point - by - point alignment), applies the one or more alignment fields, and aligns the plurality of medical images, thereby generating a plurality of aligned medical images. (d) For each particular one of the plurality of aligned medical images, the processor determines a corresponding aligned 3D hot - spot map that identifies one or more hot - spots (e.g., representing potential underlying physical lesions within the subject) within the particular aligned medical image, thereby generating a plurality of aligned 3D hot - spot maps. (e) The processor uses the plurality of 3D aligned hot - spot maps to determine one or more lesion correspondences, where each (lesion correspondence) is determined (e.g., by the processor) to identify two or more corresponding hot - spots in different medical images and represent the same underlying physical lesion within the subject. (f) Based on the plurality of 3D hot spot maps and the identification of one or more lesion correspondences by the above processor, determining one or more measured values {for example, one or more hot spot quantification measurement values and / or changes therein [for example, quantifying changes in properties such as individual hot spots (e.g., over time / between multiple medical images) and / or the volume of underlying physical lesions they represent, radiopharmaceutical uptake rate, shape, etc.], for example, patient indices (e.g., measuring the overall disease burden and / or status and / or risk for the subject) and / or changes thereof, for example, values for classifying the patient (e.g., belonging to and / or having a specific disease state, progression, category, etc.), for example, prognostic measurement values [for example, indicating and / or quantifying one or more clinical outcomes (e.g., disease state, progression, likely survival rate, treatment efficacy, and the like) (e.g., overall survival rate)], for example, predictive measurement values (e.g., indicating the predicted response to therapy and / or other clinical outcomes)} A method comprising. (Item 54) A method for analyzing a plurality of medical images of a subject, the method comprising: (a) Obtaining (e.g., receiving, and / or accessing, and / or generating) a first 3D anatomical image (e.g., CT, X-ray, MRI, etc.) and a first 3D functional image [e.g., nuclear medicine image (e.g., PET, SPECT, etc.)] of the subject by a processor of a computing device; (b) Obtaining (e.g., receiving, and / or accessing, and / or generating) a second 3D anatomical image and a second 3D functional image of the subject by the above processor; (c) Obtaining (e.g., receiving, and / or accessing, and / or generating) a first 3D anatomical compartmentalization map based on (e.g., using) the first 3D anatomical image by the above processor; (d) obtaining (e.g., receiving, and / or accessing, and / or generating) a second 3D anatomical partitioning map by the processor based on (e.g., using) the second 3D anatomical image; (e) determining, by the processor, a registration field (e.g., a full 3D registration field, e.g., point - by - point registration) using and / or based on the first 3D anatomical partitioning map and the second 3D anatomical partitioning map; (f) aligning (registering) the second 3D functional image with the first 3D functional image by the processor using the registration field, thereby generating a registered version of the second 3D functional image; (g) obtaining, by the processor, a first 3D hot - spot map associated with the first functional image; (h) determining, by the processor, a second 3D hot - spot map using the registered version of the second 3D functional image, wherein the second 3D hot - spot map is thereby aligned with the first 3D hot - spot map; (i) determining, by the processor, the identification of one or more lesion correspondences using the first 3D hot - spot map and the second 3D hot - spot map aligned therewith; (j) storing and / or providing, by the processor, the identification of one or more lesion correspondences for display and / or further processing A method comprising the above steps. (Item 55) A method for evaluating the effectiveness of an intervention, the method comprising: (a) For each specific subject in a test population presenting a specific disease (e.g., prostate cancer (e.g., metastatic castration-resistant prostate cancer)) and / or at risk thereof (e.g., registered in a clinical trial, e.g., comprising a plurality of subjects), implementing the method according to any one of the above items with respect to a plurality of medical images of the specific patient, wherein the plurality of medical images of the specific patient comprises time-series medical images acquired over a period covering the intervention under test (e.g., before, during, and / or after it), and the one or more risk indices comprise one or more endpoints indicating the patient response to the intervention under test, thereby determining for each of the one or more endpoints across the test population a plurality of values of the one or more endpoints, a step; (b) Determining the effectiveness of the intervention under test based on the values of the one or more endpoints across the test population; A method comprising. (Item 56) A method for treating a subject having a specific disease (e.g., prostate cancer (e.g., metastatic castration-resistant prostate cancer)) and / or at risk thereof, the method comprising: Administering a first cycle of a therapeutic agent to the subject; Based on the subject being imaged (e.g., before, during, and / or after the first cycle of the therapeutic agent) and being identified as a responder to the therapeutic agent using the method according to any one of Items 1 - 52 (e.g., the subject is identified / classified as a responder based on the value of the one or more risk indices determined using the method according to any one of Items 1 - 52), administering a second cycle of the therapeutic agent to the subject; A method comprising. (Item 57) A method for treating a subject having a specific disease (e.g., prostate cancer (e.g., metastatic castration-resistant prostate cancer)) and / or at risk thereof, the method comprising: Administering a cycle of a first therapeutic agent to the subject; The step of administering a cycle of a second therapeutic agent to the subject (e.g., thereby transitioning the subject to a potentially more effective therapy), based on the subject having been imaged (e.g., before, and / or during, and / or after the cycle of the first therapeutic agent) and identified as a non-responder to the first therapeutic agent using the method according to any one of items 1-52 (e.g., the subject is identified / classified as a non-responder based on a value of the risk index determined using the method according to any one of items 1-52 that is one or more than one) and A method comprising. (Item 58) A method for treating a subject having and / or at risk of a particular disease (e.g., prostate cancer (e.g., metastatic castration-resistant prostate cancer)), the method comprising: The step of administering a cycle of a therapeutic agent to the subject; and The step of interrupting the administration of the therapeutic agent to the subject (e.g., thereby transitioning the subject to a potentially more effective therapy), based on the subject having been imaged (e.g., before, and / or during, and / or after the cycle of the first therapeutic agent) and identified as a non-responder to the therapeutic agent using the method according to any one of items 1-52 (e.g., the subject is identified / classified as a non-responder based on a value of the risk index determined using the method according to any one of items 1-52 that is one or more than one) and A method comprising. (Item 59) An automated or semi-automated whole-body assessment method for a subject suffering from metastatic prostate cancer [e.g., metastatic castration-resistant prostate cancer (mCRPC) or metastatic hormone-sensitive prostate cancer (mHSPC)] for assessing disease progression and / or treatment efficacy, the method comprising: (a) A step of receiving, by a processor of a computing device, the first prostate-specific membrane antigen (PSMA)-targeted positron emission tomography (PET) image of the subject (the first PSMA-PET image) and the first 3D anatomical image of the subject [e.g., a computed tomography (CT) image, e.g., a magnetic resonance image (MRI)], wherein the first 3D anatomical image of the subject is acquired simultaneously with, or immediately after, or immediately before (e.g., on the same date as) the first PSMA-PET image such that the first 3D anatomical image and the first PSMA PET image correspond to a first date, and the images depict a sufficiently large area of the subject's body to cover the region of the body in which the metastatic prostate cancer has spread (e.g., a full-body image or a whole-body image covering multiple organs) {e.g., the PSMA-PET image is acquired using PYLARIFY (registered trademark), F-18 piflufolastat PSMA (i.e., 2-(3-{1-carboxy-5-[(6-[18F]fluoropyridine-3-carbonyl)amino]-pentyl}ureido)-pentanedioic acid, alias, [18F]F-DCFPyL), or Ga-68 PSMA-11, or other radiolabeled prostate-specific membrane antigen inhibitor contrast agents,}, step, (b) A step of receiving, by the processor, both the second PSMA-PET image of the subject and the second 3D anatomical image of the subject, both of which are acquired on a second date following the first date, (c) Using the landmarks automatically identified within the above-described first and second 3D anatomical images (e.g., an identified region representing one or more of cervical vertebrae, thoracic vertebrae, lumbar vertebrae, left and right pelvic bones, sacrum, and coccyx, left rib and left scapula, right rib and right scapula, left thigh, right thigh, sk...

