Rapid dose evaluation method and device for individualized targeted radionuclide treatment

Abstract

The invention relates to the technical field of dosimetry, in particular to a rapid dose evaluation method and device for individualized targeted radionuclide treatment.The method comprises the steps that labeled nuclides, drug ligands and image data of a patient are obtained, and corresponding standard S values are selected according to the labeled nuclides and the drug ligands; the method comprises the following steps: segmenting first-time-phase CT image data to obtain a segmentation mask of a target organ, calculating an individualized S value based on the segmentation mask, generating an individualized voxel model, registering the SPECT image data and the CT image data to obtain organ-level source item parameters and voxel-level source item parameters, and calculating an individualized S value based on the individualized voxel model to obtain an organ-level source item parameter and a voxel-level source item parameter; and according to the organ-level source item parameter and the voxel-level source item parameter, respectively obtaining an organ-level dose evaluation result and a voxel-level dose evaluation result, and comprehensively evaluating source target organ dose distribution. Therefore, the problems that the dose conversion coefficient lacks calibration, the voxel model is difficult to balance the calculation speed and precision and the like are solved, errors caused by individualized differences are reduced, and the calculation accuracy and speed are guaranteed.

Description

Technical Field

[0001] This application relates to the technical field of dosimetry, and particularly to a method and device for rapid dose assessment of individualized targeted radionuclide therapy. Background Art

[0002] Individualized precision medicine is a medical model that customizes treatment strategies based on patients' physiological characteristics and disease heterogeneity. It can promote the transformation of the medical model from "group universality" to "individual optimization", achieving efficient resource utilization and innovation of treatment paradigms.

[0003] Targeted radionuclide therapy (TRT) has become an important means of cancer treatment due to its high selectivity for tumor lesions and is one of the key development directions of individualized precision medicine. In this context, radiation dosimetry, as the cornerstone of precision treatment, provides crucial scientific basis for optimizing treatment plans and balancing tumor response rates and organ-at-risk toxicity by quantifying the distribution and energy deposition of radiopharmaceuticals in the target area and normal tissues. Research shows that the average absorbed dose of tumors in patients in the individualized dosimetry group increased by about 107.1 Gy, the median overall survival was extended to 26.6 months, and the overall response rate increased to 50%, showing significant improvement compared with the standard dosimetry group without increasing the risk of adverse reactions. With the development of targeted radionuclide therapy and diagnostic and therapeutic integrated imaging, accurate patient-specific dosimetry has become increasingly necessary to meet the needs of these advanced treatment methods.

[0004] The individualized internal radiation dose assessment system is the core tool for achieving precise control of patients' radiation doses. However, related technologies still face many challenges: (1) The dose conversion coefficient lacks local calibration (such as insufficient adaptability of organ parameters based on European and American populations); (2) The rapid calculation methods in related technologies ignore individual differences; (3) It is difficult to balance efficiency and accuracy (the calculation of high-resolution voxel models takes too long); (4) The collaborative framework of multimodal technologies is fragmented and the data stream is asynchronous, etc. Therefore, developing a system architecture that can efficiently generate individualized voxel models and balance calculation speed and accuracy has important clinical value for clinical individualized dose optimization. Summary of the Invention

[0005] This application provides a method, device, electronic device, and storage medium for rapid dose assessment of individualized targeted radionuclide therapy to solve problems such as the lack of local calibration of dose conversion coefficients, the neglect of individual differences in rapid calculation methods, and the low calculation efficiency of high-resolution voxel models, reduce systematic errors caused by ethnic differences, shorten the calculation time while ensuring calculation accuracy, and reduce deployment costs.

[0006] The first aspect of the present application provides a method for rapid dose assessment of individualized targeted radionuclide therapy, including the following steps: Obtain the labeled radionuclide, drug ligand and imaging data of the patient, wherein the imaging data includes SPECT imaging data and CT imaging data; Obtain the corresponding standard S value from a preset population-specific SAF database according to the labeled radionuclide and drug ligand, wherein the preset population-specific SAF database is constructed based on a reference human model in a preset region; Segment the first-phase CT imaging data to obtain a segmentation mask of the target organ. Based on the segmentation mask, calculate the individualized S value according to the standard S value and the target correction factor, and generate an individualized voxel model of the patient according to the segmentation mask; Register the SPECT imaging data and the CT imaging data to obtain organ-level source term parameters and voxel-level source term parameters. Based on the individualized S value, use a preset MIRD calculation method and a preset MCs calculation method to generate an organ-level dose assessment result according to the organ-level source term parameters, and input the voxel-level source term parameters and the individualized voxel model into a preset fast Monte Carlo model to obtain a voxel-level dose assessment result. Obtain the target dose assessment result according to the organ-level dose assessment result and the voxel-level dose assessment result.

[0007] Optionally, in some embodiments, the obtaining the corresponding standard S value from a preset population-specific SAF database according to the labeled radionuclide and drug ligand includes: Judge whether the preset SAF database has a standard S value matching the labeled radionuclide and the drug ligand according to the labeled radionuclide and the drug ligand; If the preset SAF database has the standard S value and the version of the standard S value meets the preset version verification condition, output the standard S value.

[0008] Optionally, in some embodiments, after judging whether the preset SAF database has a standard S value matching the labeled radionuclide and the drug ligand, it includes: If the preset SAF database does not have the standard S value, obtain the particle energy spectrum and branching ratio from a preset radionuclide decay database according to the labeled radionuclide and the drug ligand, calculate a new S value according to the particle energy spectrum and the branching ratio, and add the new S value to the preset population-specific SAF database.

[0009] Optionally, in some embodiments, the calculating the individualized S value according to the standard S value and the target correction factor based on the segmentation mask includes: Calculate the actual mass of the target organ according to the segmentation mask, and determine a target correction factor according to the actual mass and a preset standard mass; Calculate the individualized S value according to the target correction factor and the standard S value.

[0010] Optionally, in some embodiments, the registering the SPECT image data and the CT image data to obtain organ-level source term parameters and voxel-level source term parameters includes: Register the multi-phase CT image data based on the first-phase CT image data to obtain conversion matrices for each phase; Based on the conversion matrices for each phase, register the SPECT image data and the CT image data to obtain multi-phase count rates for each voxel, and based on a preset count-activity conversion factor, convert the multi-phase count rates for each voxel into activity-time curves; Obtain the organ-level source term parameters and the voxel-level source term parameters based on the activity-time curves.

