Medical system for diagnosing cognitive disease pathology and / or outcome

Abstract

To provide a medical system and method for applying machine learning techniques to aid prediction of disease pathology and clinical outcomes in subjects presenting with symptoms of cognitive decline and to expedite clinical development of novel therapeutics.SOLUTION: A medical system for predicting a disease pathology or disease status in a subject having an uncertain cognitive status comprises: a computer system 100 arranged to receive first medical data for the subject and configured as a learning machine trained on second medical data, the computer system being adapted to use the learning machine to provide a prediction of the disease pathology; and a display 190 for displaying an indication of the prediction. The first medical data does not comprise data obtained from a molecular imaging procedure on the subject, and the second medical data does not comprise data obtained from one or more molecular imaging procedures. The disease pathology is amyloid beta (Aβ) positivity in the subject's brain.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The subject matter of the present disclosure relates to a system for determining the progression of a disease in a subject. More specifically the disclosed subject matter relates to the application of machine learning techniques for digital health tools to assist in making clinical judgments that facilitate the clinical deployment of novel therapies.

Background Art

[0002] Digital health technology and its implementation in the clinical setting are matters of increasing interest. These technologies are becoming more powerful and sophisticated. In particular, machine learning is quite promising as a support for clinicians and other healthcare professionals who are striving to provide better treatment and care for patients.

[0003] Both the increasingly available diagnostic tests and the need to distinguish between abnormalities with nonspecific or overlapping symptoms allow for the accumulation of large amounts of data during the care of individual patients. Some of these data may be redundant or it may be very important for the patient's outcome. If the tests performed were simply what was necessary for them to enable appropriate judgment towards the desired outcome, then overall, it would be an advantage for the present patent and the healthcare system.

[0004] Digital technology can also be used in research environments that are particularly engaged in difficult research problems, for example, to more efficiently market novel disease modifying drugs (DMDs). Regulatory and reimbursement requirements are becoming ever more stringent, and food and drug Regulatory agencies such as the Food and Drug Administration (FDA) implement programs to trial and promote the development of drugs such as the Fast Track Program. However, the challenges faced by pharmaceutical companies during the clinical development of potential candidates are significant and represent a substantial financial draw on their resources. Bringing a therapy to market requires costly, large-scale, and time-consuming clinical trials. In addition to the complexity, there is a high attrition rate caused by the recruitment of poor-fit subjects

[0005] throughout the process. The financial risks associated with taking drug candidates through clinical testing remain high. To support the pipeline by enabling drug candidates to move more efficiently through the pipeline, increase their chances of success, and allow other candidates to enter clinical testing it is necessary to reduce these risks. When using dementia, particularly Alzheimer's disease (AD), as an example there are currently no disease-modifying drugs (DMDs) available. To date, many potential drugs have been tested, but none have shown any significant efficacy. This is partly due to the fact that the trials have focused on subjects in the mid-stage rather than the early stage of the disease. To correct this, current research is aiming to recruit subjects with mild cognitive impairment (MCI) and identify those who are likely to convert to AD during the course of a trial where the drug is most likely to show an effect. To recruit subjects at this stage, more sensitive screening and stratification tools are required.

[0006] In other words, bringing a DMD to the Alzheimer's disease market presents challenges related to the disease's heterogeneity and large scale clinical trial requirements. To address this, current research is aiming to recruit subjects with mild cognitive impairment (MCI) and identify those who are likely to convert to AD during the course of a trial where the drug is most likely to show an effect. To recruit subjects at this stage, more sensitive screening and stratification tools are required. At this stage, more sensitive screening and stratification tools are required to recruit subjects. In other words, bringing a DMD to the Alzheimer's disease market presents challenges related to the disease's heterogeneity and large scale

[0007] clinical trial requirements. It is extremely complex due to the need for long-term clinical trials on a large scale. The purine principal confounder of DMD efficacy is the appropriate selection of subjects. More sensitive screening and stratification tools are required to identify subjects suitable for a given cohort and endpoint.

[0008] Improvements in patient selection for AD trials are possible by using amyloid positron emission tomography ( PET) imaging. For example, Vizamyl (trademark) (flutemetamol F-18 injection, GE Healthcare) is a radiopharmaceutical indicated for brain (PET) imaging to evaluate β-amyloid (Aβ) neuritic plaque density in adult patients with cognitive impairment being evaluated for AD or other causes of cognitive decline. See the GE Healthcare prescribing information for Vizamyl (trademark) (flutemetamol F-18 injection). Other known agents include Neuraceq (trademark) (florbetaben F18 injection, Piramal Imaging), and Amyvid (trademark) (florbetapir, Eli Lilly and Company). A negative Aβ scan indicates that neuritic plaques are rarely absent and does not match the neuropathological diagnosis of AD at the time of image acquisition. The results of a negative scan reduce the likelihood that the patient's cognitive impairment is due to AD. A positive Aβ scan ( ​Positive Aβ scan) indicates that the frequent neuritic plaques are moderate. Neurophysiological examinations have shown that this amount of plaques is present in patients with AD, but there can also be normal cognitive elderly individuals as well as patients with other types of neurological conditions. .

[0009] Aβ PET imaging can be considered as an adjunct to other diagnostic evaluations. A positive Aβ scan does not necessarily establish the diagnosis of AD or other cognitive abnormalities by itself. Amyloid PET positive subjects may be eligible to enter clinical trials for amyloid-modulating therapies, but this test is insufficient as not all amyloid-positive subjects will progress to AD or do so within the time frame relevant to the clinical trial. Taking AD as an example, this means identifying subjects who are likely to have a positive amyloid scan.

[0010] As another example, the diagnosis of Parkinson's disease (PD) remains a challenge for patients with abnormal symptoms or those who show a lack of response to medication. PD is part of a group of diseases with common features named Parkinsonian Syndrome (PS), which includes Progressive Supranuclear Palsy (PNP) and Multiple System Atrophy (MSA). Dopamine transporter (DAT) imaging can be used to attempt to obtain an accurate diagnosis by determining the loss of dopaminergic function. Some Radiolabeled phenyltropane analogues are known for visualizing dopamine transporters and are approved for use in single photon emission tomography (SPECT) imaging. The a-I-labelled agent of the DaTscan (trademark) (GE Healthcare) product is an example. and are approved for use in single photon emission tomography (SPECT) imaging. The a -I-labelled agent of the DaTscan (trademark) (GE Healthcare) product is an example. 123 -I-labelled agent) is an example. 123 -I-labelled agent) is an example.

[0011] Furthermore, in vivo imaging techniques that utilize radiopharmaceuticals such as PET and SPECT are relatively expensive and resource-intensive diagnostic methods. It is advantageous to be able to identify subjects with a high likelihood of having an outcome that implies a pathological condition before performing such in vivo imaging techniques. Thus, for example, for PD, it is advantageous to identify subjects more likely to have dopamine hypoactivity before performing molecular imaging. For AD, it is advantageous to identify subjects more likely to have amyloid plaques before performing molecular imaging. and are relatively expensive and resource-intensive diagnostic methods. It is advantageous to be able to identify subjects with a high likelihood of having an outcome that implies a pathological condition before performing such in vivo imaging techniques. Thus, for example, for PD, it is advantageous to identify subjects more likely to have dopamine hypoactivity before performing molecular imaging. For AD, it is advantageous to identify subjects more likely to have amyloid plaques before performing molecular imaging. able to identify subjects with a high likelihood of having an outcome that implies a pathological condition before performing such in vivo imaging techniques. Thus, for example, for PD, it is advantageous to identify subjects more likely to have dopamine hypoactivity before performing molecular imaging. For AD, it is advantageous to identify subjects more likely to have amyloid plaques before performing molecular imaging. able to identify subjects with a high likelihood of having an outcome that implies a pathological condition before performing such in vivo imaging techniques. Thus, for example, for PD, it is advantageous to identify subjects more likely to have dopamine hypoactivity before performing molecular imaging. For AD, it is advantageous to identify subjects more likely to have amyloid plaques before performing molecular imaging. able to identify subjects more likely to have dopamine hypoactivity before performing molecular imaging. For AD, it is advantageous to identify subjects more likely to have amyloid plaques before performing molecular imaging. able to identify subjects more likely to have dopamine hypoactivity before performing molecular imaging. For AD, it is advantageous to identify subjects more likely to have amyloid plaques before performing molecular imaging. able to identify subjects more likely to have amyloid plaques before performing molecular imaging.

[0012] Providing a prediction model based on the optimal combination of clinical biomarkers and imaging biomarkers has a strong potential to improve the selection process. Common predictive biomarkers collected in clinical trials include hippocampal volume obtained from MR imaging for predicting brain atrophy and PET imaging for assessing amyloid pathology. These biomarkers work better than using only typical inclusion criteria. Providing a prediction model based on the optimal combination of clinical biomarkers and imaging biomarkers has a strong potential to improve the selection process. Common predictive biomarkers collected in clinical trials include hippocampal volume obtained from MR imaging for predicting brain atrophy and PET imaging for assessing amyloid pathology. These biomarkers work better than using only typical inclusion criteria. Providing a prediction model based on the optimal combination of clinical biomarkers and imaging biomarkers has a strong potential to improve the selection process. Common predictive biomarkers collected in clinical trials include hippocampal volume obtained from MR imaging for predicting brain atrophy and PET imaging for assessing amyloid pathology. These biomarkers work better than using only typical inclusion criteria. Providing a prediction model based on the optimal combination of clinical biomarkers and imaging biomarkers has a strong potential to improve the selection process. Common predictive biomarkers collected in clinical trials include hippocampal volume obtained from MR imaging for predicting brain atrophy and PET imaging for assessing amyloid pathology. These biomarkers work better than using only typical inclusion criteria. Providing a prediction model based on the optimal combination of clinical biomarkers and imaging biomarkers has a strong potential to improve the selection process. Common predictive biomarkers collected in clinical trials include hippocampal volume obtained from MR imaging for predicting brain atrophy and PET imaging for assessing amyloid pathology. These biomarkers work better than using only typical inclusion criteria. But it doesn't fully capture the complexity of the disease.