Claims

1. A method for automatically processing a 3D image of a subject and determining the value of one or more patient indices to measure disease burden and / or risk relating to the subject, wherein the method is: (a) The processor of the computing device receives a 3D functional image of the object acquired using a functional imaging modality, (b) The processor partitions a plurality of 3D hotspot volumes within the 3D functional image, each 3D hotspot volume corresponding to a local area of ​​increased intensity relative to its surroundings, representing a potential cancerous lesion within the subject, thereby obtaining a set of 3D hotspot volumes. (c) The processor calculates the value of the specific individual hotspot quantification measurement for each individual 3D hotspot volume in the set for each of the one or more individual hotspot quantification measurement values, (d) The processor determines the values ​​of the one or more patient indices, wherein each of the at least portion of the patient indices is associated with one or more specific individual hotspot quantification measurements and is a function of at least portion of the values ​​of the one or more specific individual hotspot quantification measurements calculated with respect to the set of 3D hotspot volumes. Methods that include...

2. The method according to claim 1, wherein at least one specific patient index of the one or more patient index values ​​is associated with a single specific individual hotspot quantification measurement value and is calculated as a function of all the values ​​of the specific individual hotspot quantification measurement values ​​calculated with respect to the set of 3D hotspot volumes.

3. The method according to claim 2, wherein the single specific individual hotspot quantification measurement is an individual hotspot intensity measurement that quantifies the intensity within the 3D hotspot volume.

4. The method according to claim 3, wherein the individual hotspot intensity measurements are average hotspot intensities.

5. The method according to claim 3, wherein the specific patient index is calculated as the sum of all the values ​​of the individual hotspot intensity measurements calculated with respect to the set of 3D hotspot volumes.

6. The method according to claim 6, wherein the single specific individual hotspot quantification measurement is the lesion volume.

7. The method according to claim 6, wherein the specific patient index (value) is calculated as the sum of all the lesion volume values ​​calculated with respect to the set of 3D hotspot volumes.

8. The method according to claim 1, wherein one particular of the one or more overall patient indices is associated with two or more specific individual hotspot quantification measurements and is calculated as a function of all the values ​​of the two or more specific individual hotspot quantification measurements calculated with respect to the set of 3D hotspot volumes.

9. The method according to claim 8, wherein the two or more specific individual hotspot quantification measurements comprise (i) individual hotspot intensity measurements and (ii) lesion volume.

10. The method according to claim 9, wherein each of the hotspot intensity measurements is an individual lesion index that maps the hotspot intensity values ​​to values ​​on a standardized scale.

11. The aforementioned specific patient index (value) is, For each individual 3D hotspot volume of all the aforementioned 3D hotspot volumes, the value of the lesion volume is weighted by the value of the individual hotspot intensity measurement, thereby calculating multiple intensity-weighted lesion volumes. The value of the aforementioned specific patient index is calculated as the sum of all the intensity-weighted lesion volumes. The method according to claim 9, wherein the total value of the intensity-weighted lesion volume is calculated by the method thereof.

12. The method according to claim 1, wherein the one or more individual hotspot quantification measurements include one or more individual hotspot intensity measurements that quantify the intensity within a 3D hotspot volume.

13. The quantitative measurement values ​​of one or more individual hotspots are, Average hotspot intensity and Maximum hotspot intensity and Central hotspot intensity and The method according to claim 12, comprising one or more members selected from the group consisting of the following.

14. The method according to claim 12, wherein the one or more individual hotspot intensity measurements include the peak intensity of the 3D hotspot volume.

15. The method according to claim 12, wherein the one or more individual hotspot intensity measurements include individual lesion indices that map the hotspot intensity values ​​to values ​​on a standardized scale.

16. The processor identifies one or more 3D reference volumes within the 3D functional image, each corresponding to a specific reference tissue region. Each of the aforementioned processors is associated with a specific 3D reference volume of the one or more 3D reference volumes, and determines one or more reference intensity values ​​corresponding to the intensity measurements within the specific 3D reference volume. In step (c), for each 3D hotspot volume in the set, The aforementioned processor determines the corresponding value for each specific hotspot intensity measurement, The processor determines the corresponding value of each lesion index based on the corresponding values ​​of the specific individual hotspot intensity measurement and one or more reference intensity values. To do The method according to claim 15, including the method described in claim 15.

17. Mapping each of the one or more reference intensity values ​​to the corresponding reference index value on the scale, For each 3D hotspot volume, the corresponding value of the individual lesion index is determined using the reference intensity value and the corresponding reference index value, and the corresponding individual lesion index value is interpolated onto the scale based on the corresponding value of the specific individual hotspot intensity measurement. The method according to claim 16, including the method described in claim 16.

18. The method according to claim 16, wherein the reference tissue region comprises one or more members selected from the group consisting of the liver, the aorta, and the parotid gland.

19. The first reference intensity value is (i) a blood reference intensity value associated with a reference volume corresponding to the aortic portion, and (ii) is mapped to the first reference index value. The second reference intensity value is (i) a liver reference intensity value associated with the reference volume corresponding to the liver, and (ii) mapped to the second reference index value. The second reference intensity value exceeds the first reference intensity value, and the second reference index value exceeds the first reference index value. The method according to claim 16.

20. The method according to claim 16, wherein the reference intensity value includes a maximum reference intensity value mapped to a maximum reference index value, and 3D hotspot volumes in which the corresponding value of a particular individual hotspot intensity measurement exceeds the maximum reference intensity value are assigned individual lesion index values ​​equal to the maximum reference index value.

21. Within the collection of 3D hotspot volumes, one or more subsets are identified, each associated with a specific tissue region and / or lesion classification. For each specific subset of the one or more subsets, the corresponding values ​​of one or more specific patient indices are calculated using the values ​​of the individual hotspot quantification measurements calculated with respect to the 3D hotspot volume within the specific subset. The method according to claim 1, including the method described in claim 1.

22. The method according to claim 21, wherein each of the one or more subsets is associated with a specific one of one or more tissue regions, and the method comprises, for each specific tissue region, identifying a subset of the 3D hotspot volume located within the volume of interest corresponding to the specific tissue region.

23. The method according to claim 22, wherein the one or more tissue regions comprises one or more members selected from the group consisting of a skeletal region, a lymphatic region, and a prostatic region, each comprising one or more bones of the subject.

24. The method according to claim 21, wherein each of the one or more subsets is associated with a specific one of one or more lesion subtypes, and the method comprises determining a corresponding lesion subtype for each 3D hotspot volume and assigning the 3D hotspot volume to the one or more subsets according to those corresponding lesion subtypes.

