System, control method, and computer program

Abstract

To provide a system for determining a state of dementia by using an image discrimination technique in cooperation with a cognitive ability test score.SOLUTION: A system (1) has a first input module (10) configured to obtain a first evaluation index based on data related to the physical state of a subject's brain, a second input module (20) configured to obtain a second evaluation index based on data related to the function of the subject's brain, and an estimation module (30) configured to estimate the state of dementia of the subject on the basis of an evaluation value obtained by a first evaluation function having the first evaluation index and the second evaluation index as variables.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a system capable of evaluating dementia.

Background Art

[0002] Patent Document 1 describes providing a driving suitability diagnosis device and a driving suitability diagnosis method that are less affected by factors such as the examination environment, the physical and mental state of the subject, and the arbitrariness of the examiner, and that can diagnose the driving suitability of the subject with high reliability. This driving suitability diagnosis device includes white matter lesion examination means for examining the degree of white matter lesions in the subject, and driving suitability judgment means for judging the driving suitability of the subject based on the examination results of the white matter lesion examination means. The driving suitability judgment means is characterized in that when the degree of white matter lesions examined by the white matter lesion examination means is equal to or greater than a specified value, the driving suitability of the subject is judged to be unsuitable.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] Matsuda, Hiroshi et al.'s "Differentiation Between Dementia With Lewy Bodies And Alzheimer’s Disease Using Voxel-Based Morphometry Of Structural MRI: A Multicenter Study." (Neuropsychiatric Disease and Treatment 15 (2019): 2715.) discloses a technique for differentiating Alzheimer's disease and Lewy body dementia using brain images.

[0005] It is disclosed in "Cognitive loss in dementia with Lewy bodies and Alzheimer disease." by Shimomura, Tatsuo et al. (Archives of Neurology 55.12 (1998): 1547-1552.) that there is a bias in the distribution of specific cognitive test scores (WAIS-R Block Design test scores or ADAS delayed recall scores) in Alzheimer's disease and Lewy body dementia.

[0006] However, a technique for determining the state of dementia by collaborating an image discrimination technique and a cognitive test score has not been disclosed.

Means for Solving the Problems

[0007] One aspect of the present invention is a system having a first input module configured to obtain a first evaluation index based on data related to the physical state of a subject's brain, a second input module configured to obtain a second evaluation index based on data related to the function of the subject's brain, and an estimation module configured to estimate the state of dementia and / or other brain disorders of the subject based on an evaluation value obtained by a first evaluation function using the first evaluation index and the second evaluation index as variables. It becomes possible to fold the evaluation result of the data related to the physical state of the brain and the evaluation result of the data related to the function of the brain into one collaborative evaluation result using the evaluation function, and the state of dementia can be evaluated with higher accuracy. Therefore, a system for determining or assisting in determining the affected state including the presence or absence of brain diseases of the subject can be provided. Further, when the subject is included in a group that has ingested pharmaceuticals (including investigational drugs and unapproved drugs), food and beverages, supplements, etc., the estimation module may include a function (unit) for evaluating the effect on those dementias. Further, the estimation module may include a function for estimating the affected state of the first causative disease.

[0008] One of the other aspects of the present invention is a method for controlling a system. The system includes a first input module configured to obtain a first evaluation index based on data related to the physical state of the subject's brain, a second input module configured to obtain a second evaluation index based on data related to the function of the subject's brain, and an estimation module configured to estimate the state of a brain disorder including dementia of the subject. The method includes the following steps. 1. The estimation module obtains the first evaluation index and the second evaluation index via the first input module and the second input module. 2. Based on the evaluation value obtained by a first evaluation function having the first evaluation index and the second evaluation index as variables, estimate the state of dementia of the subject. The estimated state of dementia may include at least any of the information required for the prevention and treatment of dementia, such as the presence or absence of dementia, stage, causative disease, etc.