[0013] Recent studies have demonstrated the ability to incorporate automation into quantification of brain volume and beta amyloid SUVr. The authors demonstrate the power of these tools, together with the ability to use them as a substitute for human visual interpretation of images. For example, Thurfjell L, et al., Automated quantification of 18F-Flutemetamol PET Activity for Categorizing Scans as Negative or Positive for Brain Amyloid: C oncordance with Visual Image Reads. Journal of Nuclear Medicine, 55, 1623-1628 ( Please refer to the following link (2014).

[0014] Attempts to correlate clinical data with AD have been attempted. For example, the Medical Imaging Symptoms U.S. Patent No. 9,688 to Ithapu et al., entitled "Clinical Trials for Providing Stem Disease Prognosis" No. 7,199 discloses an artificial intelligence system for analyzing clinical data. The system: It utilizes multiple ranks of machine learning modules, each of which analyzes high and low dimensional data. However, the '19 The systems disclosed in the 9 patents test subjects who are at risk of developing AD within a defined time frame. Lacking any means of identifying the body. Summary of the Invention [Problem to be solved by the invention]

[0015] Thus, innovations are needed that help address shortfalls in current methodologies. 。

Means for Solving the Problem

[0016] The following gives a brief overview of one or more embodiments to provide a basic understanding of the presently disclosed subject matter. This overview is not an extensive overview of all expected embodiments, does not identify important or essential elements of all embodiments, nor does it depict the scope of any or all embodiments. Its sole purpose is to present in a simplified form some of the concepts of one or more embodiments as an introduction to the more detailed description that will be presented later.

[0017] According to one aspect, a medical system for predicting the disease pathology or disease state of a subject having an uncertain cognitive state, comprising a computer system arranged to receive first medical data about the subject and at least partially configured as a trained learning machine trained with respect to second medical data, the trained learning machine being used to perform a prediction of disease pathology, and a display for displaying an indication of the prediction, wherein the first medical data does not include data obtained from a molecular imaging technique for the subject, and the second medical data does not include data obtained from one or more molecular imaging techniques. The medical system is disclosed. The disease pathology may be amyloid-beta (Aβ) positive in the subject's brain. The first medical data may include the results of a cognitive examination of the subject, the results of a cognitive examination of the subject, the age of the subject, the educational level of the subject, or some combination of these data.

[0018] ​​​According to another aspect, a medical system for predicting the clinical outcome of a subject having an uncertain cognitive state, arranged to receive first medical data about the subject, a first trained learning machine trained with respect to second medical data, and at least partially comprising a second learning machine trained with respect to third medical data, the first trained learning machine and the second learning machine being used to perform a disease pathology prediction, a computer system, and a display for displaying an indication of the prediction, the first medical data being able to include data obtained from a molecular imaging technique for the subject, and the first trained learning machine being trained at least in part with data obtained from one or more molecular imaging techniques, is disclosed.

[0019] According to another aspect, a method for predicting the disease pathology of a subject having an uncertain cognitive state, the method comprising the steps of collecting cohort medical data for a first set of subjects having known outcomes for the disease in an electronic memory, receiving subject medical data for the subject, using a computer system having a trained learning machine trained with respect to the cohort medical data to perform a prediction of the disease pathology based at least in part on the subject medical data, and using a display to indicate the prediction.

[0020] ​​​​​​​​​​According to another aspect, there is provided a method for predicting the clinical outcome of a subject having an uncertain cognitive state, comprising collecting first cohort medical data in an electronic memory for a first set of subjects having an outcome known for a disease, collecting second cohort medical data in an electronic memory for a second set of subjects having an outcome known for a disease, collecting second cohort medical data in an electronic memory for a second set of subjects having an outcome known for a disease, a computer system having a first trained learning machine arranged to receive subject medical data for a subject and trained with respect to the first cohort medical data, and a second trained learning machine trained with respect to the second cohort medical data, using the first and second trained learning machines to perform a prediction of disease pathology based at least in part on the subject medical data, and displaying an indication of the prediction. The first cohort medical data may include a data type different from the data type of the second cohort medical data. The first cohort medical data can include a data type at least partially the same as the data type of the second cohort medical data.

[0021] The medical system can be used for predicting the clinical outcome of a subject having an uncertain cognitive state and / or for predicting the disease pathology of a subject having an uncertain cognitive state. The ability to predict a specific disease pathology with molecular imaging as a pre-screening tool enables the enrichment of the cohort of subjects undergoing molecular imaging techniques. Health economics and the subject's exposure to radiation ​​​​​​​​​​There are also practical benefits in terms of the reduction of the explosion. Subjects selected by prescreening By performing molecular imaging on the subject, it is possible to stratify according to the rate of disease progression .

[0022] According to another aspect, a medical system for identifying a subject at risk of developing Alzheimer's disease (AD), comprising configured to receive medical data of one or more subjects with unknown outcomes for AD, a first trained learning machine for giving a first instruction and a second trained learning machine for giving a second instruction and at least partially composed of, and configured to combine the first and second instructions to identify which subjects with unknown outcomes for AD are at risk of developing AD within a defined time frame , a computer system, and a display for displaying the identification. The first instruction can include the probability that the subject is Aβ positive. The second instruction can include the probability that the subject's mild cognitive impairment is rapidly progressive. The first trained learning machine and the second trained learning machine may be implemented on the same physical hardware. The medical system can comprise an electronic memory configured to store medical data and supply the medical data to the computer system. The medical data can include in vivo image data , cognitive function memory data, and genetic data. The in vivo image data may include standardized uptake value ratio (SUVR) or T1-weighted magnetic resonance (MR:T1-weighted magnetic res onance) volume measurement. The medical data can include demographic data and / or data for each subject . The first instruction can include the probability that the subject is Aβ positive. The second instruction can include the probability that the subject's mild cognitive impairment is rapidly progressive. The first trained learning machine and the second trained learning machine may be implemented on the same physical hardware. The medical system can comprise an electronic memory configured to store medical data and supply the medical data to the computer system. The medical data can include in vivo image data , cognitive function memory data, and genetic data. The in vivo image data may include standardized uptake value ratio (SUVR) or T1-weighted magnetic resonance (MR:T1-weighted magnetic res onance) volume measurement. The medical data can include demographic data and / or data for each subject . The medical data can include in vivo image data, cognitive function memory data, and genetic data. The in vivo image data may include standardized uptake value ratio (SUVR) or T1-weighted magnetic resonance (MR:T1-weighted magnetic res onance) volume measurement. The medical data can include demographic data and / or data for each subject . The in vivo image data may include standardized uptake value ratio (SUVR) or T1-weighted magnetic resonance (MR:T1-weighted magnetic res onance) volume measurement. The medical data can include demographic data and / or data for each subject It may include the subject's electronic medical record. The subject may be a patient with mild cognitive impairment, and the defined time frame may be 3 years or less or 2 years or less. The first instruction may be the probability that the subject is Aβ positive and the second instruction may be the probability that the subject's mild cognitive impairment is rapidly progressive. The computer system may be configured to identify subjects at risk of developing AD within the defined time frame using the first probability and the second probability.

[0023] According to another aspect, a method of identifying a subject at risk of developing Alzheimer's disease (AD), comprising collecting medical data in a storage electronic memory for a first set of subjects having known outcomes for the disease, identifying subjects at risk of developing AD within a defined time frame using a first trained learning machine and a second trained learning machine, and selecting subjects at risk of developing AD within a defined time frame for AD research. The medical data may include in vivo image data, cognitive function memory data, and genetic offspring data. The in vivo image data may include standardized uptake value ratio (SUVR) or T1-weighted magnetic resonance (MR) volumetric measurements. The medical data may include demographic data and / or the subject's electronic medical record. The subject may be a patient with mild cognitive impairment, and the defined time frame may be 3 years or less or several years or less.

[0024] According to another aspect, a method of determining the prognosis of a patient with Alzheimer's disease (AD), comprising administering an amyloid protein contrast agent to a patient in need thereof. ​​​​​​​​​ scanning, and imaging amyloid protein deposits within a patient, and correlating the imaging of amyloid deposits with a training set for the patient and other variables using a prognosis known for AD are disclosed. A method is disclosed that includes

[0025] According to another aspect, a method of treating a patient with Alzheimer's disease (AD) includes, in a computer, comparing imaging data and other data obtained from the patient with a training set that includes the imaging data and other data regarding the patient and a prognosis known for AD to classify the patient into a disease cohort, identifying a drug therapy known to improve the outcome of patients with AD within the disease cohort, and treating the patient with an effective amount of the identified drug therapy. A method is disclosed that includes comparing, in a computer, imaging data and other data obtained from the patient with a training set that includes the imaging data and other data regarding the patient and a prognosis known for AD to classify the patient into a disease cohort, identifying a drug therapy known to improve the outcome of patients with AD within the disease cohort, and treating the patient with an effective amount of the identified drug therapy. A method is disclosed that includes comparing, in a computer, imaging data and other data obtained from the patient with a training set that includes the imaging data and other data regarding the patient and a prognosis known for AD to classify the patient into a disease cohort, identifying a drug therapy known to improve the outcome of patients with AD within the disease cohort, and treating the patient with an effective amount of the identified drug therapy. A method is disclosed that includes comparing, in a computer, imaging data and other data obtained from the patient with a training set that includes the imaging data and other data regarding the patient and a prognosis known for AD to classify the patient into a disease cohort, identifying a drug therapy known to improve the outcome of patients with AD within the disease cohort, and treating the patient with an effective amount of the identified drug therapy.