25. The method according to claim 1, comprising using at least a portion of the values ​​of the one or more patient indices as input to a prognostic model that generates expected values ​​and / or ranges indicating a high probability of a particular patient outcome, as output.

26. The method according to claim 1, comprising using at least a portion of the values ​​of the one or more patient indices as input to a predictive model that generates efficacy scores for each of the classes of one or more treatment options and / or therapeutic drugs, wherein the efficacy score for a particular treatment option and / or therapeutic class indicates a prediction of whether the patient will benefit from the particular treatment and / or therapeutic class.

27. The method according to claim 1, comprising generating a report that includes at least a portion of the values ​​of one or more patient indices.

28. Step (b) involves using one or more machine learning modules, The process involves detecting multiple hotspots, wherein each of the multiple 3D hotspot volumes corresponds to a specific detected hotspot and is created by partitioning the specific detected hotspot. To partition at least a portion of the plurality of 3D hotspot volumes, To classify at least a portion of the aforementioned 3D hotspot volume and The method according to claim 1, comprising performing one or more functions selected from the group consisting of the following.

29. The method according to claim 1, wherein the 3D functional image includes a PET or SPECT image acquired following the administration of the active substance to the subject.

30. The method according to claim 29, wherein the active substance comprises a PSMA binder.

31. The aforementioned active substance is, １８ The method according to claim 29, including F.

32. The method according to claim 30, wherein the active substance comprises [18F]DCFPyL.

33. The method according to claim 30, wherein the active substance comprises PSMA-11.

34. The acting substance is ９９ｍ Tc, ６８ Ga, １７７ Lu, ２２５ Ac, １１１ In, １２３ I, １２４ I, and １３１ The method according to claim 30, comprising one or more members selected from the group consisting of I.

35. A method for automated analysis of a time-series medical image of a subject, wherein the method is: (a) The processor of the computing device receives and / or accesses the time-series medical images of the subject, (b) The processor identifies multiple hotspots within each of the medical images, and the processor determines one, two, or all three of (i), (ii), and (iii), such as (i) a change in the number of identified lesions, (ii) a change in the total volume of the identified lesions, and (iii) a change in the PSMA-weighted total volume. Methods that include...

36. The method according to claim 35, wherein the time-series medical images are composite time-series images of the subject, each composite medical image is obtained at a different time and includes an anatomical image and a corresponding nuclear medicine image.

37. A method for analyzing multiple medical images of a subject, wherein the method is (a) The processor of the computing device receives and / or accesses the multiple medical images of the subject, and the processor obtains multiple 3D hotspot maps, each corresponding to a specific medical image (of the multiple), and identifies one or more hotspots within the specific medical image. (b) For each of the multiple medical images, the processor uses a machine learning module to determine a corresponding 3D anatomical compartment map that identifies a set of organ regions within the specific medical image, thereby generating multiple 3D anatomical compartment maps. (c) The processor determines the identification of one or more lesion correspondences using (i) the plurality of 3D hotspot maps and (ii) the plurality of 3D anatomical compartmentalization maps, wherein each (lesion correspondence) is determined to identify two or more corresponding hotspots in different medical images and to represent the same underlying physical lesion within the subject. (d) The processor determines the value of one or more measured values ​​based on the plurality of 3D hotspot maps and the identification of one or more lesions. Methods that include...

38. The method according to claim 37, wherein the plurality of medical images include one or more anatomical images.

39. The method according to claim 37, wherein the plurality of medical images include one or more nuclear medicine images.

40. The method according to claim 37, wherein the plurality of medical images include one or more composite images, each having an anatomical and nuclear medicine pair.

41. The method according to claim 37, wherein the plurality of medical images are time-series medical images or include time-series medical images, each of the time-series medical images is associated with a different specific time and obtained at the different specific time.

42. The method according to claim 41, wherein the time-series medical images are composite time-series images of the subject, each composite medical image is obtained at a different time and includes an anatomical image and a corresponding nuclear medicine image.

43. The method according to claim 42, wherein each 3D hotspot map corresponds to a specific composite medical image and identifies one or more hotspots within the nuclear medicine image portion of the specific composite medical image.

44. The method according to claim 43, wherein step (a) comprises generating each 3D hotspot map by partitioning the nuclear medicine image portion of the corresponding composite medical image.

45. The method according to claim 42, wherein for each specific composite medical image in the time series, the corresponding 3D anatomical compartment map identifies the set of organ regions within the anatomical image portion of the specific composite medical image.

46. The plurality of hotspot maps include (i) a first 3D hotspot map corresponding to a first composite medical image, and (ii) a second 3D hotspot map corresponding to a second composite medical image, The plurality of 3D anatomical compartment maps include (i) a first 3D anatomical compartment map that identifies the set of organ regions within the anatomical image portion of the first composite medical image, and (ii) a second 3D anatomical compartment map that identifies the set of organ regions within the anatomical image portion of the second composite medical image. Step (c) includes using the first 3D anatomical compartment map and the second 3D anatomical compartment map to (i) align the first 3D hotspot map with (ii) the second 3D hotspot map, The method according to claim 42.

47. The method according to claim 41, wherein the time-series medical images include a first medical image obtained before administering a specific therapeutic agent to the subject and a second medical image obtained after administering the specific therapeutic agent to the subject.

48. The method according to claim 47, comprising classifying the subject as responders and / or non-responders to the specific therapeutic agent based on the values ​​of one or more measurements determined in step (d).

49. The method according to claim 37, step (a) comprising generating each 3D hotspot map by partitioning at least a portion of the corresponding medical image.

50. The method according to claim 37, wherein each hotspot map comprises one or more labels that identify one or more assigned anatomical regions and / or lesion subtypes for at least a portion of the one or more hotspots identified therein.

51. The plurality of hotspot maps include (i) a first hotspot map corresponding to a first medical image and (ii) a second hotspot map corresponding to a second medical image. The plurality of 3D anatomical compartment maps include (i) a first 3D anatomical compartment map that identifies the set of organ regions in the first medical image, and (ii) a second 3D anatomical compartment map that identifies the set of organ regions in the second medical image. Step (c) includes using the first 3D anatomical compartment map and the second 3D anatomical compartment map to (i) align the first hotspot map with (ii) the second hotspot map, The method according to claim 37.

52. Step (c) is, Each of these is a member of a different hotspot map, and for two or more groups of hotspots identified within different medical images, the value of one or more lesion-corresponding metrics is determined. Based on the values ​​of the one or more lesion-corresponding measurement values, two or more hotspots in the group are determined to represent the same specific underlying physical lesion, thereby including the two or more hotspots in the group within one or more of the lesion-corresponding values. The method according to claim 37, including the method described in claim 37.

53. The method of claim 37, wherein step (d) comprises determining one, two, or all three of (i), (ii), and (iii), such as (i) a change in the number of identified lesions, (ii) a change in the total volume of identified lesions, and (iii) a change in the PSMA-weighted total volume.

54. The method according to claim 37, comprising determining the value of one or more prognostic measures indicating disease status / progression and / or treatment.

55. The method according to claim 37, comprising using the values ​​of one or more of the aforementioned measurements as inputs to a prognostic model that generates expected values ​​and / or ranges indicating a high probability of a particular patient outcome.

56. The method according to claim 37, comprising using the values ​​of one or more of the measured values ​​as an output, and as input to a response model that generates a classification indicating the patient's response to treatment.