[0009] One of the further different aspects of the present invention is a program. The program (program product) has instructions for a computer to execute obtaining a first evaluation index based on data related to the physical state of the subject's brain, obtaining a second evaluation index based on data related to the function of the subject's brain, and estimating the state of dementia of the subject based on the evaluation value obtained by a first evaluation function having the first evaluation index and the second evaluation index as variables. The program may be recorded on a computer-readable recording medium and provided.

Brief Description of the Drawings

[0010]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Figure 17

Figure 18

Figure 19

Figure 20

Figure 21

Figure 22

Figure 23

Figure 24

Figure 25

Figure 26

Figure 27A

Figure 27B

Figure 28

Figure 29

Figure 30

Mode for Carrying Out the Invention

[0011] (First Embodiment) Fig. 1 shows a system 1 for evaluating brain disorders such as dementia according to the present invention (dementia evaluation system). This system 1 includes a first input module 10 configured to obtain a first evaluation index X1 based on data related to the physical state of the subject's brain, a second input module 20 configured to obtain a second evaluation index X2 based on data related to the function of the subject's brain, and an estimation module 30 configured to estimate the state of a brain disorder (dementia and / or other brain disorders) including dementia of the subject based on an evaluation value fv obtained by a first evaluation function f1 having the first evaluation index X1 and the second evaluation index X2 as variables. The system 1 further includes a database 19 in which image data including the subject's brain image 18 is stored, and a database 29 in which clinical information 28 including the subject's cognitive test results is stored.

[0012] Brain disorders include mainly higher brain disorders such as dementia, attention disorder, memory disorder, executive function disorder, social behavior disorder, aphasia, apraxia, and agnosia. Dementia includes AD (Alzheimer Disease), DLB (Dementia with Lewy Bodies), and other degenerative dementias, such as frontotemporal dementia, progressive supranuclear palsy, corticobasal degeneration, and argyrophilic grain dementia. The state of brain disorder includes various aspects related to the brain disorder of the subject (examined person, patient, user), such as the presence or absence of brain disorder, its progression state, the presence or absence and differentiation of the causative disease (causal disease) of brain disorder such as dementia, and the progression state of single or multiple causative diseases.

[0013] One of the purposes of the present system 1 is to realize quantification by numerical values through a combination of multiple evaluations for various states of dementia, brain disorders including dementia, or brain disorders not including dementia of a subject, to enable categorization of various states, and to provide information for further evaluation and analysis for each category. Categorizing and performing evaluation and analysis on subjects with various states of brain disorder is effective not only for the treatment of subjects but also in other fields such as clinical research. For example, the present system 1 is also effective for evaluating the effects and impacts of various items ingested by people, such as pharmaceuticals, food and drink products, supplements, etc. on the brain or brain disorders including dementia, and may be used as a stratification marker. In addition, the present system 1 may be applicable to the evaluation of information devices, games, and other applications that may affect the brain or brain disorders. In the following, an example of evaluating the state of dementia will be described, including other purposes and effects of the present system.

[0014] The first input module 10 includes a first evaluation unit 11 that obtains a first evaluation index X1 obtained by statistically evaluating a first type of medical image of at least a part of the subject's brain of interest, and a second evaluation unit 12 that obtains the first evaluation index X1 obtained by evaluating the subject's medical image using a first model machine-learned to evaluate a first disease based on the first type of medical image.

[0015] As devices for diagnosing the morphology and function of a subject (test subject), various types of tomography devices (modalities) such as CT (Computed Tomography), MRI (Magnetic Resonance Imaging), PET (Positron Emission Tomography), SPECT (Single Photon Emission Computed Tomography), and PET-CT are known, and their modality images (medical images) are utilized for the diagnosis of various diseases. In particular, the modality image (medical image) 18 of the subject's brain is used to acquire data related to the physical state of the subject's brain and is utilized for the diagnosis of diseases such as dementia and Parkinson's disease. In this specification, the physical state of the brain refers to a state that can be measured by physical methods, and is a state that can be evaluated, measured, and estimated by methods such as statistical processing and learning models based on various modality images including MRI, PET, and SPECT that can measure morphology, glucose metabolism, blood flow, etc.