[0011] An embodiment of the second aspect of the present application provides a device for rapid dose assessment of individualized targeted radionuclide therapy, including: A first acquisition module, configured to acquire a labeled radionuclide, a drug ligand, and image data of a patient, where the image data includes SPECT image data and CT image data; A second acquisition module, configured to acquire a corresponding standard S value from a preset population-specific SAF database according to the labeled radionuclide and the drug ligand, where the preset population-specific SAF database is constructed based on a reference human model in a preset region; A segmentation module, configured to segment the first-phase CT image data to obtain a segmentation mask of the target organ, and based on the segmentation mask, calculate an individualized S value according to the standard S value and a target correction factor, and generate an individualized voxel model of the patient according to the segmentation mask; An evaluation module, configured to register the SPECT image data and the CT image data to obtain organ-level source term parameters and voxel-level source term parameters, and based on the individualized S value, use a preset MIRD calculation method and a preset MCs calculation method to generate an organ-level dose evaluation result according to the organ-level source term parameters, and input the voxel-level source term parameters and the individualized voxel model into a preset fast Monte Carlo model to obtain a voxel-level dose evaluation result, and obtain a target dose evaluation result according to the organ-level dose evaluation result and the voxel-level dose evaluation result.

[0012] Optionally, in some embodiments, the second acquisition module includes: A judgment unit, configured to judge whether a standard S value matching the labeled radionuclide and the drug ligand exists in the preset SAF database according to the labeled radionuclide and the drug ligand; An output unit, configured to output the standard S value when the standard S value exists in the preset SAF database and the version of the standard S value meets a preset version verification condition.

[0013] Optionally, in some embodiments, after judging whether a standard S value matching the labeled radionuclide and the drug ligand exists in the preset SAF database, the judgment unit includes: An update subunit, configured to obtain a particle energy spectrum and a branching ratio from a preset radionuclide decay database according to the labeled radionuclide and the drug ligand when the standard S value does not exist in the preset SAF database, calculate a new S value according to the particle energy spectrum and the branching ratio, and add the new S value to the preset population-specific SAF database.

[0014] Optionally, in some embodiments, the segmentation module includes: A determination unit, configured to calculate the actual mass of the target organ according to the segmentation mask, and determine a target correction factor according to the actual mass and a preset standard mass; A calculation unit, configured to calculate the individualized S value according to the target correction factor and the standard S value.

[0015] Optionally, in some embodiments, the evaluation module includes: A first registration unit, configured to register multi-phase CT image data based on the first-phase CT image data to obtain conversion matrices for each phase; A second registration unit, configured to register the SPECT image data and the CT image data based on the conversion matrices for each phase to obtain multi-phase count rates for each voxel, and convert the multi-phase count rates for each voxel into activity-time curves based on a preset count-activity conversion factor; A generation unit, configured to obtain the organ-level source term parameters and the voxel-level source term parameters based on the activity-time curves.

[0016] An embodiment of the third aspect of the present application provides an electronic device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the program to implement the method for rapid dose evaluation of individualized targeted radionuclide therapy as described in the above embodiments.

[0017] The fourth aspect of the present application provides a computer-readable storage medium, on which a computer program is stored, and the program is executed by a processor to implement the method for rapid dose assessment of individualized targeted radionuclide therapy as described in the above embodiments.

[0018] Therefore, the present application has the following beneficial effects: (1) In view of the problems existing in the traditional dose conversion system, such as the deviation of population anatomical characteristics, the bottleneck of clinical high-load operation efficiency, and the rapid conversion requirements of new targeted drugs (such as 225 Ac-PSMA, 64 Cu-DOTATATE) from preclinical research to phase III trials, the Chinese population-specific dose conversion coefficient library constructed in the embodiments of the present application realizes multi-dimensional optimization through a hierarchical heterogeneous architecture design. While maintaining the rigor of calculating the marketed target drug dose in the MIRD paradigm, this architecture significantly improves the dose evaluation efficiency of new target drugs in preclinical and phase III experiments.

[0019] (2) The individualized dose conversion coefficient calculation method proposed in the present application realizes the migration from a population model to individualized calculation through multi-modal image fusion technology. This technical path ensures clinical operability while correcting the changes in dose conversion coefficients caused by anatomical variations of individual patients, etc., and helps to improve the accuracy of patient dose evaluation.

[0020] (3) The present application innovatively integrates heterogeneous computing and dynamic source term reconstruction technology to achieve a breakthrough in the efficiency and accuracy of nuclear medicine dose assessment. The GPU-accelerated Monte Carlo system based on the CUDA architecture reduces the single-case calculation time from 60 hours to 0.5 hours, with an acceleration of more than 80 times. By means of the dual strategies of CT-based elastic registration and SPECT rigid registration, the problem of spatio-temporal mismatch of multi-temporal images is overcome, and the voxel activity integration time is reduced to 1 / 10 of the traditional time by combining the piecewise fitting-trapezoidal integration algorithm. The full-process automation significantly reduces the clinical operation time, and has both high efficiency and clinical applicability, providing a systematic solution for precise diagnosis and treatment.

[0021] (4) The present application integrates a variety of differentiated dose assessment schemes. Clinicians can not only perform rapid organ dose evaluation, but also carry out precise sub-organ level dose assessment relying on the patient-specific voxel model, achieving a double improvement in clinical efficiency and calculation accuracy. By deeply integrating core technologies such as image processing, activity modeling, and dose calculation, an automated dose assessment system is established. The system significantly reduces the human-computer interaction requirements in the source term information generation and dose calculation links, and all calculation modules achieve individualized algorithm optimization, providing reliable technical support for precise nuclear medicine diagnosis and treatment.

[0022] The additional aspects and advantages of the present application will be partially given in the following description, partially become obvious from the following description, or be understood through the practice of the present application. Description of the Drawings

[0023] The above and / or additional aspects and advantages of the present application will become apparent and be readily understood from the following description of embodiments in conjunction with the accompanying drawings, where: Figure 1 FIG. is a flowchart of a method for rapid dose assessment of individualized targeted radionuclide therapy provided according to an embodiment of the present application; Figure 2 FIG. is a schematic diagram of the principle for constructing a preset population-specific SAF database provided according to an embodiment of the present application; Figure 3 FIG. is a schematic diagram of the principle for calculating an individualized S value provided according to an embodiment of the present application; Figure 4 FIG. is a schematic diagram of the principle of a preset fast Monte Carlo model provided according to an embodiment of the present application; Figure 5 FIG. is a schematic diagram of a multi-modal tightly coupled system architecture provided according to an embodiment of the present application; Figure 6 FIG. is a block diagram of a device for rapid dose assessment of individualized targeted radionuclide therapy provided according to an embodiment of the present application; Figure 7 FIG. is a schematic diagram of the structure of an electronic device provided according to an embodiment of the present application. Detailed Description of the Embodiments

[0024] Embodiments of the present application will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to explain the present application and should not be construed as limiting the present application.