57. The method according to claim 37, comprising using the values ​​of one or more measurements as an output, as input to a predictive model that generates efficacy scores for each of one or more treatment options and / or therapy drug classes, wherein the efficacy score for a particular treatment option and / or therapy class indicates a prediction of whether the subject will benefit from the particular treatment option and / or therapy class.

58. A method for analyzing multiple medical images of a subject, wherein the method is (a) The processor of the computing device acquires a first 3D hotspot map relating to the object, (b) The processor obtains a first 3D anatomical compartment map associated with the first 3D hotspot map, (c) The processor obtains a second 3D hotspot map relating to the object, (d) The processor obtains a second 3D anatomical compartment map associated with the second 3D hotspot map, (e) The processor determines the alignment field using the first 3D anatomical compartment map and the second 3D anatomical compartment map, (f) The processor aligns the first 3D hotspot map and the second 3D hotspot map using the determined alignment field, thereby generating a pair of co-aligned 3D hotspot maps. (g) The processor determines the identification of one or more lesions using the co-aligned pair of 3D hotspot maps, (h) The processor stores and / or provides the identification of one or more lesions for display and / or further processing. Methods that include...

59. A method for analyzing multiple medical images of a subject, wherein the method is (a) The processor of the computing device receives and / or accesses the plurality of medical images of the subject, (b) For each of the multiple medical images, the processor uses a machine learning module to determine a corresponding 3D anatomical compartment map that identifies a set of organ regions within the specific medical image, thereby generating multiple 3D anatomical compartment maps. (c) The processor determines one or more alignment fields using the plurality of 3D anatomical compartment maps, applies the one or more alignment fields to align the plurality of medical images, and thereby generates a plurality of aligned medical images. (d) The processor determines a corresponding aligned 3D hotspot map for each specific one of the plurality of aligned medical images, which identifies one or more hotspots within the specific aligned medical image, thereby generating a plurality of aligned 3D hotspot maps. (e) The processor determines the identification of one or more lesion correspondences using the plurality of 3D-aligned hotspot maps, wherein each (lesion correspondence) identifies two or more corresponding hotspots in different medical images and represents the same underlying physical lesion within the subject. (f) The processor determines the value of one or more measured values ​​based on the plurality of 3D hotspot maps and the identification of one or more lesions. Methods that include...

60. A method for analyzing multiple medical images of a subject, wherein the method is (a) The processor of the computing device acquires a first 3D anatomical image and a first 3D functional image of the subject, (b) The processor acquires a second 3D anatomical image and a second 3D functional image of the object, (c) The processor obtains a first 3D anatomical compartment map based on the first 3D anatomical image, (d) The processor obtains a second 3D anatomical compartment map based on the second 3D anatomical image, (e) The processor determines the alignment field using the first 3D anatomical compartment map and the second 3D anatomical compartment map, (f) The processor uses the alignment field to align (match) the second 3D functional image with the first 3D functional image, thereby generating an aligned version of the second 3D functional image. (g) The processor obtains a first 3D hotspot map associated with the first functional image, (h) The processor determines a second 3D hotspot map using the aligned version of the second 3D functional image, the second 3D hotspot map being aligned with the first 3D hotspot map. (i) The processor determines the identification of one or more lesions using the first 3D hotspot map and the second 3D hotspot map aligned thereto, (j) The processor stores and / or provides the identification of one or more lesions for display and / or further processing. Methods that include...

61. A method for evaluating the effectiveness of an intervention, wherein the method is (a) For each specific subject of the study population exhibiting and / or at risk of a specific disease, the processor performs the method of any one of claims 1 to 60 with respect to a plurality of medical images of the specific patient, wherein the plurality of medical images of the specific patient include time-series medical images acquired over a period extending to the intervention under study, and the one or more risk indices comprises one or more endpoints indicating the patient's response to the intervention under study, thereby determining a plurality of values ​​for each of the one or more endpoints across the study population; (b) Determining the effectiveness of the intervention under the trial based on the values ​​of one or more endpoints across the trial population. Methods that include...

62. An automated or semi-automated systemic assessment method for a subject with metastatic prostate cancer to assess disease progression and / or treatment effectiveness, wherein the method is: (a) The processor of the computing device receives a first prostate-specific membrane antigen (PSMA) targeted positron emission tomography (PET) image of the subject (the first PSMA-PET image) and a first 3D anatomical image of the subject, wherein the first 3D anatomical image of the subject is acquired simultaneously with, immediately after, or immediately before, the first PSMA-PET image, such that the first 3D anatomical image and the first PSMA-PET image correspond to a first date, and the image depicts an area of ​​the subject's body that is sufficiently large to cover the area of ​​the body where the metastatic prostate cancer is spreading. (b) The processor receives a second PSMA-PET image of the subject and a second 3D anatomical image of the subject, both of which are acquired on a second date following the first date. (c) The processor automatically determines a registration field using landmarks automatically identified in the first and second 3D anatomical images, and the processor aligns the first and second PSMA-PET images using the determined registration field. (d) Using the first and second PSMA-PET images thus matched, the processor automatically detects the changes in the disease from the first date to the second date. Methods that include...

63. The method according to claim 62, wherein the method comprises one or more members selected from the group consisting of lesion location assignment, tumor staging, lymph node staging, distant metastasis staging, assessment of intraprostatic lesions, and determination of PSMA expression score.

64. The method of claim 62, wherein the subjects have been administered therapy to them one or more times between the first date and the second date (after the first image is taken and before the second image is taken) for the treatment of the metastatic prostate cancer, so that the method is used to assess the therapeutic effect.

65. The processor obtains one or more further PSMA-PET images and corresponding 3D anatomical images of the subject following the second date, The processor uses corresponding 3D anatomical images to match one or more additional PSMA-PET images with the first and second PSMA-PET images, thereby obtaining the matched additional PSMA-PET images. The processor automatically assesses the disease progression and / or treatment effectiveness using the further matched PSMA PET images. The method according to claim 62, further comprising:

66. The method according to claim 62, further comprising the processor determining and rendering, at least in part, a predicted PSMA-PET image depicting the predicted progression and / or remission of the disease to a future date, based on the detected changes in the disease from the first date to the second date.

67. A method for quantifying and reporting disease burden in patients with cancer and / or at risk of cancer, wherein the method is: (a) Acquiring medical images of the patient using the processor of a computing device, (b) The processor detects one or more hotspots in the medical image, each hotspot corresponding to a specific 3D volume in the medical image and representing a potential underlying physical lesion within the subject, (c) For each specific lesion class of multiple lesion classes that represent a specific tissue region and / or lesion subtype, The processor identifies a corresponding subset of the one or more hotspots as belonging to the specific lesion class, The processor determines, based on a corresponding subset of the hotspots, one or more patient index values ​​that quantify the disease burden within and / or associated with the particular lesion class. To do, (d) The processor causes the display of a graphical representation of the patient index value calculated for each of the plurality of lesion classes, thereby providing the user with a graphical report summarizing the tumor burden within a specific tissue region and / or associated with a specific lesion subtype. Methods that include...