[0026] According to another aspect, using medical data for one or more subjects having an unknown outcome for Alzheimer's disease (AD) to determine which subjects are at risk of developing AD, the computer system being at least partially configured as a first trained learning machine that gives a first instruction and a second trained learning machine that gives a second instruction, the combination of the first and second instructions being configured to specify which subjects having an unknown outcome for AD are at risk of developing AD within a defined time frame, the method including supplying the medical data to the computer system. A use is disclosed that includes using medical data for one or more subjects having an unknown outcome for Alzheimer's disease (AD) to determine which subjects are at risk of developing AD, the computer system being at least partially configured as a first trained learning machine that gives a first instruction and a second trained learning machine that gives a second instruction, the combination of the first and second instructions being configured to specify which subjects having an unknown outcome for AD are at risk of developing AD within a defined time frame, the method including supplying the medical data to the computer system. A use is disclosed that includes using medical data for one or more subjects having an unknown outcome for Alzheimer's disease (AD) to determine which subjects are at risk of developing AD, the computer system being at least partially configured as a first trained learning machine that gives a first instruction and a second trained learning machine that gives a second instruction, the combination of the first and second instructions being configured to specify which subjects having an unknown outcome for AD are at risk of developing AD within a defined time frame, the method including supplying the medical data to the computer system. A use is disclosed that includes using medical data for one or more subjects having an unknown outcome for Alzheimer's disease (AD) to determine which subjects are at risk of developing AD, the computer system being at least partially configured as a first trained learning machine that gives a first instruction and a second trained learning machine that gives a second instruction, the combination of the first and second instructions being configured to specify which subjects having an unknown outcome for AD are at risk of developing AD within a defined time frame, the method including supplying the medical data to the computer system. A use is disclosed that includes

[0027] According to another aspect, a medical system for classifying a subject as having mild cognitive impairment (MCI) or Alzheimer's disease (AD) is arranged to receive medical data about one or more subjects having an unknown classification for MCI or AD and is at least partially configured as a first trained learning machine that gives a first instruction and a second trained learning machine that gives a second instruction, and is configured to combine the first and second instructions to identify which subjects have MCI and which subjects have A D, a computer system, and a display for presenting the classification. The computer system may be configured to give a further classification to subjects classified as MCI as to whether the MCI is early MCI or late M CI.

[0028] According to another aspect, a medical system for classifying a subject as having Alzheimer's disease (AD) or some other form of dementia is arranged to receive medical data about one or more subjects having an unknown classification for MCI or AD and is at least partially configured as a first trained learning machine that gives a first instruction and a second trained learning machine that gives a second instruction, and is configured to combine the first instruction and the second instruction to identify which subjects have AD and which subjects have another form of dementia and a computer system and a display for presenting the classification.

[0029] ​​​​​​According to another aspect, a medical system for identifying additional instructions for a drug, configured to receive medical data for one or more subjects taking the drug, at least partially comprising a first trained learning machine that gives a first instruction and a second trained learning machine that gives a second instruction, and configured to determine whether the first and second instructions can be combined and assigned to an instruction in which the drug is added to the existing instructions, a computer system, and a display for displaying the additional instructions are disclosed.

[0030] Further features and advantages of the present invention, as well as the structure and operation of various embodiments of the present invention, will be described in more detail below with reference to the accompanying drawings. It should be noted that the present invention is not limited to the specific embodiments described herein. Such embodiments are merely shown herein for explanation. Further embodiments will be apparent to those skilled in the art based on the

[0031] teaching contained herein. The accompanying drawings, which are incorporated herein and form a part of the specification, illustrate the subject matter and, together with the description in words, further serve to explain the principles of the subject matter and enable those skilled in the art to make and use the

[0032]

Figure 1

Figure 2

Figure 3

Figure 4

[0033] Next, various embodiments will be described with reference to the accompanying drawings. Like reference numerals are used throughout to refer to like elements. In the following description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of one or more embodiments. However, it will be apparent in some or all instances that any of the embodiments described below may be practiced without adopting the specific design details described below. In other instances, well-known structures and devices are shown in block diagram form in order to facilitate the description of one or more embodiments. The following presents a simplified summary of one or more embodiments in order to provide a basic understanding of the embodiments. This summary is not an extensive overview of all contemplated embodiments, nor does it identify key or essential elements of all embodiments, nor delineate the scope of any or all embodiments. references to an "embodiment," "one embodiment," "an example embodiment," etc., in the specification indicate that the embodiment described may include a particular feature, structure, or characteristic, but not every embodiment necessarily includes the particular feature, structure, or characteristic. Moreover, such phrases are not necessarily referring to the same embodiment. Moreover, such phrases are not necessarily referring to the same embodiment. However, it will be apparent in some or all instances that any of the embodiments described below may be practiced without adopting the specific design details described below. In other instances, well-known structures and devices are shown in block diagram form in order to facilitate the description of one or more embodiments. The following presents a simplified summary of one or more embodiments in order to provide a basic understanding of the embodiments. This summary is not an extensive overview of all contemplated embodiments, nor does it identify key or essential elements of all embodiments, nor delineate the scope of any or all embodiments. The following presents a simplified summary of one or more embodiments in order to provide a basic understanding of the embodiments. This summary is not an extensive overview of all contemplated embodiments, nor does it identify key or essential elements of all embodiments, nor delineate the scope of any or all embodiments. This summary is not an extensive overview of all contemplated embodiments, nor does it identify key or essential elements of all embodiments, nor delineate the scope of any or all embodiments. This summary is not an extensive overview of all contemplated embodiments, nor does it identify key or essential elements of all embodiments, nor delineate the scope of any or all embodiments. nor delineate the scope of any or all embodiments.

[0034] References to an "embodiment," "one embodiment," "an example embodiment," etc., in the specification indicate that the embodiment described may include a particular feature, structure, or characteristic, but not every embodiment necessarily includes the particular feature, structure, or characteristic. Moreover, such phrases are not necessarily referring to the same embodiment. Moreover, such phrases are not necessarily referring to the same embodiment. Moreover, such phrases are not necessarily referring to the same embodiment. Yes. Further, when a particular feature, structure, or characteristic is described in connection with one embodiment, it should be understood that, whether explicitly described or not, it is within the knowledge of those skilled in the art to effect such feature, structure, or characteristic in connection with other embodiments.

[0035] Embodiments of the present invention can be implemented in hardware, firmware, software, or any combination thereof. Embodiments of the present invention can also be implemented as instructions stored on a machine-readable medium that can be read and executed by one or more processors. The machine-readable medium can include any mechanism for storing or transmitting information in a form readable by a machine (e.g., a computing device). For example, the machine-readable medium can include read-only memory (ROM), random access memory (RAM), magnetic disk storage media, optical storage media, flash memory devices, electrical, optical, acoustic, or other forms of non-transitory machine-readable media. Further, firmware, software, routines, instructions can be described herein as performing certain operations. However, such descriptions are for convenience only and in reality, such operations can arise from computing devices, processors, controllers, or other devices executing firmware, software, routines, instructions, etc. It should be understood. To more clearly and concisely describe and point out the subject matter of the claimed invention, the following definitions are provided for certain terms used throughout this specification and the claims. However, such descriptions are for convenience only and in reality, such operations can arise from computing devices, processors, controllers, or other devices executing firmware, software, routines, instructions, etc. It should be understood that such operations result from other devices that execute firmware, software, routines, instructions, etc., such as computing devices, processors, controllers, or firmware, software, routines, instructions. Please understand.

[0036] To more clearly and concisely describe and point out the subject matter of the claimed invention, the following definitions are given for specific terms used throughout this specification and the claims. Any instances of a particular term herein should be considered as a non-limiting example.

[0037] The term "uncertain cognitive status" refers to symptoms suggestive of a disease or condition associated with cognitive decline. Symptoms include confusion, poor motor coordination, and loss of sense of self. Confusion, impaired judgment, subjective memory loss, inability to concentrate, and slurred speech. Known diseases and conditions associated with cognitive decline Non-limiting examples of conditions include AD, PD, MCI, TBI (traumatic brain injury), jury), and Chronic Traumatic Encephalopathy (CTE). can be.

[0038] The term "disease pathology" refers to pathology associated with a disease or condition typically associated with cognitive decline. The term "pathology" is used herein to refer to the biological characteristics of the disease pathology contemplated by the present invention. Non-limiting examples include amyloid beta positive (i.e., abnormal presence of amyloid), dopamine Deficiency of phosphodiesterase activity and the formation of neurofibrillary tangles (NFTs) ) is included.

[0039] The term "subject" is used herein to refer to any human or animal subject. In one aspect of the embodiment, the subject of the invention is a mammal. In another aspect of the embodiment, the subject is an intact mammalian body in vivo. The body is human. The subject is a Subjective Memory Complainer (SMC). who may be a Complainer or may suffer from mild cognitive impairment (MCI), and the possibility of developing Alzheimer's disease (AD) is being investigated.

[0040] A "trained learning machine" is, for example, a computing system trained using machine learning with respect to a set of training data in order to have the ability to give predictions, conclusions, and classifications when given new data. The computing system learns on its own without being explicitly programmed with the relationship between the data and the predictions, conclusions, and classifications.