[0025] Before introducing the method for rapid dose assessment of individualized targeted radionuclide therapy according to the embodiments of the present application, the relevant technical principles will be introduced first.

[0026] In current clinical practice, patient dose assessment is mainly based on the MIRD (Medical Internal Radiation Dose) method, also known as the organ S-value method. This method multiplies the organ cumulative activity by the organ S-value provided in the MIRD handbook (which describes the average absorbed dose produced in the target area per unit cumulative activity) to obtain the average organ dose. However, its core limitations are reflected in two aspects: First, the organ S-values are calculated based on a standard reference human model, ignoring the anatomical differences of individual patients and the heterogeneity of radionuclide spatial distribution, resulting in systematic errors in individualized dose assessment; Second, the reference human model on which the relevant S-value calculations depend is based on the typical physical parameters of the Caucasian race, and the lack of anatomical feature adaptability and localization correction further amplifies the systematic errors in dose assessment between the tumor target area and normal tissues. Taking the liver as an example, the mass of the liver in the Chinese male reference human model is 1.46 kg, while the mass of the liver in the ICRP (International Commission on Radiological Protection) male reference human is 2.36 kg, resulting in a calculation deviation of approximately 10%-25% in the photon specific absorption fraction.

[0027] Monte Carlo simulation (MCs) is considered the gold standard for achieving reliable dose calculations in a clinical environment. Combining advanced medical molecular imaging techniques (such as PET / CT and SPECT / CT) can accurately obtain the voxel-level three-dimensional dose distribution map of individual patients. However, the problems of slow calculation speed and high cost limit its application in routine clinical procedures. Related technologies optimize simulation parameters to improve accuracy and efficiency by combining methods with multiple resolutions and cut-off energies, but still require about 60 hours of calculation time. Such high computational loads are difficult to adapt to in a clinical scenario, especially in the current situation of a large patient population in China, and the practical application feasibility is low.

[0028] The Voxel S-Value (VSV) method regards each voxel as a separate source voxel and the target voxel of adjacent voxels, and uses MCs to calculate the voxel S-value (dose conversion coefficient for a specific radionuclide and voxel size) and the dose kernel matrix M, and accurately calculates the energy transfer and deposition in each voxel through convolution with the three-dimensional source term distribution data. It is regarded as an efficient alternative to MCs because of its low computational cost and fast speed. The calculation of VSVs usually assumes that the point source is located in an infinitely homogeneous soft tissue or water medium, but due to the lack of consideration of tissue and organ heterogeneity, errors are often introduced at the tissue or tissue junction with a large density difference from water.

[0029] Current mainstream nuclear medicine dose calculation systems (such as HERMES MEDICAL SOLUTIONS, MIM, PMOD, and QDOSE, etc.) basically implement the dose prediction function using the above publicly disclosed technical framework, but there is a general disconnection between the algorithm efficiency and clinical needs: on the one hand, although the Monte Carlo method can ensure the accuracy of dose calculation, its huge computational amount leads to a single case analysis taking up to several hours (such as HERMES); on the other hand, although fast analytical algorithms (such as the MIRD method and the voxel S-value method) can be shortened to the minute level, the simplified assumptions for heterogeneous tissues and non-uniform activity distributions will introduce systematic biases (such as QDOSE and MIM). In addition, for the purpose of commercial data control, the core servers of the above systems are all deployed overseas, resulting in the need for cross-border transmission of patients' sensitive images and dose data, and it is easy to cause privacy leakage risks due to vulnerabilities in overseas servers or malicious attacks.

[0030] The individualized dose calculation of targeted radiopharmaceuticals needs to rely on a multi-modal technology collaborative framework. Its core lies not only in the optimization of the dose algorithm, but also in the organic coupling of key modules such as accurate medical image segmentation, dynamic modeling of individualized voxel models (mapping of tissue density and activity distribution), and pharmacokinetic parameter inversion based on multi-time point molecular imaging data. However, the systems in related technologies generally have the pain points of modular fragmentation (difficulty in cross-platform migration due to the lack of data interface standards), high operation complexity (requiring manual execution of multi-software switching and parameter resetting), and imbalance between computational efficiency and accuracy (most integrated platforms use fast analytical algorithms to replace Monte Carlo simulation to shorten the process time), which seriously restricts the clinical practical value.

[0031] In summary, the related technologies have the following defects: (1) Data model mismatch: double deviation of population characteristics and individualized parameters.

[0032] Ⅰ. Population deviation: The dose conversion coefficient (such as the S-value) depends on the Caucasian anatomical database, and there are significant differences in key parameters such as organ volume between Asian populations, resulting in large dose assessment errors.

[0033] Ⅱ. Individual inaccuracy: The anatomical parameters of the reference standard human model are fixed, and the changes in the dose conversion coefficient caused by anatomical variations of individual patients are not corrected.

[0034] (2) Limitations of the calculation model: the contradiction between the uniform assumption at the organ level and the non-uniform demand at the voxel level.

[0035] Ⅰ. Lack of spatial distribution information: Most related dose systems are based on the assumption of uniform distribution of radionuclides in organs, taking the cumulative activity of organs as the source term input to calculate the uniform dose distribution results at the organ level, ignoring the non-uniform spatial distribution characteristics highly expressed in the tumor target area, resulting in the under-treatment risk of underestimated target area dose and overestimated dose of organs at risk; there is a lack of dose calculation based on voxel-level non-uniform source terms (such as the activity distribution at multiple time points of SPECT / CT).

[0036] (3)Imbalance between algorithm efficiency and accuracy: The opposition between Monte Carlo and analytical methods.

[0037] Ⅰ. High cost of high precision: Although Monte Carlo simulation can support voxel-level dose calculation, the calculation time for a single case is as long as 60 hours, and it requires high-cost hardware such as a CPU cluster.

[0038] Ⅱ. Defects of fast algorithms: Although analytical algorithms (MIRD method, voxel S-value method) compress the time to the minute level, they simplify tissue non-uniformity (CT value - density mapping error > 10%) and geometric approximation, resulting in systematic biases.

[0039] (4)Fragmentation of system architecture: Double obstacles of modular redundancy and hardware dependence.