68. The aforementioned multiple lesion classes are as follows: (i) A local tumor class which identifies potential lesions and / or parts thereof located within one or more local tumor-associated tissue regions that are associated with a localized tumor site within the patient and / or adjacent to a localized tumor site within the patient, and for the local tumor class, a corresponding subset of the hotspots represents the potential lesions and / or parts thereof. (ii) A local lymph node class in which the local lymph node class identifies potential lesions located within local lymph nodes adjacent to and / or near the original tumor site, and for the local lymph node class, the corresponding subset of the hotspots represents the potential lesions, (iii) One or more metastatic tumor classes, wherein the one or more metastatic tumor classes identify potential metastases and / or their subtypes, and the corresponding subset of the hotspots for the one or more metastatic tumor classes represents the potential metastases and / or their subtypes. The method according to claim 67, comprising one or more of the above.

69. The aforementioned one or more metastatic tumor classes are as follows: A distant lymph node metastasis class, wherein the distant lymph node metastasis class identifies potential lesions that have metastasized to distant lymph nodes, and the corresponding subset of the hotspots for the distant lymph node metastasis class represents potential lesions, A distant bone metastasis class, wherein the distant bone metastasis class identifies potential lesions located within one or more bones of the patient, and a corresponding subset of the hotspots represents the potential lesions for the distant bone metastasis class, A visceral metastasis class identifies potential lesions located in one or more organs or in other non-lymphatic soft tissue areas outside the local tumor-associated tissue area, and a corresponding subset of the hotspots represents potential lesions for the visceral metastasis class. The method according to claim 67, comprising one or more of the above.

70. Step (c) is the following patient index for each specific lesion class, namely, A lesion count that quantifies the number of lesions represented by a subset of hotspots corresponding to the aforementioned specific lesion class, The maximum acquisition value quantifies the maximum acquisition rate within the corresponding set of the aforementioned hotspots, The average acquisition value quantifies the overall average acquisition rate within the corresponding subset of the hotspot, Total lesion volume, which quantifies the total volume of lesions belonging to the aforementioned specific lesion class, The intensity-weighted tumor volume (ILTV) score is calculated as the weighted sum of all individual lesion volumes weighted by those intensity measurements, and The method according to claim 67, comprising determining one or more of the following values.

71. The method according to claim 67, comprising determining an alphanumeric code for classifying the overall load within a particular lesion for each lesion class.

72. The method according to claim 71, wherein step (e) includes causing the occurrence and / or display of the alphanumeric code representation for each specific lesion class.

73. The method of claim 67, further comprising determining an overall disease stage for the patient indicating an overall disease status and / or burden for the patient based on the plurality of lesion classes and corresponding subsets of hotspots, and causing the processor to render a graphical representation of the overall disease stage for inclusion in the report.

74. The processor determines one or more reference intensity values ​​representing the physiological uptake rate of the radiopharmaceutical, each calculated based on the intensity of image voxels in the corresponding reference volume identified in the medical image within a specific reference tissue region within the patient. In step (d), the processor causes the rendering of one or more representations of the reference intensity values ​​for inclusion in the report. The method according to claim 67, further comprising:

75. A method for characterizing and reporting individual lesions detected based on imaging assessments of patients with cancer and / or at risk of cancer, wherein the method is: (a) Acquiring medical images of the patient using the processor of a computing device, (b) The processor detects one or more sets of hotspots in the medical image, wherein each hotspot in the set corresponds to a specific 3D volume in the medical image and represents a potential underlying physical lesion within the subject. (c) The processor assigns one or more lesion class labels to each of the one or more hotspots in the set, wherein each lesion class label represents a specific tissue region and / or lesion subtype, and identifies the hotspot as representing a potential lesion located within the specific tissue region and / or belonging to the lesion subtype. (d) The processor calculates the value of the specific individual hotspot quantification measurement for each individual hotspot in the set, for each specific individual hotspot quantification measurement among one or more individual hotspot quantification measurements. (e) The processor causes for each specific hotspot of at least a portion of the hotspots in the set to display a graphical representation that includes the identification of the specific hotspot, the one or more lesion class labels assigned to the specific hotspot, and the one or more individual hotspot quantification measurements calculated with respect to the specific hotspot. Methods that include...

76. The lesion class label is as follows: (i) A local tumor class which identifies potential lesions and / or parts thereof located within one or more local tumor-associated tissue regions that are associated with a localized tumor site within the patient and / or adjacent to a localized tumor site within the patient, and for the local tumor class, a corresponding subset of the hotspots represents the potential lesions and / or parts thereof. (ii) A local lymph node class in which the local lymph node class identifies potential lesions located within local lymph nodes adjacent to and / or near the original tumor site, and for the local lymph node class, the corresponding subset of the hotspots represents the potential lesions, (iii) One or more metastatic tumor classes, wherein the one or more metastatic tumor classes identify potential metastases and / or their subtypes, and the corresponding subset of the hotspots for the one or more metastatic tumor classes represents the potential metastases and / or their subtypes. The method according to claim 75, comprising a sign representing one or more of the following.

77. The aforementioned one or more metastatic tumor classes are as follows: A distant lymph node metastasis class, wherein the distant lymph node metastasis class identifies potential lesions that have metastasized to distant lymph nodes, and the corresponding subset of the hotspots for the distant lymph node metastasis class represents the potential lesions, A distant bone metastasis class, wherein the distant bone metastasis class identifies potential lesions located within one or more bones of the patient, and a corresponding subset of the hotspots represents the potential lesions for the distant bone metastasis class, A visceral (also referred to as distant soft tissue) metastasis class in which the visceral (also referred to as distant soft tissue) metastasis class identifies potential lesions located in one or more organs or in other non-lymphatic soft tissue areas outside the local tumor-associated tissue area, and for the visceral (also referred to as distant soft tissue) metastasis class, the corresponding subset of the hotspots represents the potential lesions. The method according to claim 76, comprising one or more of the above.

78. The method according to claim 75, wherein the lesion class label includes one or more tissue labels that identify a specific organ or bone where the lesion (represented by the hotspot) is determined to be located.

79. The method according to claim 75, wherein the one or more individual hotspot quantification measurements include one or more of the following: maximum intensity, peak intensity, mean intensity, lesion volume, and lesion index.

80. A method for quantifying and reporting disease progression and / or risk over time in patients who have cancer and / or are at risk of cancer, wherein the method is: (a) The processor of a computing device acquires multiple medical images of the patient, each medical image representing a scan of the patient taken at a specific time; (b) For each specific medical image among the plurality of medical images, the processor detects a corresponding set of one or more hotspots within the specific medical image, wherein each hotspot corresponds to a specific 3D volume within the medical image and represents a potential underlying physical lesion within the subject. (c) For each specific patient index of one or more patient indices that measure the overall disease burden within a patient at a specific time, the processor determines the value of the specific patient index for each specific medical image of the plurality of medical images based on the corresponding set of hotspots detected with respect to the specific medical image, thereby determining a set of values ​​that track changes in disease burden over time as measured by the specific patient index value for each specific patient index of one or more patient indices. (d) The processor causes the display of a graphical representation of the set of values ​​relating to at least one or more of the patient index values, thereby conveying a time-series measurement of the disease progression for the patient. Methods that include...