[0041] A "display" can mean any device capable of displaying information in the form of alphanumeric or graphic characters and typically includes a screen, circuitry, casing, and power source. Non-limiting examples of displays include computer monitors, tablet screens, and smartphone screens. The display may be local, i.e., directly connected to a local computer serving as the host of the system, or it may be remote, i.e., part of a user system where a computer serving as the host of the system communicates over a network, or it may be provided as a cloud-based utility, for example, by an Internet having system services.

[0042] As used herein, the term "prediction instruction" is intended to refer to information displayed on a display in the form of alphanumeric or graphic characters. Non-limiting examples of instructions include graphs, tables, and bar graphs.

[0043] ​​​​​​​​​​​​​As used herein, the term "molecular imaging techniques" refers to techniques that are used to study cellular function or or in vivo imaging that allows visualization of molecular processes (i.e., generating molecular images). "In vivo imaging technique" refers to a technique for imaging the internal aspects of a subject. A technique for non-invasively generating images of all or part of a molecule contemplated by the present invention. Non-limiting examples of imaging techniques include amyloid beta imaging, dopamine transduction These include tau imaging, reporter imaging, and tau imaging.

[0044] The term "amyloid beta (Aβ) positive" refers to, for example, When observed on the obtained Aβ molecular images, frequent amyloid neuritic plaques were observed. que) is moderate. Prediction of the possibility of amyloid positivity was based on molecular imaging data. It is possible to implement aspects of the present invention using data that does not include such data. Non-limiting examples of data include activities of daily living (ADL) baseline, digit span variance, and digit span backward, logical memory 30 min after story baseline I I (logical memory II 30min after story baseline), Trail Making Part A trail making part A-time (seconds), education, male gender, left caudate volume The right amygdala volume, right amygdala volume, and right caudate volume. In an embodiment, such data includes age, sex, Mini Mental Score (MMSE), clinical Dementia Scale (CDR), Clinical Dementia Severity Scale (CDR-SB), ApoE gene Inspection status, local left and right brain volumes of the hippocampus, parahippocampus, amygdala, olfactory endothelium Quality, including the medial temporal lobe, gyrus rectus, ventricle, and angular gyrus.

[0045] The term "cognitive test" refers to tests performed on a subject to assist in determining the cognitive state. Typical tests are well known to those skilled in the art. Non-limiting examples of tests include various forms of IQ tests, memory, attention, and drawing concentration. Non-limiting examples of tests typically performed on non-human animal subjects include the mirror test and the T-maze test. The term "cognitive test" refers to tests performed on a subject to assist in determining the cognitive state. Typical tests are well known to those skilled in the art. Non-limiting examples of tests include various forms of IQ tests, memory, attention, and drawing concentration. Non-limiting examples of tests typically performed on non-human animal subjects include the mirror test and the T-maze test. The term "cognitive test" refers to tests performed on a subject to assist in determining the cognitive state. Typical tests are well known to those skilled in the art. Non-limiting examples of tests include various forms of IQ tests, memory, attention, and drawing concentration. Non-limiting examples of tests typically performed on non-human animal subjects include the mirror test and the T-maze test. The term "cognitive test" refers to tests performed on a subject to assist in determining the cognitive state. Typical tests are well known to those skilled in the art. Non-limiting examples of tests include various forms of IQ tests, memory, attention, and drawing concentration. Non-limiting examples of tests typically performed on non-human animal subjects include the mirror test and the T-maze test. The term "cognitive test" refers to tests performed on a subject to assist in determining the cognitive state. Typical tests are well known to those skilled in the art. Non-limiting examples of tests include various forms of IQ tests, memory, attention, and drawing concentration. Non-limiting examples of tests typically performed on non-human animal subjects include the mirror test and the T-maze test.

[0046] As used herein, the term "cohort medical data" is taken to mean the collection of data from a defined cohort of subjects (e.g., subjects potentially having a particular disease or condition and / or subjects being considered for inclusion in a clinical study). As used herein, the term "cohort medical data" is taken to mean the collection of data from a defined cohort of subjects (e.g., subjects potentially having a particular disease or condition and / or subjects being considered for inclusion in a clinical study). As used herein, the term "cohort medical data" is taken to mean the collection of data from a defined cohort of subjects (e.g., subjects potentially having a particular disease or condition and / or subjects being considered for inclusion in a clinical study). As used herein, the term "cohort medical data" is taken to mean the collection of data from a defined cohort of subjects (e.g., subjects potentially having a particular disease or condition and / or subjects being considered for inclusion in a clinical study).

[0047] The term "mild cognitive impairment" (MCI) is used to refer to a condition involving problems with cognitive functions such as memory, language, thinking, and judgment, which are often greater than normal age-related changes. The term "mild cognitive impairment" (MCI) is used to refer to a condition involving problems with cognitive functions such as memory, language, thinking, and judgment, which are often greater than normal age-related changes.

[0048] The term "rapidly progressive" refers to the relative progression from the initial symptoms of a disease or condition to the diagnosis of the disease or condition in a subject. Non-limiting examples are that the subject progresses from MCI to dementia within a relatively short time frame, for example, within about 3 years. Some data may be useful for predicting rapid progression. Non-limiting examples of such data are ADL baseline, Mini-Mental State Examination (MMSE) baseline, semantic fluency task - animal category The term "rapidly progressive" refers to the relative progression from the initial symptoms of a disease or condition to the diagnosis of the disease or condition in a subject. Non-limiting examples are that the subject progresses from MCI to dementia within a relatively short time frame, for example, within about 3 years. Some data may be useful for predicting rapid progression. Non-limiting examples of such data are ADL baseline, Mini-Mental State Examination (MMSE) baseline, semantic fluency task - animal category The term "rapidly progressive" refers to the relative progression from the initial symptoms of a disease or condition to the diagnosis of the disease or condition in a subject. Non-limiting examples are that the subject progresses from MCI to dementia within a relatively short time frame, for example, within about 3 years. Some data may be useful for predicting rapid progression. Non-limiting examples of such data are ADL baseline, Mini-Mental State Examination (MMSE) baseline, semantic fluency task - animal category The term "rapidly progressive" refers to the relative progression from the initial symptoms of a disease or condition to the diagnosis of the disease or condition in a subject. Non-limiting examples are that the subject progresses from MCI to dementia within a relatively short time frame, for example, within about 3 years. Some data may be useful for predicting rapid progression. Non-limiting examples of such data are ADL baseline, Mini-Mental State Examination (MMSE) baseline, semantic fluency task - animal category The term "rapidly progressive" refers to the relative progression from the initial symptoms of a disease or condition to the diagnosis of the disease or condition in a subject. Non-limiting examples are that the subject progresses from MCI to dementia within a relatively short time frame, for example, within about 3 years. Some data may be useful for predicting rapid progression. Non-limiting examples of such data are ADL baseline, Mini-Mental State Examination (MMSE) baseline, semantic fluency task - animal category Lee, composite amyloid standard uptake value ratio (SUVR) (pons), and hippocampus including the product.

[0049] The present invention can be used by pharmaceutical companies at various points in the clinical trial workflow to help reduce time and costs in bringing new drugs to market. More specifically it is by enriching the inclusion criteria (screening and stratification tools), or by identifying the most appropriate cohort (market access tool) that is most likely to respond to treatment.

[0050] For example, after an initial diagnostic test is performed, a patient's eligibility for a trial can be considered by assessing their suitability against the inclusion criteria. Typically, this is done through a screening process that may involve some inexpensive diagnostic tests and in some cases more expensive diagnostic tests, such as the use of positron emission tomography (PET) imaging. Phase 3 trials involve hundreds of subjects. If the screening for a trial is inefficient, more subjects will be screened than need to be entered into the trial. Thus, traditional screening methods are a source of economic burden for pharmaceutical companies in terms of the higher costs associated with some of the tests, and for the many subjects who pass through this stage of the process. The present invention uses clinical data collected by less expensive tests to predict the outcome of more expensive tests. In this way, patients are prioritized through the screening process. If necessary, additional functionality of the tool can further stratify subjects based on the rate of disease progression, or the subjects most likely to respond to therapy. In this way, the inclusion criteria are enriched by the prediction of the outcome.

[0051] If the trial is successful, the market access team will ensure that payers settle for the drug and will require data as evidence of drug effectiveness. Here, the present invention helps to identify which patients will benefit from the drug, analyzes existing real-world evidence from multiple sources, and creates the outcome data needed by payers and service managers for adoption.

[0052] Applying the present invention to data from the recruitment phase can reduce the number of patients required to demonstrate drug effectiveness. Further, these techniques can predict the subjects most likely to show an effect of the drug, and thus improve the number and suitability of subjects needed to achieve a statistically relevant effect size. These efficiencies affect the overall cost of the trial by shortening the recruitment phase and reducing the overall cost associated with screening tests. This is just one example of a context in which the teachings of the present disclosure can be advantageously applied.

[0053] A general-purpose computer system 100 is shown in FIG. 1. The system 100 of FIG. 1 includes a processor 110 connected to a bus 120. Also connected to this bus are a memory (e.g., firmware, RAM, or ROM) 130, a hard drive 140, a network interface 150, and an input / output (I / O) interface 160. The network interface 150 is connected to a network 170. The I / O interface 160 can be connected to one or more input / output devices such as an input device 180 and a display 190. The system 100 can include ports, printers, CD / It can have additional components such as a DVD ROM reader / writer, etc.