[0040] Ⅰ. Fragmentation of processes: Modules such as organ segmentation, voxel modeling, and pharmacokinetic parameter inversion are scattered and independent, and manual execution of data export and format conversion is required, resulting in long operation time and increased risk of data loss.

[0041] Ⅱ. High hardware cost: To meet the general Monte Carlo acceleration calculation requirements, it is necessary to deploy a multi-node CPU cluster, which far exceeds the affordability of primary medical institutions.

[0042] In view of the above problems, the present application provides a method for rapidly evaluating the dose of individualized targeted radionuclide therapy. In this method, by obtaining the labeled radionuclide, drug ligand and imaging data of the patient, and obtaining the corresponding standard S value from the preset population-specific SAF database according to the labeled radionuclide and drug ligand, segmenting the first-phase CT imaging data to obtain the segmentation mask of the target organ, based on the segmentation mask, calculating the individualized S value according to the standard S value and the target correction factor, then generating the individualized voxel model of the patient, registering the SPECT imaging data and the CT imaging data to obtain the organ-level source term parameters and voxel-level source term parameters, based on the individualized S value, using the preset MIRD calculation method and the preset MCs calculation method to generate the organ-level dose evaluation result according to the organ-level source term parameters, and inputting the voxel-level source term parameters and the individualized voxel model into the preset fast Monte Carlo model to obtain the voxel-level dose evaluation result, and obtaining the target dose evaluation result according to the organ-level dose evaluation result and the voxel-level dose evaluation result. Thus, the problems of lack of population adaptability, imbalance between calculation accuracy and efficiency, and discretization of the system architecture in dose evaluation are solved, the systematic error caused by ethnic differences can be reduced, the calculation time can be shortened while ensuring the calculation accuracy, and the deployment cost can be reduced.

[0043] Specifically, Figure 1 is a schematic flow chart of a method for rapidly evaluating the dose of individualized targeted radionuclide therapy provided by an embodiment of the present application.

[0044] As Figure 1 shown, the method for rapidly evaluating the dose of individualized targeted radionuclide therapy includes the following steps: In step S101, obtain the labeled radionuclide, drug ligand and imaging data of the patient, where the imaging data includes SPECT imaging data and CT imaging data.

[0045] Among them, the SPECT imaging data can be replaced by PET imaging data.

[0046] In step S102, obtain the corresponding standard S value from the preset population-specific SAF database according to the labeled radionuclide and drug ligand, where the preset population-specific SAF database is constructed based on the reference human model in the preset region.

[0047] Preferably, the preset population-specific SAF database in the embodiment of the present application is the Chinese population dose conversion coefficient library established in advance by relevant personnel, and the reference human model in the preset region is the Chinese reference human model established in advance by relevant personnel.

[0048] It should be noted that the dose conversion coefficient S value refers to the ratio of the absorbed dose of the target organ to the time integral of the source organ activity under the unit cumulative activity. The S value calculation process is as follows: (1) Divide the energy range from 10 keV to 4 MeV into 200 equally logarithmically spaced energy points, and construct the Specific Absorption Fraction (SAF) for monoenergetic photons, electrons, and α particles respectively. Among them, SAF is the ratio of the energy absorbed by the target organ to the total energy and its mass, and can be calculated by the following formula

[0049] Among them, is the specific absorption fraction of the target organ R , is the energy (MeV) deposited in the target organ R , is the initial energy (MeV) of the particles emitted by the source organ, is the mass (kg) of the target organ.

[0050] (2) According to the decay energy spectrum and its branching ratio of the radionuclide, accumulate the SAF to obtain the S value of the target organ, according to the following formula. R Among them,

[0051]

[0052] Among them, is the R value of the source organ to the target organ S , is the particle energy (MeV) generated by a single decay of the radionuclide, is the decay branching ratio of this energy ray, indicates that this nuclide produces i types of energy particles.

[0053] At the clinical application level, to implement internal radiation dose, first quantitatively obtain the time distribution data of the radioactivity of the source organ through PET / SPECT molecular imaging technology, and then calculate the cumulative activity , and finally the target organ dose is obtained by the following formula:

[0054] Among them, is the target organ dose, is the cumulative activity of the source organ, is the source organ j to the target organ R of the S value.

[0055] In the calculation process of the S value, the SAF is usually calculated based on the Monte Carlo Simulation (MCs) method. The accuracy of its particle sampling and transport process is directly restricted by anatomical parameters such as organ geometric configuration, spatial distribution characteristics, and mass density.

[0056] In view of this characteristic, establishing a digital phantom with population anatomical representativeness has become the key to improving the accuracy of dose calculation. The dose conversion coefficient research for the Chinese population uses the self-developed Chinese Reference Adult Male / Female voxel model (CRAM / CRAF). The height, weight, and mass of the main organs of this model meet the relevant technical requirements, and the spatial topological relationship of the organs conforms to the statistical distribution of the anatomical characteristics of the Chinese population. When this dose conversion coefficient based on the local phantom is combined with the MIRD method framework, it can effectively reduce the dose assessment deviation caused by anatomical differences and significantly improve the rationality of dose calculation.

[0057] Optionally, in some embodiments, the corresponding standard S value is obtained from a preset population-specific SAF database according to the labeled radionuclide and the drug ligand, including: judging whether there is a standard S value in the preset SAF database that matches the labeled radionuclide and the drug ligand according to the labeled radionuclide and the drug ligand; if there is a standard S value in the preset SAF database and the version of the standard S value meets the preset version verification condition, output the standard S value.

[0058] Furthermore, in some embodiments, after judging whether there is a standard S value in the preset SAF database that matches the labeled radionuclide and the drug ligand, it includes: if there is no standard S value in the preset SAF database, obtain the particle energy spectrum and branching ratio from the preset radionuclide decay database according to the labeled radionuclide and the drug ligand, calculate a new S value according to the particle energy spectrum and the branching ratio, and add the new S value to the preset population-specific SAF database.

[0059] Specifically, the preset population-specific SAF database in the embodiments of the present application adopts a hybrid architecture design, and realizes data update through a dynamic iteration mechanism that combines offline pre-storage and online calculation, as shown in the appendix Figure 2 As shown, the system architecture includes two core modules.

[0060] For the offline pre-storage module, the preset population-specific SAF database pre-stores the SAF matrix calculated based on the CRAM phantom; the complete decay parameters of the radionuclides recommended by ICRP Publication 107; integrated commonly used clinical radioactive drugs (such as 177The organ S-value dataset of Lu-PSMA), which significantly shortens the conventional targeted radionuclide drug dose evaluation cycle through the ready-to-use call of the standardized dose conversion coefficient.