[0016] An example of the type of medical image is CT and MRI, and these images can reflect highly accurate morphological information. Another example of the type of medical image is PET and SPECT, and these images are generated by administering a radioactive drug into the subject's body by intravenous injection or the like and imaging the radiation emitted from the drug in the body. According to the image using the drug, not only the morphology of each part of the body but also how the drug administered into the body is distributed or the state of accumulation of substances in the body that react with the drug can be grasped by the doctor, which can contribute to improving the diagnostic accuracy of diseases. For example, by using the so-called Pittsburgh compound B as a PET radioactive drug (tracer) to image a PET image and measuring the degree of accumulation of amyloid-β protein in the brain based on the imaged PET image, it can be useful for differential diagnosis or early diagnosis of Alzheimer's dementia.

[0017] As an example of a SPECT image, there is an imaging method called DatSCAN (Dopamine transporter SCAN) that visualizes the distribution of the dopamine transporter (DAT) in a SPECT examination after administration of a radiopharmaceutical called 123I-ioflupane. The purposes of this imaging include early diagnosis of Parkinson's syndrome (PS) in Parkinson's disease (hereinafter PD), auxiliary diagnosis of dementia with Lewy bodies (DLB), and determination of the type of medication treatment called levodopa in cases where there is dopamine nerve loss in the striatum.

[0018] The first evaluation unit 11 uses statistical processing for the evaluation of medical images. The first evaluation unit 11 may perform statistical comparison and evaluation between the brain image of the subject and the brain image of a healthy person. As a method for evaluating brain atrophy using a brain image, voxel-based morphometry (VBM), which performs image processing in units of voxels, which are three-dimensional pixels, obtained by imaging the subject's head, is known. A typical statistical process is to generate a Z-score map. Therefore, the first evaluation unit 11 may obtain the Z-score as the first evaluation index X1.

[0019] Taking an MR image as an example, from the MR image of a normal case that has undergone brain morphological standardization processing, the values of the data (normal standard brain) in which the mean value and standard deviation are calculated for each voxel to create a mean image and a standard deviation image, and the values of the image data (processed image) of the subject are substituted into the following formula for calculating the Z-score. z=(M(x,y,z)-I(x,y,z)) / SD(x,y,z) M and SD represent the average image and the standard deviation image of a normal standard brain, respectively, and I represents the processed image. By using the Z-score map, it is possible to quantitatively analyze at which sites and what kinds of changes have occurred in the processed image compared to the normal standard brain. For example, voxels with positive values in the Z-score map indicate regions with atrophy compared to the normal standard brain, and it can be interpreted that the larger the value, the greater the statistical deviation. For example, if the Z-score is "2", it means exceeding twice the standard deviation from the average value, and it is evaluated that there is a statistically significant difference at a risk rate of about 5%. To quantitatively evaluate atrophy within a region, in the region of interest, M, SD, and I are calculated respectively, and the average of all positive Z-scores can be obtained.

[0020] As statistical processing methods, various methods have been proposed, such as comparing the volume or area of each part of the brain, or using the T-test with the General Linear Model (GLM).

[0021] Using the so-called Pittsburgh Compound B as a radioactive agent (tracer) for PET, by measuring the degree of amyloid-β protein accumulation in the brain based on the captured PET images, it can be useful for differential diagnosis or early diagnosis of Alzheimer's disease. As statistical processing, this PET image can adopt the SUVR value (SUVR, Standardized Uptake Value Ratio, cerebellum ratio SUVR), which represents the ratio of the sum of the accumulation degrees (SUV, Standardized Uptake Value) of amyloid-β protein in some cerebral gray matter of the brain to the accumulation degree (SUV) of amyloid-β protein in the cerebellum. SUVR can be defined by the following formula.

Equation

[0022] For the statistical processing of DatSCAN using SPECT images, BR (Binding Ratio) can be adopted as an evaluation (index value), which is expressed by the following formula.