[0054] A general-purpose computer system 100, such as that shown in FIG. 1, is configured for machine learning, i.e., configured to perform tasks without specific programming for machine learning. In this aspect, a set of training data ("features") is given to the computer system 100, based on which the computer system 100 can learn to make predictions. Subsequently, this computer system can make predictions based on new data.

[0055] One aspect of the disclosed subject matter includes the use of machine learning-based analysis techniques that facilitate the identification of appropriate subjects based on that specific imaging data, genetic data, psychometric data, and demographic data, thereby determining optimal stratification of clinical trial subjects with mild cognitive impairment (MCI). This technique involves the application of sequential or individual machine learning models to exclude patients lacking appropriate biomarkers, progression rates, or both.

[0056] Specifically, the process of selecting subjects for the study is shown in FIG. 2. In step S20, subjects with mild cognitive impairment (MCI) are identified using standard test screening through psychological tests, collection of subject demographics, and acquisition of MRIs. In step S30, a machine learning model is used to predict the probability that the identified subjects are beta-amyloid (Aβ) positive. In step S40, amyloid PET images are obtained for subjects whose probability of being Aβ positive exceeds a certain threshold. In step S 50 will use machine learning models to predict the progression of subjects to AD within the clinical trial timeframe. In step S60, subjects whose probability of transition exceeds a certain threshold are Selected for inclusion in the study.

[0057] Figure 2 shows the results of two trained, i.e., machine learning, models (one predicting Aβ positivity). one to measure the probability of transition and the other to predict the probability of transition), but in succession, i.e., one after the other, to select subjects. These models are trained in this paper and the process used These trained learning machines, also called machines, are implemented on separate hardware. It will be appreciated that the above may be implemented in the same hardware. Each model can be used by itself to customize the selection process as needed. In this way, subjects who would not benefit from the DMD under investigation may be spared unnecessary testing. This dramatically reduces the cost of clinical trials and improves overall trial efficiency. Another possibility is to use models in parallel, with the results of each being used to determine the effect of subject selection. This is shown in FIG. 3, in step S20. Standard testing procedures were performed through psychological testing, collection of subject demographics, and acquisition of MRIs. There is again a step of identifying subjects with mild cognitive impairment using screening. In step S70, Aβ positive markers are collected. In step S80, machine learning Step S90: Using the model to predict the probability that the identified subject is Aβ positive. In step S100, this probability is used as an input to subject selection. In step 110, positive markers for progressing to MCI are collected. Then, using the machine learning model, predict the probability that a specified subject has MCI, which is an ongoing fact. In step S90, this probability is used as another input to subject selection.

[0058] As an example, a model for predicting Aβ status and a model for predicting rapidly progressive MCI subjects defined as subjects who will transition to a probable AD (pAD: probable AD) status within a 36-month time frame were learned with respect to a Phase III clinical trial dataset including 87 Aβ-positive subjects, 81 subjects who transitioned to AD within 36 months after baseline imaging examination, and 232 MCI subjects. The main purpose of this Phase III clinical trial was to compare the pAD transition of MCI subjects with normal and abnormal 18 F] flutemetamol uptake. The long-term follow-up of the subject transition status to pAD makes this a good dataset for the purpose of model learning.

[0059] The goal of the amyloid-positive model is to select subjects who are likely to have an amyloid-positive PET imaging scan before they accept the cost and inconvenience of undergoing a PET scan. Phase III trial subjects lacking an amyloid PET scan were removed from the training set, leaving 227 subjects, and the amyloid positivity rate was 38%. The goal of the rapidly progressive MCI model is to select subjects who are likely to transition to pAD within the scope of a 3-year clinical trial. Within a time period of less than 3 years ​​​​​​​​​​​​​​​Among the Phase III trial subjects lacking a conversion label, or those with a "not-converted" conversion label, 182 subjects were excluded from the training set, resulting in a conversion rate of 45%. (not-converted) and the conversion rate was 45%.

[0060] Model generation was performed using features available from the Phase III study. Imaging data was quantified using automated quantification software, which generated quantified brain volumes from T1 MR images and SUVr values for the pons reference region from amyloid PET images. These quantified regions were then selected below using a feature selection algorithm for use in the model in combination with the collected demographic, neuropsychological, and genetic data. To handle subjects who deviated from the data other than the target level, an interpolation method was utilized. Figure 2 graphically shows the generalized process for using this model. data was quantified using automated quantification software, which generated quantified brain volumes from T1 MR images and SUVr values for the pons reference region from amyloid PET images. These quantified regions were then selected below using a feature selection algorithm for use in the model in combination with the collected demographic, neuropsychological, and genetic data. To handle subjects who deviated from the data other than the target level, an interpolation method was utilized. Figure 2 graphically shows the generalized process for using this model. ence region) and SUVr values for the pons reference region from amyloid PET images. These quantified regions were then selected below using a feature selection algorithm for use in the model in combination with the collected demographic, neuropsychological, and genetic data. To handle subjects who deviated from the data other than the target level, an interpolation method was utilized. Figure 2 graphically shows the generalized process for using this model. and the collected demographic, neuropsychological, and genetic data, and were selected below using a feature selection algorithm for use in the model. To handle subjects who deviated from the data other than the target level, an interpolation method was utilized. Figure 2 graphically shows the generalized process for using this model. data were then selected below using a feature selection algorithm for use in the model in combination with the collected demographic, neuropsychological, and genetic data. To handle subjects who deviated from the data other than the target level, an interpolation method was utilized. Figure 2 graphically shows the generalized process for using this model. An interpolation method was used to handle subjects who deviated from the data other than the target level. Figure 2 graphically shows the generalized process for using this model. Figure 2 graphically shows the generalized process for using this model.

[0061] The Aβ-positive model showed the potential for a 43% improvement in the test efficiency of selecting Aβ-positive subjects, a 79% improvement in model accuracy, and an 85% improvement in specificity when the improvement in test efficiency was measured as the positive predictive value (PPV) of the model compared to the original inclusion of Aβ-positive subjects. when the improvement in test efficiency was measured as the positive predictive value (PPV) of the model compared to the original inclusion of Aβ-positive subjects. and a 79% improvement in model accuracy, and an 85% improvement in specificity. The rapidly progressive MCI model showed an improvement in the selection of rapidly progressive MCI subjects of 24%, an improvement in accuracy of 86%, and an improvement in specificity of 92% when the improvement in test efficiency was measured as the PPV of the model compared to subjects identified as Aβ-positive by amyloid PET imaging alone. when the improvement in test efficiency was measured as the PPV of the model compared to subjects identified as Aβ-positive by amyloid PET imaging alone. and an improvement in accuracy of 86%, and an improvement in specificity of 92%. improvement.

[0062] The report of the performance measurement criteria of the model brought about through tolerance verification represents the results of machine learning is an effective method for, but it is important to know that these can be adapted to independent data sets and are not unique to the learning data set.

[0063] To evaluate this ability, the Aβ-positive model was reconstructed using features common to both the Phase III data set and the data set of the Australian Imaging, Biomarker and Lifestyle Flagship Study of Ageing (AIBL). The AIBL validation for Aβ-positive with n = 551 showed an extrapolation to other populations and disease states by incorporating healthy and AD subjects within the AIBL data set, resulting in an accuracy score of 75% and a specificity of 87%. The use of machine learning for subject screening and stratification in clinical trials can increase the success rate of indicating drug efficacy and the rate at which DMD therapies that produce the desired results become available. The performance of the model shown suggests that the incorporation efficiency can be improved by more than 50%. Furthermore, the validation of the Aβ-positive model using the AIBL data set demonstrates the feasibility of extending a model learned on one population to a different population and the ability to extend the positive risk score learned only on MCI subjects to both healthy and AD subjects. e Flagship Study of Ageing) data set. reconstructed using features common to both the Phase III data set and the data set of the Australian Imaging, Biomarker and Lifestyle Flagship Study of Ageing (AIBL). For Aβ-positive with n = 551, the AIBL validation showed extrapolation to other populations and disease states by incorporating healthy and AD subjects within the AIBL data set, resulting in an accuracy score of 75% and a specificity of 87%.

[0064] The use of machine learning for subject screening and stratification in clinical trials can increase the success rate of indicating drug efficacy and the rate at which DMD therapies that produce the desired results become available. The performance of the model shown suggests that the incorporation efficiency can be improved by more than 50%. Furthermore, the validation of the Aβ-positive model using the AIBL data set demonstrates the feasibility of extending a model learned on one population to a different population and the ability to extend the positive risk score learned only on MCI subjects to both healthy and AD subjects.

[0065] Next, a method for constructing a model according to one aspect of the present invention will be described. This method consists of two main ​It can be considered to include the next phase. The first one prepares the data, that is, "cleans" the data. The second one evaluates the data when it is prepared.

[0066] The features included for consideration can be obtained from any one or more categories such as psychometrics, demographics, genetics, AβPET, and T1 MRI. For psychometrics, the features can include CDR, CDR-SB, ADL, MMSE, semantic fluency task (animals ), semantic fluency task (vegetables), ADAS-cog, digit span backward, digit span forward, digit symbol substitution test, logical memory II (30 minutes after the story ), logical memory II (immediately after the story), trail making part A, and / or can include data from trail making part B. Demographic features can include education and / or age and / or gender. Genetic features can include Apo E. AβPET features can include AβSUVR-PONS in front of the frontal lobe, anterior cingulate AβSUVR-PONS, precuneus post cingulate AβSUVR -PONS, parietal AβSUVR-PONS, temporal lateral AβSUVR -PONS, temporal mesial AβSUVR-PONS, occipital AβSUVR -PONS, sensorimotor AβSUVR-PONS, and / or composite AβSUV R-PONS. T1 MR can include hippocampal volume, thalamic volume, tonsil volume, putamen volume, caudate volume, parahippocampal volume, olfactory cortex volume, medial temporal lobe volume, ventricular volume, gyrus rectus volume, angular gyrus volume, total gray matter, and / or total white matter. Generally, ADL, M ​ MSE, semantic fluency task (animals), composite Aβ SUVR-PONS, and hippocampus Subsets of these features such as product are used.