81. The one or more patient indices mentioned above are: A lesion count, which quantifies the number of lesions represented by a collection of hotspots detected within a specific medical image, and A maximum acquisition value that quantifies the maximum acquisition rate of a hotspot related to a specific medical image within the corresponding set, The average acquisition value quantifies the overall average acquisition rate within the corresponding set of hotspots, Total volume lesion volume, which quantifies the total volume of lesions detected within the subject at a specific point in time, The intensity-weighted tumor volume (ILTV) score is calculated as the weighted sum of all individual lesion volumes, with each individual lesion volume weighted according to its intensity measurement. The method according to claim 80, including the method described in claim 80.

82. The method according to claim 80, further comprising: determining an overall disease stage indicating the overall disease status and / or burden for the patient at a particular point in time, based on the corresponding set of hotspots for each specific medical image of the plurality of medical images; and causing the processor to render a graphical representation of the overall disease stage at each point in time.

83. The processor determines, for each of the plurality of medical images, a set of one or more reference intensity values ​​that represent the physiological uptake rate of the radiopharmaceutical in a specific reference tissue region, each calculated based on the intensity of the image voxel in the corresponding reference volume identified within the medical image in the patient, The processor causes rendering of one or more representations of the reference intensity values. The method according to claim 80, further comprising:

84. A method for automatically processing a 3D image of a subject and determining the value of one or more patient indices to measure disease burden and / or risk relating to the subject, wherein the method is: (a) The processor of the computing device receives a 3D functional image of the object acquired using a functional imaging modality, (b) The processor partitions a plurality of 3D hotspot volumes within the 3D functional image, each 3D hotspot volume corresponding to a local area of ​​increased intensity relative to its surroundings, representing a potential cancerous lesion within the subject, thereby obtaining a set of 3D hotspot volumes. (c) The processor calculates, for each specific hotspot quantification metric among one or more individual hotspot quantification metrics, the value of the specific individual hotspot quantification metric for each individual 3D hotspot volume in the set, wherein, with respect to a specific individual 3D hotspot volume, each hotspot quantification metric quantifies the properties of the specific 3D hotspot volume and is a specific function of the intensity and / or number of individual voxels within the specific 3D hotspot volume. (d) The processor determines the values ​​of the one or more patient indices, each of which at least a portion of the patient indices is associated with one or more specific individual hotspot quantification measurements and is calculated using a particular function with the intensity and / or number of voxels in a combined hotspot volume comprising at least a portion of the set of 3D hotspot volumes. Methods that include...

85. A method for automatically determining the prognosis of a subject suffering from prostate cancer from one or more medical images of the subject, wherein the method is: (a) The processor of the computing device receives one or more images of the subject and / or accesses one or more images of the subject, (b) The processor automatically determines a quantitative assessment of one or more prostate cancer lesions from one or more images, (c)(b) The prognosis for the subject is to be automatically determined from the quantitative assessment in (c)(b), wherein the prognosis includes one or more of the following for the subject: (I) expected survival rate, (II) expected time to disease progression, (III) expected time to radiological progression, (IV) risk of simultaneous metastasis, and (V) risk of future metastatic metastasis. Methods that include...

86. The method according to claim 85, wherein the quantitative assessment of the one or more prostate cancer lesions determined in step (b) comprises one or more of the following: (A) total tumor volume, (B) change in tumor volume, (C) total SUV, and (D) PSMA score, and the prognosis for the subject determined in step (c) comprises one or more of the following: (E) expected survival rate, (F) time to progression, and (G) time to radiological progression.

87. The method of claim 85, wherein the quantitative assessment of the one or more prostate cancer lesions determined in step (b) includes one or more characteristics of PSMA expression in the prostate, and the prognosis for the subject determined in step (c) includes the risk of synchronous metastasis and / or the risk of future metachronous metastasis.

88. A method for automatically determining the response to treatment of a subject suffering from prostate cancer from multiple medical images of the subject, wherein the method is: (a) The processor of the computing device receives and / or accesses a plurality of images of the subject, wherein at least a first image of the plurality of images is acquired prior to the administration of the treatment, and at least a second image of the plurality of images is acquired following the administration of the treatment, (b) The processor automatically determines a quantitative assessment of one or more prostate cancer lesions from the image, (c) The processor automatically determines, from the quantitative assessment in (b), whether the subject is responding to the treatment and / or the extent to which the subject is responding to the treatment. Methods that include...

89. A method for automatically identifying whether a subject with prostate cancer is likely to benefit from a specific treatment for prostate cancer, using multiple medical images of the subject, wherein the method is: (a) The processor of the computing device receives and / or accesses the multiple images of the subject, (b) The processor automatically determines a quantitative assessment of one or more prostate cancer lesions from the image, (c) The processor automatically determines, based on the quantitative assessment in (b), whether the subject is likely to benefit from the specific treatment for prostate cancer. Methods that include...

90. A system for automatically processing a 3D image of a subject and determining the value of one or more patient indices to measure disease burden and / or risk related to the subject, wherein the system The processor of a computing device, A memory wherein the memory has instructions stored thereon, and when an instruction is executed by the processor, the processor, (a) Receiving a 3D functional image of the object acquired using a functional imaging modality, (b) Partitioning a plurality of 3D hotspot volumes within the 3D functional image, wherein each 3D hotspot volume corresponds to a local area of ​​increased intensity relative to its surroundings, represents a potential cancerous lesion within the subject, and thereby obtains a set of 3D hotspot volumes. (c) For each specific hotspot quantification measurement value among one or more individual hotspot quantification measurements, the value of the specific individual hotspot quantification measurement value for each individual 3D hotspot volume in the set is calculated, (d) Determining the values ​​of one or more patient indices, wherein each of at least a portion of the patient indices is associated with one or more specific individual hotspot quantification measurements and is a function of at least a portion of the values ​​of one or more specific individual hotspot quantification measurements calculated with respect to the set of 3D hotspot volumes. To make it do so, memory and A system equipped with these features.

91. A system for automated analysis of time-series medical images of a target, wherein the system is The processor of a computing device, A memory wherein the memory has instructions stored thereon, and when an instruction is executed by the processor, the processor, (a) Receiving and / or accessing the time-series medical images of the subject, (b) Identifying multiple hotspots within each of the medical images, and the processor determining one, two, or all three of (i), (ii), and (iii), such as (i) a change in the number of identified lesions, (ii) a change in the total volume of the identified lesions, and (iii) a change in the PSMA-weighted total volume. To make it do so, memory and A system equipped with these features.

92. A system for analyzing multiple medical images of a target, wherein the system is The processor of a computing device, A memory wherein the memory has instructions stored thereon, and when an instruction is executed by the processor, the processor, (a) Receiving and / or accessing the multiple medical images of the subject, and the processor obtaining multiple 3D hotspot maps, each corresponding to a specific medical image (of the multiple), and identifying one or more hotspots within the specific medical image, (b) For each of the multiple medical images, a machine learning module is used to determine a corresponding 3D anatomical compartment map that identifies a set of organ regions within the specific medical image, thereby generating multiple 3D anatomical compartment maps. (c) (i) determining the identification of one or more lesion correspondences using the plurality of 3D hotspot maps and (ii) the plurality of 3D anatomical compartmentalization maps, wherein each (lesion correspondence) is determined to identify two or more corresponding hotspots in different medical images and to represent the same underlying physical lesion within the subject. (d) Determining the value of one or more measured values ​​based on the plurality of 3D hotspot maps and the identification of one or more lesions. To make it do so, memory and A system equipped with these features.