[0061] For the online calculation module, an adaptive calculation process is constructed: the user submits the radionuclide and ligand, and the system retrieves the matching organ S-value list in the pre-stored library. If it exists and passes the version verification, the dose conversion coefficient in the standard format is directly output.

[0062] If the target combination is not retrieved, an online report is returned, and then the coverage integrity of the CRAM phantom organ library is verified and the SAF matrix offline database is optimized; for the defined source organ, the radionuclide decay database in ICRP Report No. 107 is called to obtain the particle energy spectrum and branching ratio, and the target organ dose coefficient is generated based on the SAF accumulation equation. Finally, the new calculation results are dynamically incorporated into the preset population-specific SAF database after statistical verification.

[0063] In the embodiment of the present application, in view of the group anatomical feature deviation, the clinical high-load operation time bottleneck existing in the traditional dose conversion system, and the rapid conversion requirements of new targeted drugs (such as 225 Ac-PSMA, 64 Cu-DOTATATE) from preclinical research to phase III trials, the Chinese population-specific dose conversion coefficient library constructed in the embodiment of the present application realizes multi-dimensional optimization through a hierarchical heterogeneous architecture design. While maintaining the rigor of calculating the marketed target drug dose in the MIRD paradigm, this architecture significantly improves the dose evaluation efficiency of new target drugs in preclinical and phase III experiments.

[0064] In step S103, the first-phase CT image data is segmented to obtain the segmentation mask of the target organ. Based on the segmentation mask, the individualized S-value is calculated according to the standard S-value and the target correction factor, and the individualized voxel model of the patient is generated according to the segmentation mask.

[0065] Optionally, in some embodiments, calculating the individualized S-value according to the standard S-value and the target correction factor based on the segmentation mask includes: calculating the actual mass of the target organ according to the segmentation mask, determining the target correction factor according to the actual mass and the preset standard mass; calculating the individualized S-value according to the target correction factor and the standard S-value.

[0066] Specifically, the calculation of the individualized S-value proposed in the embodiment of the present application realizes the migration from the population model to the individualized calculation through multi-modal image fusion technology.

[0067] Such as Figure 3As shown, first, based on the first-phase CT image data in DICOM format, the system automatically analyzes the patient-specific organ topology through a deep learning segmentation engine (3D U-Net architecture). The automatic segmentation results are corrected through human-computer collaboration: the physician manually corrects the contours of key organs (such as ectopic thyroid), and the automatic / manual segmentation masks are fused through algorithms to generate the segmentation contours of the target organs, which conform to the three-dimensional contours of the source / target organs. Subsequently, the pre-stored organ S-value matrix is called, and the tissue density database of ICRP Publication 110 is retrieved synchronously.

[0068] Based on the source / target organ masks of the patient, the total mass of the organ is calculated through the voxel mass integration formula, and the formula is as follows: ; Among them, is the actual mass of the target organ, is the density of the material corresponding to the voxel, is the voxel position label, is the voxel volume microelement.

[0069] Subsequently, the anatomical difference compensation calculation of the organ S-value is performed, and the organ mass ratio correction factor is established, and the formula is as follows: ; Among them, is the target correction factor, is the organ mass in the CRAM model, is the actual organ mass of the patient.

[0070] Finally, the standard S-value of the organ based on the CRAM model is multiplied by the correction factor to obtain the individualized organ S-value. This technical path ensures clinical operability and helps improve the accuracy of patient dose evaluation under the condition of correcting the changes in dose conversion coefficients caused by anatomical variations of individual patients.

[0071] In step S104, the SPECT image data and CT image data are registered to obtain organ-level source term parameters and voxel-level source term parameters. Based on the individualized S-value, the organ-level dose evaluation results are generated using the preset MIRD calculation method and the preset MCs calculation method according to the organ-level source term parameters, and the voxel-level source term parameters and the individualized voxel model are input into the preset fast Monte Carlo model to obtain the voxel-level dose evaluation results. The target dose evaluation result is obtained according to the organ-level dose evaluation result and the voxel-level dose evaluation result.

[0072] Further, in some embodiments, registering SPECT image data and CT image data to obtain organ-level source term parameters and voxel-level source term parameters includes: registering multi-phase CT image data with the first-phase CT image data as a reference to obtain conversion matrices for each phase; based on the conversion matrices for each phase, registering the SPECT image data and the CT image data to obtain the multi-phase count rates of each voxel, and based on a preset count-activity conversion factor, converting the multi-phase count rates of each voxel into activity-time curves; and obtaining the organ-level source term parameters and the voxel-level source term parameters based on the activity-time curves.

[0073] It should be noted that the preset fast Monte Carlo model is as Figure 4 shown. The GPU-accelerated Monte Carlo dose system developed by the embodiments of the present application based on a heterogeneous computing architecture uses the Visual Studio 2022 integrated development environment and the CUDA 10.0 toolchain, and realizes multi-level optimization through C++ / CUDA C mixed programming. According to the characteristics of the hardware hierarchy in the memory, thread hierarchy, and execution model under the CUDA programming model of the GPU graphics card, the program framework, particle transport, step size sampling, data structure, and cross-section access are designed and optimized. On the premise of maintaining the integrity of the physical process, accelerated by an RTX 3090, it is accelerated by more than 50 times compared with traditional MCNP and Geant4 single-core simulations, and the single-case simulation time is compressed from 60 hours to within 0.5 hours.

[0074] The calculation accuracy of the fast Monte Carlo method directly depends on the quality of the input individualized voxel model and dynamic source term distribution. Therefore, a multi-phase SPECT / CT data processing module is constructed: based on the sequence image data in DICOM format, the system first calls a deep learning segmentation network to extract the first-phase CT image data to obtain the contour of the target organ, and generates an individualized voxel model of the patient through piecewise linear conversion of the Hounsfield unit-tissue density.

[0075] In the process of constructing the voxel-level dynamic source term distribution, the system needs to overcome two major technical bottlenecks: First, the spatial mismatch of multi-temporal CT images caused by the patient's body position change and organ deformation during the treatment process; Second, due to the lack of anatomical structure information in SPECT functional images, and as the treatment progresses, the accumulation amount of the drug in the lesion area will gradually decrease, resulting in inaccurate direct registration of cross-temporal functional images. Therefore, for the problem of multi-temporal source term reconstruction, the system adopts a dual registration strategy - taking the first-phase CT as the reference, using the itk-based elastic registration framework, and through the built-in image processing toolbox, aligning the subsequent-phase CTs to the same coordinate system; for SPECT functional images, according to the established CT phase conversion matrix, the functional-anatomical image registration of the corresponding phase is realized, and finally the multi-temporal count rate of each voxel is obtained. After completing the spatio-temporal registration, based on the count-activity conversion factor, the voxel count rate is converted into an activity-time curve.