[0067] Regarding data preparation, it can also be regarded as a group of techniques, for example, feature selection (narrowing from a first set of potential features to a subset of potential features), feature engineering, subject selection, and complementing outlier features. Feature selection and feature en gineering involve selecting a subset of features to be removed in a first instance based on prior knowledge about these features. For example, some features may be part of the test incorporation criteria and thus do not provide much variance for use in model building. Some features are known to be strongly correlated with other features and are therefore removed to prevent highly correlated features from biasing the model. Finally, some features may prove inappropriate for use as features in the model through experimentation.

[0068] The experimental phase of model building can also involve using recursive feature elimination, random forest feature selection, or other similar methods known to those skilled in the art before the model building phase. In this way, less important features may be removed to simplify the model. This technique is useful when building the rapidly progressive MCI model described below. After the experimental phase, the model features can be set in the model without further feature selection. Feature selection is available, for example, in the scikit-learn.org scikit-learn machine It can be performed using tools available from the learning library.

[0069] Further feature engineering can include averaging features from MR scans of different cerebral hemispheres. This approach is also useful when constructing the rapidly progressive MCI model described below. For example, the right hippocampal volume and the left hippocampal volume may be averaged to generate hippocampal volume features, and the left ventricle and the right ventricle may be combined to generate ventricular volume features. When multiple datasets are used, it is useful to use intersections across the feature sets to utilize only the features common to all datasets. For subject selection, in the case of the rapidly progressive MCI model, subjects with a "No" progression status before the 3-year time point were excluded from the model construction phase. For one reason or another, these subjects did not continue throughout the study, and thus their progression status is not reliable. In the case of the amyloid-positive model, subjects were given a negative (amyloid-negative) or positive (amyloid-positive) status based on their composite SUVr relative to a threshold of 0.62, enabling the generation of transition labels. Subjects lacking the transition labels or PET scans whose true status is unknown were removed from the model construction population. Regarding the complementation of outlying features, imputation of the median value may be performed, or subjects may be removed, to take into account subjects lacking values for the features used in the model.

[0070] When multiple datasets are used, it is useful to use intersections across the feature sets to utilize only the features common to all datasets. When multiple datasets are used, it is useful to use intersections across the feature sets to utilize only the features common to all datasets.

[0071] Regarding subject selection, in the case of the rapidly progressive MCI model, subjects with a "No" progression status before the 3-year time point were excluded from the model construction phase. For one reason or another, these subjects did not continue throughout the study, and thus their progression status is not reliable. For one reason or another, these subjects did not continue throughout the study, and thus their progression status is not reliable. In the case of the amyloid-positive model, subjects were given a negative (amyloid-negative) or positive (amyloid-positive) status based on their composite SUVr relative to a threshold of 0.62, enabling the generation of transition labels. In the case of the amyloid-positive model, subjects were given a negative (amyloid-negative) or positive (amyloid-positive) status based on their composite SUVr relative to a threshold of 0.62, enabling the generation of transition labels. In the case of the amyloid-positive model, subjects were given a negative (amyloid-negative) or positive (amyloid-positive) status based on their composite SUVr relative to a threshold of 0.62, enabling the generation of transition labels. Subjects lacking the transition labels or PET scans whose true status is unknown were removed from the model construction population. Subjects lacking the transition labels or PET scans whose true status is unknown were removed from the model construction population.

[0072] Regarding the complementation of outlying features, imputation of the median value may be performed, or subjects may be removed, to take into account subjects lacking values for the features used in the model. Regarding the complementation of outlying features, imputation of the median value may be performed, or subjects may be removed, to take into account subjects lacking values for the features used in the model. Alternatively, subjects may be assigned the most common value for a feature. For example, if ApoE status is a feature, subjects lacking ApoE status may be excluded from the data. Instead of performing an imputation on the intermediate values ​​of Alternatively, if a feature is outlying, the variables can be imputed. The model is applied without specific characteristics. Based on the knowledge of the important characteristics, If characteristics of a subject or a cohort of subjects are not available, the modeling software In some cases, this may be the case if such tests are performed. It is known that these methods can improve the accuracy of predictions (e.g., determining ApoE status). This allows adding the corresponding value if

[0073] Once the data is prepared, it is evaluated to build a model of rapidly progressive MCI. For this, a stratified K-fold (5-fold) can be used. Both amyloid positive and transition labels are If present in the dataset used to construct the pool, the population is amyloid-positive converters. , amyloid positive non-converters, amyloid negative converters, and amyloid negatives. The split can be stratified to have an equal distribution of non-converters. , which more similarly matches the actual population of subjects than stratifying purely on migration status. The stratified k-fold implementation of Scikit-Learn can be used for this purpose. obtain.

[0074] Logistic regression can also be used to build models of rapidly progressive MCI. For example, a psychic learning run on logistic regression can be used to determine optimal hyperparameters. To determine the parameters used in the model construction, we used the grid search CV implementation of ScikitLearn. The model can be reconstructed for each fold to determine the average statistic, and then The final model used in can be reconstructed using all available data.

[0075] Model statistics for constructing the rapidly progressive MCI model can be calculated for each fold. The results are then averaged and presented with 95% confidence intervals for the reported statistics. Statistics collected were accuracy, f1 score, specificity, recall, pr-AUC, NPV, and and accuracy.

[0076] To build the amyloid positive model, a stratified K-fold (5 fold) can again be used. Both amyloid positive and transition labels were present in the dataset used for model building. If present, the population is classified as amyloid positive converters, amyloid positive non-converters, have an equal distribution of amyloid-negative converters and amyloid-negative non-converters In this way, the division can be stratified purely on the basis of amyloid positivity. The actual population of subjects can be stratified to more similarly match than by using a transition label. is missing the dataset used for construction, where the subjects are As in the population, each partition has an equal number of amyloid-positive and amyloid-negative subjects. The psychiatric distribution can be stratified using only amyloid positives to ensure that the A stratified k-fold implementation of Trahn can be used for this purpose.

[0077] Sequential feedforward feature selection is used to build an amyloid-positive model. It is also possible. Feedforward feature selection can be performed within each split during the experimental phase to determine the optimal features for each split. After the optimal features are determined for each split, the features selected in three or more splits can be selected for use in the model building phase. After the experimental phase, feedforward feature selection can be removed from the processing steps, and the features used are not selected by the algorithm each time. The mlxtend (Machine Learning Extensions) Python library can be used for the implementation of feedforward feature selection. It is also possible to use a Gaussian Naive Bayes classifier to build an amyloid-positive model. The circuit run implementation of Gaussian Naive Bayes can be used for this with the default options maintained. The model can be reconstructed for each split to determine the mean statistics, and then the final model used can be reconstructed using all available data. Model statistics can be calculated for each split and then averaged for the reported statistics and presented with a 95% confidence interval. The calculated statistics can include accuracy, f1-score, specificity, recall, pr-AUC, NPV, and precision.

[0078] A machine learning model developed for the identification and stratification of ideal subjects for clinical trials considerably improves the ability of the automated process, thereby assisting clinical trials for the benefit of both the subjects and the institutions conducting the trials. The amyloid-positive model has shown the ability to improve test efficiency by up to 43%, while the rapidly progressive MCI model has a transition efficiency of only 17% over three years. It is also possible. The circuit run implementation of Gaussian Naive Bayes can be used for this with the default options maintained. The model can be reconstructed for each split to determine the mean statistics, and then the final model used can be reconstructed using all available data. Model statistics can be calculated for each split and then averaged for the reported statistics and presented with a 95% confidence interval. The calculated statistics can include accuracy, f1-score, specificity, recall, pr-AUC, NPV, and precision. A machine learning model developed for the identification and stratification of ideal subjects for clinical trials considerably improves the ability of the automated process, thereby assisting clinical trials for the benefit of both the subjects and the institutions conducting the trials. The amyloid-positive model has shown the ability to improve test efficiency by up to 43%, while the rapidly progressive MCI model has a transition efficiency of only 17% over three years.

[0079] Model statistics can be calculated for each split and then averaged for the reported statistics and presented with a 95% confidence interval. The calculated statistics can include accuracy, f1-score, specificity, recall, pr-AUC, NPV, and precision. A machine learning model developed for the identification and stratification of ideal subjects for clinical trials considerably improves the ability of the automated process, thereby assisting clinical trials for the benefit of both the subjects and the institutions conducting the trials. The amyloid-positive model has shown the ability to improve test efficiency by up to 43%, while the rapidly progressive MCI model has a transition efficiency of only 17% over three years.

[0080] A machine learning model developed for the identification and stratification of ideal subjects for clinical trials considerably improves the ability of the automated process, thereby assisting clinical trials for the benefit of both the subjects and the institutions conducting the trials. The amyloid-positive model has shown the ability to improve test efficiency by up to 43%, while the rapidly progressive MCI model has a transition efficiency of only 17% over three years. A machine learning model developed for the identification and stratification of ideal subjects for clinical trials considerably improves the ability of the automated process, thereby assisting clinical trials for the benefit of both the subjects and the institutions conducting the trials. The amyloid-positive model has shown the ability to improve test efficiency by up to 43%, while the rapidly progressive MCI model has a transition efficiency of only 17% over three years. showed the ability to stratify subjects to increase the rate.