93. A system for analyzing multiple medical images of a target, wherein the system is The processor of a computing device, A memory wherein the memory has instructions stored thereon, and when an instruction is executed by the processor, the processor, (a) Obtaining a first 3D hotspot map relating to the subject, (b) Obtaining a first 3D anatomical compartment map associated with the first 3D hotspot map, (c) Obtaining a second 3D hotspot map relating to the subject, (d) Obtaining a second 3D anatomical compartment map associated with the second 3D hotspot map, (e) Using the first 3D anatomical compartment map and the second 3D anatomical compartment map, determine the alignment field, (f) Using the alignment field, the first 3D hotspot map and the second 3D hotspot map are aligned, thereby generating a pair of co-aligned 3D hotspot maps. (g) Using the co-aligned pair of 3D hotspot maps, determine the identification of one or more lesions, (h) To store and / or provide the identification of one or more lesions for display and / or further processing To make it do so, memory and A system equipped with these features.

94. A system for analyzing multiple medical images of a target, wherein the system is The processor of a computing device, A memory wherein the memory has instructions stored thereon, and when an instruction is executed by the processor, the processor, (a) Receiving and / or accessing the multiple medical images of the subject, (b) For each of the multiple medical images, a machine learning module is used to determine a corresponding 3D anatomical compartment map that identifies a set of organ regions within the specific medical image, thereby generating multiple 3D anatomical compartment maps. (c) Using the plurality of 3D anatomical compartment maps, determine one or more alignment fields, apply the one or more alignment fields, align the plurality of medical images, and thereby generate a plurality of aligned medical images. (d) For each specific one of the plurality of aligned medical images, determine a corresponding aligned 3D hotspot map that identifies one or more hotspots within the specific aligned medical image, thereby generating a plurality of aligned 3D hotspot maps. (e) Using the plurality of 3D-aligned hotspot maps, determine the identification of one or more lesion correspondences, wherein each (lesion correspondence) identifies two or more corresponding hotspots in different medical images and is determined to represent the same underlying physical lesion within the subject. (f) Determining the value of one or more measured values ​​based on the plurality of 3D hotspot maps and the identification of one or more lesions. To make it do so, memory and A system equipped with these features.

95. A system for analyzing multiple medical images of a target, wherein the system is The processor of a computing device, A memory wherein the memory has instructions stored thereon, and when an instruction is executed by the processor, the processor, (a) To acquire a first 3D anatomical image and a first 3D functional image of the subject, (b) Obtaining a second 3D anatomical image and a second 3D functional image of the subject, (c) Obtaining a first 3D anatomical compartment map based on the first 3D anatomical image, (d) Obtaining a second 3D anatomical compartment map based on the second 3D anatomical image, (e) Using the first 3D anatomical compartment map and the second 3D anatomical compartment map, determine the alignment field, (f) Using the 3D alignment field, the second 3D functional image is aligned with the first 3D functional image, thereby generating an aligned version of the second 3D functional image. (g) Obtaining a first 3D hotspot map associated with the first functional image, (h) Determining a second 3D hotspot map using the aligned version of the second 3D functional image, wherein the second 3D hotspot map is thereby aligned with the first 3D hotspot map. (i) Using the first 3D hotspot map and the second 3D hotspot map aligned thereto, determine the identification of one or more lesions, (j) To store and / or provide the identification of one or more lesions for display and / or further processing To make it do so, memory and A system equipped with these features.

96. An automated or semi-automated system for whole-body assessment of a patient with metastatic prostate cancer to assess disease progression and / or treatment effectiveness, wherein the system is: The processor of a computing device, A memory wherein the memory has instructions stored thereon, and when an instruction is executed by the processor, the processor, (a) Receiving a first prostate-specific membrane antigen (PSMA)-targeted positron emission tomography (PET) image of the subject (the first PSMA-PET image) and a first 3D anatomical image of the subject, wherein the first 3D anatomical image of the subject is acquired simultaneously with, immediately after, or immediately before, the first PSMA-PET image, such that the first 3D anatomical image and the first PSMA-PET image correspond to a first date, and the image depicts an area of ​​the subject's body that is sufficiently large to cover the area of ​​the body in which the metastatic prostate cancer is spreading. (b) Both receive a second PSMA-PET image of the subject and a second 3D anatomical image of the subject, acquired on the second date following the first date. (c) Automatically determine the alignment field using the markers automatically identified in the first and second 3D anatomical images, and the processor aligns the first and second PSMA-PET images using the determined alignment field. (d) Using the first and second PSMA-PET images that have been matched in this manner, the changes in the disease from the first date to the second date are automatically detected. To make it do so, memory and A system equipped with these features.

97. A system for quantifying and reporting disease burden in patients with cancer and / or at risk of cancer, wherein the system The processor of a computing device, A memory wherein the memory has instructions stored thereon, and when an instruction is executed by the processor, the processor, (a) Obtaining medical images of the patient, (b) Detecting one or more hotspots in the medical image, each hotspot corresponding to a specific 3D volume in the medical image and representing a potential underlying physical lesion within the subject, (c) For each specific lesion class of multiple lesion classes that represent a specific tissue region and / or lesion subtype, Identifying a corresponding subset of one or more hotspots as belonging to a specific lesion class, Based on the corresponding subset of the hotspots, determine one or more patient index values ​​that quantify the disease burden within and / or associated with the particular lesion class. To do, (d) To cause a graphical representation of the patient index value calculated for each of the plurality of lesion classes, thereby providing the user with a graphical report summarizing the tumor burden within a specific tissue region and / or associated with a specific lesion subtype. To make it do so, memory and A system equipped with these features.

98. A system for characterizing and reporting individual lesions detected based on imaging assessments of patients with cancer and / or at risk of cancer, wherein the system The processor of a computing device, A memory wherein the memory has instructions stored thereon, and when an instruction is executed by the processor, the processor, (a) Obtaining medical images of the patient, (b) Detecting a collection of one or more hotspots in the medical image, wherein each hotspot in the collection corresponds to a specific 3D volume in the medical image and represents a potential underlying physical lesion within the subject, (c) Assigning one or more lesion class labels to each of the one or more hotspots in the set, wherein each lesion class label represents a specific tissue region and / or lesion subtype, and the hotspots are identified as representing potential lesions located within the specific tissue region and / or belonging to the lesion subtype. (d) For each specific hotspot quantification measurement value among one or more individual hotspot quantification measurements, the value of the specific individual hotspot quantification measurement value for each individual hotspot in the set is calculated, (e) For each specific hotspot, at least a portion of the hotspots in the set, to cause a graphical representation to be displayed that includes the identification of the specific hotspot, the one or more lesion class labels assigned to the specific hotspot, and the one or more individual hotspot quantification values ​​calculated with respect to the specific hotspot. To make it do so, memory and A system equipped with these features.