[0076] In radioactive pharmacokinetic modeling, for the scenario of uniform activity distribution within an organ, the count rates of all voxels within a specific volume of interest (VOI) are accumulated to generate an organ-level time-activity curve. Subsequently, the mono / double exponential decay model is fitted using the non-linear least squares method. Finally, the time integral activity coefficient is obtained through numerical integration, and the organ-level source term parameters are obtained. This process is simple to calculate and takes a short time.

[0077] However, the reconstruction of the voxel-level non-uniform distribution source term faces a stepwise increase in computational complexity. Fitting and integrating each voxel (the voxel dimension is usually on the order of 10 6 pieces) requires a large amount of computing power, storage space, and time. Therefore, the embodiment of this application designs an analytical calculation and trapezoidal integration technique, and the calculation formula is as follows: ; where is the cumulative activity, is the number of time points, is the i th moment is the activity at and

[0078] is the decay coefficient.

[0079] Based on the coupled data of the individualized voxel model and dynamic source term mapping, after the user sets the key parameters on the interactive interface, the system performs GPU-accelerated Monte Carlo simulation. After the simulation is completed, the system outputs a three-dimensional dose distribution matrix with the same resolution as SPECT and automatically generates an organ dose report. The dose volume histogram (DVH) generated based on the voxel-level non-uniform source term can quantify the spatial heterogeneous distribution of the dose within the organ, breaking through the statistical limitations of the traditional uniform distribution model; combined with the GPU-accelerated Monte Carlo algorithm, the overall calculation efficiency is increased by hundreds of times, providing a key basis for individualized radiopharmaceutical dose planning.

[0080] In addition, the embodiment of the present application can also establish a multi-modal tightly coupled system architecture as shown in Figure 5 Targeted radiopharmaceutical dose evaluation, as an interdisciplinary subject integrating biomolecular imaging, radiation dosimetry, medical imaging, and computational physics, has significant professional and technical thresholds. Driven by the improvement of the production capacity of medical radionuclides and the localization process of high-end imaging equipment, nuclear medicine is undergoing a paradigm shift from empirical and extensive diagnosis and treatment to precise quantitative assessment. This transformation poses stringent requirements for nuclear medicine physicians to integrate multidisciplinary knowledge. The multi-platform switching operations (involving multiple independent modules such as multi-modal image fusion, biokinetic model construction, and dose calculation) of the traditional dose evaluation system result in a single evaluation taking up to several hours, severely restricting the efficiency of clinical decision-making. The multi-modal tightly coupled intelligent system constructed by the embodiment of the present application realizes the full-process automation upgrade of dose evaluation through three core technology breakthroughs. This system significantly reduces the intensity of manual interaction, compresses the individualized dose evaluation time to within 1 hour, and provides real-time support for the dynamic optimization of clinical treatment plans.

[0081] In the first stage of processing, the multi-modal tightly coupled system uses deep learning methods to achieve automatic registration of multi-modal images and anatomical structure segmentation. This algorithm supports dual-hardware modes of CPU and GPU and can achieve high-speed operation in a GPU environment with a video memory capacity > 8GB. At present, the segmentation of 117 organs and tissue structures takes no more than 4 minutes. The system is equipped with an intelligent interactive interface, integrating three-dimensional visualization VOI cropping and correction tools for images to assist clinicians in visually verifying and manually correcting the tumor target area and the automatic segmentation results.

[0082] For multi-phase nuclear medicine image data, the system automatically calls the image processing module in Core Technology 3 to achieve precise registration of multi-phase nuclear medicine images, converts the registered data into a three-dimensional activity distribution map, and simultaneously constructs a patient-specific voxel model, laying a personalized data foundation for subsequent dose calculation.

[0083] The system automatically obtains the cumulative activity data of the source organ / target organ and the region of interest (VOI) through the built-in exponential fitting and integral calculation module, and synchronously outputs the source term parameters in dual modes of organ level (uniform distribution) and voxel level (non-uniform distribution). This module realizes the full-automatic generation of source term information, completely avoiding the manual parameter conversion link in the traditional process.

[0084] The system integrates three different dose evaluation schemes: the locally improved MIRD standard method, the direct MCs using the CRAM model, and the Monte Carlo high-precision algorithm accelerated by GPU parallel. Clinicians can not only quickly compare the doses between the MIRD method and the OLINDA system, but also carry out accurate dose evaluation at the sub-organ level relying on the patient-specific voxel model, achieving a double improvement in clinical efficiency and calculation accuracy.

[0085] In the embodiments of this application, an automated dose evaluation system is established by deeply integrating core technologies such as image processing, activity modeling, and dose calculation. Compared with foreign similar platforms, the system greatly reduces the human-computer interaction requirements in the source term information generation and dose calculation links, and all calculation modules have been optimized with local algorithms, providing reliable technical support for precise nuclear medicine diagnosis and treatment.

[0086] Thus, the embodiments of this application solve the problem of multi-modal data flow breakage, integrate modules such as intelligent image segmentation, multi-modal registration, cumulative activity evaluation, dose calculation, and patient information management, construct an automated tightly coupled workflow, and achieve seamless cross-platform data transmission through a standardized protocol, finally forming an integrated closed-loop system covering "data input-dose modeling-clinical decision-making", providing efficient and reliable technical support for individualized precise treatment.

[0087] The rapid dose assessment method for individualized targeted radionuclide therapy proposed according to the embodiments of the present application obtains the labeled radionuclide, drug ligand, and imaging data of the patient, and obtains the corresponding standard S value from the preset population-specific SAF database according to the labeled radionuclide and drug ligand. The first-phase CT imaging data is segmented to obtain the segmentation mask of the target organ. Based on the segmentation mask, the individualized S value is calculated according to the standard S value and the target correction factor, and then the individualized voxel model of the patient is generated according to the segmentation mask. The SPECT imaging data and CT imaging data are registered to obtain the organ-level source term parameters and voxel-level source term parameters. Based on the individualized S value, the organ-level dose assessment result is generated according to the preset MIRD calculation method and the preset MCs calculation method based on the organ-level source term parameters, and the voxel-level source term parameters and the individualized voxel model are input into the preset fast Monte Carlo model to obtain the voxel-level dose assessment result. The target dose assessment result is obtained according to the organ-level dose assessment result and the voxel-level dose assessment result. Thus, the problems of lack of population adaptability, imbalance between calculation accuracy and efficiency, and discretization of the system architecture in dose assessment are solved, the systematic error caused by ethnic differences can be reduced, the calculation time can be shortened while ensuring the calculation accuracy, and the deployment cost can be reduced.