[0081] These models enable the ability to be fine-tuned to meet the specific requirements of clinical trials, This model may be reconstructed to determine the priority of various statistical measures such as sensitivity, F1, or accuracy, and the subject risk threshold may be modified to target subjects demonstrating the ideal pathology for a particular test. Further, this model can be used individually or sequentially before expending resources on subjects who do not fit well with the test, thereby demonstrating a clinical trial in a stepwise manner to include subjects in the test. For example, an amyloid-positive model can be used to select subjects to undergo an amyloid PET scan, and then a rapidly progressive MCI model can determine whether a subject is likely to progress at an ideal rate over a certain period of the test before including the subject in a lengthy and expensive protocol. The models have the ability to be combined in this way, but the models can also be used individually without sacrificing the performance of the model in any way.

[0082] More datasets may be included in the model construction phase. Further, model construction utilizing raw MR and PET imaging data can be used in conjunction with or in place of the automated quantification algorithms described herein. The term "raw data" is intended to include direct pixel or voxel data.

[0083] The teachings of the present disclosure can be implemented as a web-based application. This is shown in FIG. 4. A client including a web browser or dedicated application 405​​​​​​​​​​​ The computer 400 communicates with a distributed content information system 420 such as the Internet. The virtual private cloud 410 communicates with the The application program 410 provides an interface to the execution framework 440. The execution framework 440 may include a model store. Using the trained model from 450, we apply the application layer 430 and operates based on data provided through and stored in database 460. , for example, through a web browser 400 to an application layer 430. The information gives the probability that the patient will develop AD.

[0084] The preceding example illustrates the use of a system for selecting subjects for clinical studies based on their prognosis. However, this may be used for other applications. It may be used to predict disease pathology in subjects with uncertain cognitive status, A computer system uses trained learning machines to predict disease pathology, and displaying an indication of a prediction, the indication being a molecular imaging procedure for the subject. The data obtained from the study will not be used. Disease pathology is the formation of amyloid beta ( The first medical data may include a result of a cognitive test of the subject, The results of the study, the age of the subject, the education level of the subject, or some combination of these data It may include

[0085] As another example, a health care system may need to predict clinical outcomes for subjects with uncertain cognitive status. and can be used therefor. The system receives first medical data about a subject and is arranged to be trained with respect to second medical data, a first trained learning machine, and at least partially comprises a computer system configured as a second learning machine trained with respect to third medical data, the computer system being arranged to perform a prediction of disease pathology using the first trained learning machine and the second learning machine. The first medical data can include data obtained from molecular imaging techniques with respect to the subject, and the first trained learning machine is trained at least in part with data obtained from one or more molecular imaging techniques.

[0086] For example, as described above, the diagnosis of Parkinson's disease (PD) in patients presenting with abnormal symptoms or showing lack of response to medication can be difficult. Imaging with DaTscan ( trademark) can be used to obtain an accurate diagnosis by determining the loss of dopaminergic function. Furthermore, in vivo imaging techniques using radiopharmaceuticals such as PET and single photon emission computed tomography (SPECT) are relatively expensive and resource-intensive diagnostic techniques. The use of trained learning as described above using in vivo imaging scan results enhances the ability to identify subjects likely to have an outcome indicating a pathological condition. For PD, it is advantageous to identify

[0087] As another example, the teachings herein can be used to implement a method for predicting the disease pathology of a subject having an uncertain cognitive state, and the cohort medical data is collected and stored in an electronic memory for a first set of subjects having an outcome known for the disease, and the computer system is arranged to receive subject medical data for a subject and comprises a trained learning machine trained with respect to the cohort medical data, wherein the computer system is configured to perform a prediction of the disease pathology based at least in part on the subject medical data using the trained learning machine to provide a prediction of the disease pathology.

[0088] As another example, the teachings herein can be used to implement a method for predicting the clinical outcome of a subject having an uncertain cognitive state, and the first cohort medical data is collected and stored in an electronic memory for a first set of subjects having an outcome known for the disease, and the second cohort medical data is collected and stored in an electronic memory for a second set of subjects having an outcome known for the disease, and the computer system uses the subject medical data for the subject and comprises a first trained learning machine trained with respect to the first cohort medical data and a second trained learning machine trained with respect to the second cohort medical data. The computer system uses the first and second trained learning machines to perform a prediction of the disease pathology based at least in part on the subject medical data and displays an indication of the prediction. The first cohort medical data has a data type different from the data type of the second cohort medical data. may also include. The first cohort medical data can include data types that are at least partially the same as the data types of the second cohort medical data.

[0089] As another example, the teachings herein can be used to predict the clinical outcomes of subjects having an uncertain cognitive state and / or to predict the disease pathology of subjects having an uncertain cognitive state.

[0090] The teachings of the present disclosure can also be applied to determine the prognosis of patients with AD. An amyloid protein contrast agent can be administered to a patient, and then the amyloid protein deposition in the patient can be imaged to obtain image data. The image data can be correlated with data from the training set of the patient and other data using a prognosis known for AD.

[0091] The teachings of the present disclosure can also be applied to treat patients with AD. A computer can be used to compare imaging data and other data obtained from the patient with a training set including imaging data and other data regarding the patient with respect to a prognosis known for AD to classify the patient into a disease cohort. Using the results of the comparison, known drug therapies can be identified, thereby improving the outcome of patients with AD within the disease cohort, and then the patient can be treated with an effective amount of the identified drug therapy.

[0092] The teachings of the present disclosure can be used with the medical data of one or more subjects having an unknown outcome for AD to determine which subjects are at risk of developing AD. ​​​​​​​​​​​​​​ can also be used. The medical data is supplied to a computer system that is at least partially configured as a first trained learning machine that gives a first instruction and a second trained learning machine that gives a second instruction. The instructions are combined to identify which subjects with an unknown outcome for AD are at risk of developing AD within a defined time frame.

[0093] The teachings of the present disclosure can also be applied to classify subjects as having mild cognitive impairment (MCI) or Alzheimer's disease (AD). The computer receives medical data for one or more subjects having an unknown classification for MCI or AD, is arranged to be at least partially configured as a first trained learning machine that gives a first instruction and a second trained learning machine that gives a second instruction, and is combined with the first and second instructions to identify which subjects have MCI and which subjects have AD. The computer system may be further configured to give a further classification to subjects classified as MCI as to which MCI is early MCI or late MCI.

[0094] The teachings of the present disclosure are for a medical system for classifying subjects having Alzheimer's disease (AD) or some other form of dementia, wherein a computer system is arranged to receive medical data for one or more subjects having an unknown classification for MCI or AD, and is at least partially configured as a first trained learning machine that gives a first instruction and a second trained learning machine that gives a second instruction. At least partially configured as a second trained learning machine that gives instructions Together, it is configured to identify which subjects have AD by combining the first and second instructions, and Which subjects have some other form of dementia, and is configured to It can also be applied to medical systems.

[0095] The teachings of the present disclosure are medical systems for identifying additional instructions for drugs, The computer system is configured to receive medical data for one or more subjects taking the drug A first trained learning machine that gives a first instruction and At least partially configured as a second trained learning machine that gives a second instruction And combining the first and second instructions to determine whether the drug can be assigned to an instruction to be added to an existing instruction It can also be applied to medical systems configured as such. Can be.

[0096] This specification discloses the invention, including the best mode, using examples, and enables those skilled in the art to make and use any device Or system, and to practice the invention, including practicing any incorporated method. The patentable scope of the present invention is defined by the claims And may include other examples that occur to those skilled in the art. Such Other examples are intended to be within the scope of the claims if they have structural elements that do not differ from the literal terms of the claims Or if they include equivalent structural elements that differ slightly from the literal terms of the claims Elements. All patents and patent applications mentioned in the text are incorporated by reference in their entirety as if individually incorporated Included. All patents and patent applications are incorporated by reference in their entirety as if individually incorporated It is incorporated into this specification.

Explanation of Signs

[0097] 100 General-purpose computer system, computer system, system 110 Processor 120 Bus 130 Memory 140 Hard drive 150 Network interface 160 Input / output (I / O) interface, I / O interface 170 Network 180 Input device 190 Display 400 Client computer, web browser 405 Web browser or dedicated application 410 Virtual private cloud 420 Distributed content information system 430 Application layer 440 Execution framework 450 Model storage unit 460 Database

Claims

1. A medical system for predicting disease pathology in a subject with an uncertain cognitive status, 、 configured to receive first medical data about the subject and a second At least in part as a trained learning machine that is trained on medical data and using the trained learning machine to predict the disease pathology. A computer system configured to: a display for displaying an indication of the prediction; The first medical data is data obtained from a molecular imaging technique for the subject. and said second medical data does not include said second medical data, said second medical data being derived from one or more molecular imaging techniques. The health care system does not include data that can be used to

2. The method according to claim 1, wherein the disease pathology is associated with a disease or condition associated with cognitive decline. The medical system.

3. 2. The method of claim 1, wherein the disease pathology is a loss of dopamine producing brain cells. system.

4. The disease pathology is amyloid beta (Aβ) positivity in the brain of the subject.

2. The medical system described in 1.

5. The first medical data includes an activities of daily living (ADL) baseline, a digit span, Backward, 30 minutes of logical memory II after story baseline, trail making Part A time (seconds), education, male gender, left caudate volume, right amygdala volume, and right caudate volume. The medical system of claim 4 .