99. A system for quantifying and reporting disease progression and / or risk over time in patients who have cancer and / or are at risk of cancer, wherein the system is: The processor of a computing device, A memory wherein the memory has instructions stored thereon, and when an instruction is executed by the processor, the processor, (a) Acquiring multiple medical images of the patient, wherein each medical image represents a scan of the patient acquired at a specific time; (b) For each specific medical image among the plurality of medical images, to detect a corresponding set of one or more hotspots within the specific medical image, wherein each hotspot corresponds to a specific 3D volume within the medical image and represents a potential underlying physical lesion within the subject, (c) For each specific patient index of one or more patient indices that measure the overall disease burden within a patient at a specific time, the value of the specific patient index for each specific medical image of the plurality of medical images is determined based on the corresponding set of hotspots detected with respect to the specific medical image, thereby determining for each specific patient index of one or more patient indices a set of values ​​that track changes in disease burden over time as measured by the specific patient index value, (d) To cause a graphical representation of a set of values ​​relating to at least one of the patient index values, thereby conveying a time-series measurement of the disease progression for the patient. To make it do so, memory and A system equipped with these features.

100. A system for automatically processing a 3D image of a subject and determining the value of one or more patient indices to measure disease burden and / or risk related to the subject, wherein the system The processor of a computing device, A memory wherein the memory has instructions stored thereon, and when an instruction is executed by the processor, the processor, (a) Receiving a 3D functional image of the object acquired using a functional imaging modality, (b) Partitioning a plurality of 3D hotspot volumes within the 3D functional image, wherein each 3D hotspot volume corresponds to a local area of ​​increased intensity relative to its surroundings, represents a potential cancerous lesion within the subject, and thereby obtains a set of 3D hotspot volumes. (c) For each specific hotspot quantification metric among one or more individual hotspot quantification metric values, the value of the specific individual hotspot quantification metric for each individual 3D hotspot volume in the set, wherein, with respect to a specific individual 3D hotspot volume, each hotspot quantification metric quantifies the properties of the specific 3D hotspot volume and is a specific function of the intensity and / or number of individual voxels within the specific 3D hotspot volume. (d) Determining the values ​​of one or more patient indices, each of the at least portion of the patient indices being associated with one or more specific individual hotspot quantification measurements and calculated using the particular function, which includes the intensities and / or numbers of voxels in a combined hotspot volume comprising at least a portion of the set of 3D hotspot volumes. To make it do so, memory and A system equipped with these features.

101. A system for automatically determining the prognosis of a subject suffering from prostate cancer from one or more medical images of the subject, wherein the system is The processor of a computing device, A memory wherein the memory has instructions stored thereon, and when an instruction is executed by the processor, the processor, (a) Receiving one or more images of the subject and / or accessing one or more images of the subject, (b) Automatically determine the quantitative assessment of one or more prostate cancer lesions from one or more images, (c)(b) The prognosis for the subject is to be automatically determined from the quantitative assessment in (c)(b), wherein the prognosis includes one or more of the following for the subject: (I) expected survival rate, (II) expected time to disease progression, (III) expected time to radiological progression, (IV) risk of simultaneous metastasis, and (V) risk of future metastatic metastasis. To make it do so, memory and A system equipped with these features.

102. The system according to claim 101, wherein the quantitative assessment of the one or more prostate cancer lesions determined in step (b) comprises one or more of the following: (A) total tumor volume, (B) change in tumor volume, (C) total SUV, and (D) PSMA score, and the prognosis for the subject determined in step (c) comprises one or more of the following: (E) expected survival rate, (F) time to progression, and (G) time to radiological progression.

103. The system according to claim 101, wherein the quantitative assessment of the one or more prostate cancer lesions determined in step (b) includes one or more characteristics of PSMA expression in the prostate, and the prognosis for the subject determined in step (c) includes the risk of synchronous metastasis and / or the risk of future metachronous metastasis.

104. A system for automatically determining the response to treatment for a subject suffering from prostate cancer from multiple medical images of the subject, wherein the system is The processor of a computing device, A memory wherein the memory has instructions stored thereon, and when an instruction is executed by the processor, the processor, (a) The processor of the computing device receives and / or accesses a plurality of images of the subject, wherein at least a first image of the plurality of images is acquired prior to the administration of the treatment, and at least a second image of the plurality of images is acquired following the administration of the treatment, (b) Automatically determine the quantitative assessment of one or more prostate cancer lesions from the aforementioned images, (c)(b) The quantitative assessment in (c)(b) automatically determines whether the subject is responding to the treatment and / or the extent to which the subject is responding to the treatment. To make it do so, memory and A system equipped with these features.

105. A system for automatically identifying whether a subject with prostate cancer is likely to benefit from a specific treatment for prostate cancer, using multiple medical images of the subject, wherein the system The processor of a computing device, A memory wherein the memory has instructions stored thereon, and when an instruction is executed by the processor, the processor, (a) Receiving and / or accessing multiple images of the subject, (b) Automatically determine the quantitative assessment of one or more prostate cancer lesions from the aforementioned images, (c)(b) The quantitative assessment in (c)(b) automatically determines whether the subject is likely to benefit from the specific treatment for prostate cancer. To make it do so, memory and A system equipped with these features.

106. A therapeutic agent for use in the treatment of subjects having and / or at risk of having a particular disease, wherein the subject (i) has been administered a first cycle of the therapeutic agent and imaged, and (ii) has been identified as a responder to the therapeutic agent using the method of any one of claims 1 to 58.

107. A second therapeutic agent for use in the treatment of subjects having and / or at risk of having a particular disease, wherein the subject (i) has been administered and imaged an initial cycle of the first therapeutic agent, and (ii) has been identified as a non-responder to the first therapeutic agent using the method of any one of claims 1 to 58.

108. A method for analyzing a time-series composite medical image of a subject, wherein each composite medical image is obtained at a different time and includes an anatomical image and a corresponding nuclear medicine image, and the method is (a) The processor of the computing device receives and / or accesses the time-series composite medical images of the subject, and the processor obtains a plurality of 3D hotspot maps, each corresponding to a specific composite medical image, and identifies one or more hotspots within the nuclear medicine image portion of the specific composite medical image. (b) For each specific composite medical image in the time series of composite medical images, the processor determines a corresponding 3D anatomical compartment map that identifies a set of organ regions within the anatomical image portion of the specific composite medical image, thereby generating a plurality of 3D anatomical compartment maps. (c) The processor determines, using (i) the plurality of 3D hotspot maps and (ii) the plurality of 3D anatomical compartment maps, to identify one or more lesion correspondences, each of which identifies two or more corresponding hotspots in different composite medical images and is determined to represent the same underlying physical lesion within the subject. (d) The processor determines the value of one or more measured values ​​based on the plurality of 3D hotspot maps and the identification of one or more lesions. Includes, The plurality of 3D hotspot maps include (i) a first 3D hotspot map corresponding to a first composite medical image, and (ii) a second 3D hotspot map corresponding to a second composite medical image. The plurality of 3D anatomical compartment maps include (i) a first 3D anatomical compartment map that identifies the set of organ regions within the anatomical image portion of the first composite medical image, and (ii) a second 3D anatomical compartment map that identifies the set of organ regions within the anatomical image portion of the second composite medical image. Step (c) is a method comprising using the first 3D anatomical compartment map and the second 3D anatomical compartment map to (i) align the first hotspot map with (ii) the second hotspot map.

109. The method according to claim 108, wherein each composite medical image is a PET / CT image including a CT image and a PET image that are co-aligned with each other.

110. The method according to claim 108, wherein the one or more measured values ​​quantify the change in value and / or the radiopharmaceutical uptake rate of individual hotspots and / or the underlying physical lesions they represent over time.

111. The method according to claim 108, wherein the one or more measured values ​​measure the overall disease burden and / or changes thereof.

Images