[0088] Next, a rapid dose assessment device for individualized targeted radionuclide therapy proposed according to the embodiments of the present application will be described with reference to the accompanying drawings.

[0089] Figure 6 It is a block diagram of a rapid dose assessment device for individualized targeted radionuclide therapy according to an embodiment of the present application.

[0090] As Figure 6 shown, the rapid dose assessment device 10 for individualized targeted radionuclide therapy includes: a first acquisition module 100, a second acquisition module 200, a segmentation module 300, and an evaluation module 400.

[0091] Among them, the first acquisition module 100 is used to acquire the labeled radionuclide, drug ligand, and imaging data of the patient, where the imaging data includes SPECT imaging data and CT imaging data.

[0092] The second acquisition module 200 is used to obtain the corresponding standard S value from the preset population-specific SAF database according to the labeled radionuclide and drug ligand, where the preset population-specific SAF database is constructed based on a reference human model in a preset region.

[0093] The segmentation module 300 is used to segment the first-phase CT imaging data to obtain the segmentation mask of the target organ, calculate the individualized S value according to the standard S value and the target correction factor based on the segmentation mask, and generate the individualized voxel model of the patient according to the segmentation mask.

[0094] An evaluation module 400 is configured to register SPECT image data and CT image data to obtain organ-level source term parameters and voxel-level source term parameters. Based on the individualized S value, an organ-level dose evaluation result is generated according to the organ-level source term parameters by using a preset MIRD calculation method and a preset MCs calculation method, and the voxel-level source term parameters and the individualized voxel model are input into a preset fast Monte Carlo model to obtain a voxel-level dose evaluation result. A target dose evaluation result is obtained according to the organ-level dose evaluation result and the voxel-level dose evaluation result.

[0095] Optionally, in some embodiments, the second acquisition module 200 includes: a judgment unit and an output unit.

[0096] The judgment unit is configured to judge whether a standard S value matching the labeled radionuclide and the drug ligand exists in a preset SAF database according to the labeled radionuclide and the drug ligand.

[0097] The output unit is configured to output the standard S value when a standard S value exists in the preset SAF database and the version of the standard S value meets a preset version verification condition.

[0098] Optionally, in some embodiments, after judging whether a standard S value matching the labeled radionuclide and the drug ligand exists in the preset SAF database, the judgment unit includes: an update subunit.

[0099] The update subunit is configured to obtain a particle energy spectrum and a branching ratio from a preset radionuclide decay database according to the labeled radionuclide and the drug ligand when the standard S value does not exist in the preset SAF database, calculate a new S value according to the particle energy spectrum and the branching ratio, and add the new S value to a preset population-specific SAF database.

[0100] Optionally, in some embodiments, the segmentation module 300 includes: a determination unit and a calculation unit.

[0101] The determination unit is configured to calculate the actual mass of the target organ according to a segmentation mask, and determine a target correction factor according to the actual mass and a preset standard mass.

[0102] The calculation unit is configured to calculate an individualized S value according to the target correction factor and the standard S value.

[0103] Optionally, in some embodiments, the evaluation module 400 includes: a first registration unit, a second registration unit, and a generation unit.

[0104] The first registration unit is configured to register multi-phase CT image data based on the first-phase CT image data to obtain conversion matrices for each phase.

[0105] A second registration unit, configured to register SPECT image data and CT image data based on each phase conversion matrix to obtain the multi-phase count rates of each voxel, and convert the multi-phase count rates of each voxel into activity-time curves based on a preset count-activity conversion factor.

[0106] A generation unit, configured to obtain organ-level source term parameters and voxel-level source term parameters based on the activity-time curves.

[0107] It should be noted that the foregoing explanation of the embodiments of the method for rapid dose assessment of individualized targeted radionuclide therapy also applies to the device for rapid dose assessment of individualized targeted radionuclide therapy in this embodiment, and will not be elaborated here.

[0108] According to the device for rapid dose assessment of individualized targeted radionuclide therapy provided by the embodiments of the present application, by acquiring the labeled radionuclide, drug ligand and image data of a patient, and obtaining the corresponding standard S value from a preset population-specific SAF database according to the labeled radionuclide and drug ligand, segmenting the first-phase CT image data to obtain a segmentation mask of the target organ, calculating the individualized S value based on the segmentation mask, the standard S value and the target correction factor, generating an individualized voxel model of the patient according to the segmentation mask, registering the SPECT image data and the CT image data to obtain organ-level source term parameters and voxel-level source term parameters, based on the individualized S value, using a preset MIRD calculation method and a preset MCs calculation method to generate an organ-level dose assessment result according to the organ-level source term parameters, and inputting the voxel-level source term parameters and the individualized voxel model into a preset fast Monte Carlo model to obtain a voxel-level dose assessment result, and obtaining a target dose assessment result according to the organ-level dose assessment result and the voxel-level dose assessment result. Thereby, the problems of lack of population adaptability, imbalance between calculation accuracy and efficiency, and discretization of the system architecture in dose assessment are solved, the systematic error caused by ethnic differences can be reduced, the calculation time can be shortened while ensuring the calculation accuracy, and the deployment cost can be reduced.

[0109] Figure 7 The structural schematic diagram of the electronic device provided by the embodiments of the present application. The electronic device may include: A memory 701, a processor 702, and a computer program stored on the memory 701 and executable on the processor 702.

[0110] When the processor 702 executes the program, it implements the method for rapid dose assessment of individualized targeted radionuclide therapy provided in the foregoing embodiments.

[0111] Further, the electronic device further includes: A communication interface 703, configured for communication between the memory 701 and the processor 702.

[0112] A memory 701 for storing a computer program that can run on a processor 702.

[0113] The memory 701 may include a high-speed RAM (Random Access Memory) memory, and may also include a non-volatile memory, such as at least one disk memory.

[0114] If the memory 701, the processor 702, and the communication interface 703 are implemented independently, the communication interface 703, the memory 701, and the processor 702 can be interconnected through a bus and communicate with each other. The bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of representation, Figure 7 only a thick line is used to represent it in the figure, but it does not mean that there is only one bus or one type of bus.