6. The first medical data includes age, sex, Mini Mental Score (MMSE), clinical dementia, Scale (CDR), Clinical Dementia Severity Scale (CDR-SB), ApoE Gene Testing Scale Status, regional left and right brain volumes of hippocampus, parahippocampal gyrus, amygdala, entorhinal cortex, medial temporal lobe, rectus gyrus, ventricle and the angular gyrus.

7. The medical system of claim 1 , wherein the clinical outcome is rapid progression of MCI.

8. The first medical data includes ADL baseline, Mini-Mental State Examination (MMSE), ) Baseline, Semantic Fluency Task - Animal Category, Composite Amyloid Standards 8. The medical system of claim 7, comprising: Tem.

9. The medical device of claim 1 , wherein the first medical data comprises results of a cognitive test of the subject. system.

10. The medical system of claim 1 , wherein the first medical data includes an age of the subject.

11. 10. The medical system of claim 1, wherein the first medical data includes the subject's years of education. Tem.

12. The medical system of claim 1 , wherein the first medical data includes a result of an ApoE genetic test. Stem.

13. Claims for use in a method for predicting disease pathology in a subject with an uncertain cognitive status. Item 1. A medical system according to item 1.

14. Claims for use in a method for predicting clinical outcomes in subjects with uncertain cognitive status.

2. The medical system described in 1.

15. The subject is a subjective memory comprehender (SMC) or has mild cognitive impairment (MCI). The patient is suffering from mild cognitive impairment (MCI) and is being investigated for the possibility of developing Alzheimer's disease (AD).

2. The medical system according to claim 1.

16. The medical system of claim 1 , wherein the subject is cognitively normal.

17. A medical system for predicting a clinical outcome of a subject with an uncertain cognitive status, comprising: configured to receive first medical data about the subject; and A first trained learning machine is trained on a third medical data and a second learning machine configured at least in part as a second learning machine trained on the data. , determining disease pathology using the first trained learning machine and the second learning machine. a computer system configured to perform a prediction of a genomics; a display for displaying an indication of the prediction; The first medical data is data obtained from a molecular imaging technique for the subject. the first trained learning machine includes one or more molecular images A medical system that is trained at least in part on data obtained from a Hmm.

18. The subject is a subjective memory comprehender (SMC) or has mild cognitive impairment (MCI). The patient is suffering from mild cognitive impairment (MCI) and is being investigated for the possibility of developing Alzheimer's disease (AD). The medical system of claim 17.

19. 18. The method of claim 17, wherein the prediction is that the subject will develop Alzheimer's disease. health care system.

20. Claims for use in a method for predicting disease pathology in a subject with an uncertain cognitive status. Item 18. The medical system described in item 17.

21. Claims for use in a method for predicting clinical outcomes in subjects with uncertain cognitive status.

18. The medical system described in 17.

22. 1. A method for predicting disease pathology in a subject with uncertain cognitive status, comprising: A cohort medical data set is generated for a first set of subjects with a known outcome for the disease. collecting the data in an electronic memory; a cohort medical data processor configured to receive subject medical data about the subject; a trained learning machine that is trained on data, and analyzing the subject's medical history using the trained learning machine to make a prediction of pathology. and making a prediction of said disease pathology based at least in part on the data. using a computer system; and indicating said prediction using a display.

23. 1. A method for predicting a clinical outcome in a subject with an uncertain cognitive status, comprising: A first cohort for a first set of subjects with a known outcome for the disease. collecting medical data in an electronic memory; A second cohort of subjects with a known outcome for the disease. collecting the patient's medical data in an electronic memory; a first cohort of subjects, the first cohort being configured to receive subject medical data about the subject; A first trained learning machine that is trained on the previous medical data; A second trained learning system that is trained on the second cohort medical data. and detecting the first and second trained learning machines to determine whether the first and second trained learning machines are capable of detecting the first and second trained learning machines. and making the prediction of the clinical outcome based at least in part on the treatment data. using a computer system; and indicating said prediction using a display.

24. The first cohort medical data is a different data type than the second cohort medical data.

24. The method of claim 23, comprising different data types.

25. The first cohort medical data has a data type that is slightly different from that of the second cohort medical data.

24. The method of claim 23, wherein the plurality of data elements include at least partially the same data types.

26. A medical system for identifying subjects at risk for developing Alzheimer's disease (AD) It is Receive medical data for one or more subjects with an unknown outcome for AD. a first trained learning machine arranged to trust a first instruction and a second and a second trained learning machine configured at least in part as a second trained learning machine that provides instructions to the second and combining the first and second indications with an unknown outcome for AD. To identify which subjects are at risk of developing AD within a defined time frame. A computer system configured as follows: and a display for displaying the identification.

27. 27. The medical system of claim 26, wherein the first indication comprises a probability that the subject is Aβ positive. Tem.

28. 27. The method of claim 26, wherein the second indication comprises a probability that the subject has a progressive form of mild cognitive impairment. The medical system described in.

29. The first trained learning machine and the second trained learning machine 27. The medical system of claim 26, wherein: are implemented on the same physical hardware.

30. storing said medical data and providing said medical data to said computer system; 27. The medical system of claim 26, further comprising an electronic memory configured to:

31. The medical data includes in vivo imaging data, cognitive function memory data, and genetic data.

27. The medical system of claim 26, further comprising:

32. The in vivo image data includes amyloid beta positive status, normal uptake ratio, (SUVR), or T1-weighted magnetic resonance (MR) volumetric measurements. Treatment system.

33. The medical data includes demographic data and / or electronic medical records of each subject. The medical system of claim 26.

34. The subject is a patient with mild cognitive impairment and the defined time frame is 3 years or less.

27. The medical system of claim 26.

35. 35. The medical system of claim 34, wherein the defined time frame is two years or less.

36. 2. The subject is a healthy control, a subject with mild cognitive impairment, and a subject with Alzheimer's disease.

7. The medical system described in 6.

37. The first indication is a probability that the subject is Aβ positive and the second indication is a probability of the subject being Aβ positive. a probability that the mild cognitive impairment is progressive, and the computer system and a second probability is used to identify subjects at risk of developing AD within a defined time frame. The medical system of claim 26 configured to identify.

38. 1. A method for identifying a subject at risk for developing Alzheimer's disease (AD), comprising: Record medical data for a first set of subjects with a known outcome for the disease. collecting the information in a storage type electronic memory; Using a first trained learning machine and a second trained learning machine identifying a subject at risk of developing said AD within a time frame determined by the method; Subjects at risk of developing AD within a time frame defined for a study on said AD. selecting a body.

39. The medical data includes in vivo imaging data, cognitive function memory data, and genetic data.

40. The method of claim 38, further comprising:

40. The in vivo image data includes amyloid positive status, normalized uptake ratio (SU 40. The method of claim 39, comprising measuring the volume of the ventricle by CT, CT scan, or T1-weighted magnetic resonance (MR) volumetry.

41. The medical data includes demographic data and / or electronic medical records of each subject. The method of claim 38.

42. The subject is a patient with mild cognitive impairment and the defined time frame is 3 years or less.

39. The method of claim 38.

43. 43. The method of claim 42, wherein the defined time frame is two years or less.

44. 3. The subject is a healthy control, a subject with mild cognitive impairment, and a subject with Alzheimer's disease.

8. The method according to claim 8.

45. 1. A method for determining the prognosis of a patient with Alzheimer's disease (AD), comprising: The step of administering an amyloid protein contrast agent to a patient in need of it. P and imaging amyloid protein deposits in the brain of said patient; Imaging the amyloid deposits in the patient is compared with known prognoses for AD. A training set of patient data and a trained machine learning model and correlating with other variables using.

46. 1. A method of treating a patient suffering from Alzheimer's disease (AD), comprising: To classify the patients into disease cohorts, a computer is The data and other data and imaging data obtained from the patient and known AD. The data are compared to a training set that contains other data about patients with known outcomes. and Medications known to improve patient outcomes for AD within the disease cohort identifying a law; and treating the patient with the identified medication.

47. To determine which of the subjects are at risk for developing AD, The present invention relates to medical data for one or more subjects with an unknown outcome for Alzheimer's disease (AD). a first trained learning machine that provides a first instruction; and a second trained learning machine configured at least in part as a second instruction providing means for providing the second instruction. and combining the first and second indications to have an unknown outcome for AD. To identify which subjects are at risk of developing AD within a defined time frame. providing said medical data to a computer system configured to ,use.

48. Subjects are classified as having mild cognitive impairment (MCI) or Alzheimer's disease (AD). A medical system for classifying A medical examination of one or more subjects with an unknown classification for MCI or AD. a first trained learning device configured to receive the medical data and provide a first instruction; and a second trained learning machine that provides the second instruction. and determining whether the subject has MCI by combining the first and second instructions. and which subjects have AD, A computer system; and a display for displaying the classification.

49. The computer system determines whether the MCI is in the early MCI state for a subject classified as having MCI. and providing a further classification of CI or late MCI.

49. The medical system of claim 48.

50. To classify a subject as having Alzheimer's disease (AD) or some other form of dementia A health care system for: A medical examination of one or more subjects with an unknown classification for MCI or AD. a first trained learning device configured to receive the medical data and provide a first instruction; and a second trained learning machine that provides the second instruction. The first instruction and the second instruction are combined to determine which subject has AD. and which subjects have another form of dementia. A computer system comprising: and a display for displaying the identification.

51. 1. A medical system for identifying additional instructions for a medication, comprising: configured to receive medical data about one or more subjects taking the drug a first trained learning machine that provides a first instruction and a second trained learning machine that provides a second instruction At least in part configured as a second trained learning machine, Combining the first and second instructions and assigning them to an instruction in which the drug is added to an existing instruction. a computer system configured to determine whether and a display for displaying the additional instructions.

Images