Number

[0023] The second evaluation unit 12 evaluates the brain image 18 of the subject using a first model (learning model) that is machine - learned to evaluate a first disease, for example, AD (Alzheimer Disease), DLB (Dementia with Lewy Bodies), based on medical images of the same type or different types that are common to the first evaluation unit 11.

[0024] Disease discrimination is performed on a subject's medical image using a model (learning model) obtained by performing machine learning based on the information of the medical image. In "Deep-learning-based imaging-classification identified cingulate island sign in dementia with Lewy bodies." by Iizuka, Tomomichi et al. (Scientific reports 9.1 (2019): 1-9.), in an experiment using a convolutional neural network for perfusion SPECT images, an even higher accuracy of 89.32% was achieved, and it was reported that in the discrimination by deep learning, attention was paid to the blood flow findings in the occipital lobe that have been conventionally used for radiography. In "A survey on deep learning in medical image analysis." by Litjens, Geert et al. (Medical image analysis 42 (2017): 60-88.) and "Convolutional Neural Networks for Classification of Alzheimer's Disease: Overview and Reproducible Evaluation." by Wen, Junhao et al. (CoRR abs / 1904.07773 (2019)), it was reported that the application of recent deep learning techniques has shown high accuracy in the discrimination of AD.

[0025] The second evaluation unit 12 may obtain the output softmax value Xa of the activation function when estimating the first causative disease by the deep learning discrimination model as the first evaluation index X1. The first input module 10 may output the following values obtained by the first evaluation unit 11 and / or the second evaluation unit 12 as the first evaluation index X1. Xa: The output softmax value of the activation function when estimating the first causative disease by the deep learning discrimination model that takes a brain image as an input. Xb: The softmax value of the activation function when estimating the first causative disease by a deep learning discrimination model using the image obtained by filtering the brain image in the region of interest by statistical processing of the brain image. Xc: The volume value or blood flow volume of the region of interest by statistical processing of the brain image. Xd: The Z-score value of the volume or blood flow evaluation of the region of interest by statistical processing of the brain image. Xe: The volume value or blood flow volume of the region of interest when estimating the first causative disease by a deep learning discrimination model taking the brain image as an input. Xf: The Z-score value of the volume or blood flow evaluation of the region of interest when estimating the first causative disease by a deep learning discrimination model taking the brain image as an input.

[0026] The second input module 20 includes a configuration for obtaining an evaluation of clinical information 28 including a cognitive test as a second evaluation index X2. This second input module 20 includes a unit 22 for evaluating the cognitive test and a unit 21 for evaluating other clinical information related to the attributes of the user.

[0027] The cognitive test is a means for obtaining data related to the function of the subject's brain, and in particular, is used as a test for checking the cognitive function of the brain. The content of the cognitive test can include, for example, calculations such as addition and subtraction, Stroop, N-Back, word memorization, etc., but is not limited thereto. Specific examples of the cognitive test are disclosed, for example, in Japanese Patent Application Laid-Open No. 2019-75071 of the applicant of the present application, and include tests related to "forward recitation" and "reverse recitation", tests related to "Stroop", tests related to "addition" and "subtraction", tests related to N-Back (e.g., 1-Back), and tests related to "immediate reproduction" (word recall). Note that the cognitive test is not limited to these, and tests or tests for measuring the health state of the brain (including the state of cognitive function and the presence and degree of brain diseases and mental diseases) listed in FIGS. 3 and 4 or other similar forms of tests or tests may be adopted.

[0028] By arbitrarily combining or combining all of these tests, it is possible to provide a cognitive ability test suitable for judging the overall picture of brain function and the state of suffering from a specific causative disease. In this specification, the function of the brain refers to an ability that can be judged by artificial behaviors involving the brain, such as expression and comprehension ability, other than the physical state of the brain. Typically, the function of the brain may be appropriately judged based on the results of an appropriate cognitive ability test.