[0115] Optionally, in a specific implementation, if the memory 701, the processor 702, and the communication interface 703 are integrated on a chip, the memory 701, the processor 702, and the communication interface 703 can communicate with each other through an internal interface.

[0116] The processor 702 may be a CPU (Central Processing Unit), or an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present application.

[0117] The embodiments of the present application also provide a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the above-mentioned method for rapid evaluation of the dose of individualized targeted radionuclide therapy is implemented.

[0118] In the description of this specification, the descriptions referring to terms such as "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of this application. In this specification, the schematic expressions of the above terms are not necessarily directed to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or N embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0119] In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" can explicitly or implicitly include at least one of such features. In the description of this application, the meaning of "N" is at least two, such as two, three, etc., unless otherwise specifically defined.

[0120] Any process or method description shown in a flowchart or described in other ways herein can be understood to represent a module, segment, or part of code including one or more N executable instructions for implementing a customized logical function or process, and the scope of the preferred embodiments of this application includes additional implementations, where the functions can be executed in a substantially simultaneous manner or in a reverse order according to the involved functions, rather than in the order shown or discussed, which should be understood by those skilled in the technical field to which the embodiments of this application belong.

[0121] It should be understood that each part of this application can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, the N steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following technologies well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays, field programmable gate arrays, etc.

[0122] Those of ordinary skill in the art of this technology can understand that all or part of the steps carried by the method for implementing the above embodiments can be completed by instructing relevant hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0123] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present application. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present application.

Claims

1. A method for rapid dose assessment of individualized targeted radionuclide therapy, characterized in that, Including the following steps: Obtain the labeled radionuclide, drug ligand, and image data of the patient, where the image data includes SPECT image data and CT image data; Obtain the corresponding standard S value from a preset population-specific SAF database according to the labeled radionuclide and drug ligand, where the preset population-specific SAF database is constructed based on a reference human model in a preset region; Segment the first-phase CT image data to obtain a segmentation mask of the target organ. Based on the segmentation mask, calculate the individualized S value according to the standard S value and the target correction factor, and generate an individualized voxel model of the patient according to the segmentation mask; Register the SPECT image data and the CT image data to obtain organ-level source term parameters and voxel-level source term parameters. Based on the individualized S value, use a preset MIRD calculation method and a preset MCs calculation method to generate an organ-level dose evaluation result according to the organ-level source term parameters, and input the voxel-level source term parameters and the individualized voxel model into a preset fast Monte Carlo model to obtain a voxel-level dose evaluation result. Obtain the target dose evaluation result according to the organ-level dose evaluation result and the voxel-level dose evaluation result.

2. The method according to claim 1, wherein The obtaining the corresponding standard S value from a preset population-specific SAF database according to the labeled radionuclide and drug ligand includes: Judge whether there is a standard S value in the preset SAF database that matches the labeled radionuclide and the drug ligand according to the labeled radionuclide and the drug ligand; If the standard S value exists in the preset SAF database and the version of the standard S value meets the preset version verification condition, output the standard S value.

3. The method according to claim 2, wherein After judging whether there is a standard S value in the preset SAF database that matches the labeled radionuclide and the drug ligand, it includes: If the standard S value does not exist in the preset SAF database, obtain the particle energy spectrum and branching ratio from a preset radionuclide decay database according to the labeled radionuclide and the drug ligand, calculate a new S value according to the particle energy spectrum and the branching ratio, and add the new S value to the preset population-specific SAF database.

4. The method according to claim 1, characterized in that, The calculating the individualized S value according to the standard S value and the target correction factor based on the segmentation mask includes: Calculate the actual mass of the target organ according to the segmentation mask, and determine the target correction factor according to the actual mass and a preset standard mass; Calculate the individualized S value according to the target correction factor and the standard S value.

5. The method according to claim 1, wherein The registering the SPECT image data and the CT image data to obtain organ-level source term parameters and voxel-level source term parameters includes: Register the multi-phase CT image data with the first-phase CT image data as the reference to obtain phase conversion matrices; Based on the phase conversion matrices, register the SPECT image data with the CT image data to obtain the multi-phase count rates of each voxel, and based on a preset count-activity conversion factor, convert the multi-phase count rates of each voxel into activity-time curves; The organ-level source term parameters and the voxel-level source term parameters are obtained based on the activity-time curve.

6. An apparatus for rapid dose assessment of individualized targeted radionuclide therapy, characterized in that, It includes: A first acquisition module for acquiring the labeled radionuclide, drug ligand, and image data of the patient, where the image data includes SPECT image data and CT image data; A second acquisition module for obtaining the corresponding standard S value from a preset population-specific SAF database according to the labeled radionuclide and drug ligand, where the preset population-specific SAF database is constructed based on a reference human model in a preset region; A segmentation module for segmenting the first-phase CT image data to obtain a segmentation mask of the target organ, based on the segmentation mask, calculating an individualized S value according to the standard S value and the target correction factor, and generating an individualized voxel model of the patient according to the segmentation mask; An evaluation module for registering the SPECT image data and the CT image data to obtain organ-level source term parameters and voxel-level source term parameters, based on the individualized S value, using a preset MIRD calculation method and a preset MCs calculation method to generate an organ-level dose evaluation result according to the organ-level source term parameters, and inputting the voxel-level source term parameters and the individualized voxel model into a preset fast Monte Carlo model to obtain a voxel-level dose evaluation result, and obtaining a target dose evaluation result according to the organ-level dose evaluation result and the voxel-level dose evaluation result.

7. The device according to claim 6, characterized in that, The second acquisition module includes: A judgment unit for judging whether there is a standard S value in the preset SAF database that matches the labeled radionuclide and the drug ligand according to the labeled radionuclide and the drug ligand; An output unit for outputting the standard S value when the standard S value exists in the preset SAF database and the version of the standard S value meets the preset version verification condition.

8. The device according to claim 7, characterized in that, After judging whether there is a standard S value in the preset SAF database that matches the labeled radionuclide and the drug ligand, the judgment unit includes: An update subunit for, when the standard S value does not exist in the preset SAF database, obtaining the particle energy spectrum and branching ratio from a preset radionuclide decay database according to the labeled radionuclide and the drug ligand, calculating a new S value according to the particle energy spectrum and the branching ratio, and adding the new S value to the preset population-specific SAF database.

9. An electronic device, characterized in that, It includes: A memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor executes the program to implement the method for rapid dose evaluation of individualized targeted radionuclide therapy according to any one of claims 1-5.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to be used for implementing the method for rapid dose evaluation of individualized targeted radionuclide therapy according to any one of claims 1-5.

Images