[0029] The results of the cognitive ability test can score the state of brain function based on, for example, the reaction time (response time) of the subject to the cognitive ability test or the number of correct answers (hereinafter, this scored result is referred to as the "cognitive ability score"). The information of the cognitive ability score may be used as a second evaluation index X2 for evaluating the function of the subject's brain. By representing the results (cognitive ability scores) of the cognitive ability test as a normal distribution, the brain age can be estimated, and the evaluation index may be corrected using the age of the subject. The evaluation index for evaluating the state of brain function can be obtained based on clinical information including the cognitive ability test score. As clinical information, in addition to the cognitive ability test score, any of age, gender, educational background, work history, genes (such as ApoE), blood test results, and interview results (such as ADL interview) may be included.

[0030] The second evaluation index X2 based on the data related to the function of the brain is also effective as an index that accurately indicates the risk of progression from pre-MCI (the stage before Mild Cognitive Impairment) to MCI (Mild Cognitive Impairment) and AD (Alzheimer's Disease).

[0031] Based on the evaluation value fv obtained by the first evaluation function f1 with the above-mentioned first evaluation index X1 and second evaluation index X2 as variables, a typical example of the estimation module 30 for estimating the dementia state of the subject is a function (disease estimation function, unit) 35 for estimating the morbidity state of the first causative disease, for example, AD or DLB. The first input module 10 includes a configuration for obtaining a first evaluation index X1 related to the discrimination of the first causative disease, and the second input module 20 includes a configuration for obtaining a second evaluation index X2 related to the discrimination of the first causative disease. The second input module 20 may include a configuration for obtaining, as the second evaluation index X2, the result 28 of a cognitive ability test suitable for the discrimination of the first causative disease.

[0032] Another example of the estimation module 30 is a function (clinical evaluation function, clinical evaluation unit) 36 for evaluating the effect of the ingested substance on dementia when the subject is included in a group that has ingested at least one of pharmaceuticals, food and drink products, and supplements. The evaluation result in the evaluation function 36 can be used as a stratification marker corresponding to the biomarker in stratified medicine. Therefore, the system 1 may include a module for providing information as a stratification marker based on the estimation of the estimation module 30. In this system 1, as will be described in more detail below, it is possible to provide a quantified stratification marker by the cooperation of two or more factors, and classify the patient into a category according to the purpose by the quantified marker, or find a category necessary for evaluation such as research or medical treatment. In the following, the unit 35 for estimating the morbidity state will be further described as an example.

[0033] The morbidity estimation unit 35 includes an odds determination unit 31 for determining the presence or absence of morbidity by odds and a probability determination unit 32 for determining the probability of morbidity. In the odds determination unit 31, the first evaluation index X1 is used as the odds x1 of the first causative disease, and the second evaluation index X2 is used as the odds x2 of the first causative disease, and the evaluation value s is obtained by the following first evaluation function f1a. s = x1 × x2 ··· (f1a)

[0034] When the probability determination unit 32 uses the first evaluation index X1 and the second evaluation index X2 as xi and their respective value weight coefficients as wi, it includes a configuration for obtaining the disease probability p as an evaluation value by the following first evaluation function f1b.

Equation

[0035] According to the logistic regression model, the log odds of the disease probability p of two classes, that is, one of the causative diseases (causative disorders) A (AD or DLB), is expressed by the following equation.

Equation

[0036] According to the evaluation function f1b, if the disease probability p is greater than 0.5, it can be determined that the disease is A, and if it is 0.5 or less, it can be determined that the disease is not A. When evaluating a plurality of causative diseases (multiple classes), multiple sets of one vs all can be created, and it may be determined that the causative disease (class) with the maximum p value corresponds. The desired first evaluation function f1c for the causative disease based on multinomial logistic regression may be as follows.

Equation

[0037] Figure 2 shows, in a flowchart, an evaluation method using the dementia evaluation system 1. The dementia evaluation system 1 can be provided as an information processing apparatus having computer resources including a memory and a CPU, and the control method can be provided as a program having instructions executable on a computer. The program (program product) may be recorded on a computer-readable recording medium and provided, or may be provided in a downloadable state from the Internet or the like. Further, the dementia evaluation system 1 may be provided as a service (SaaS (Software as a Service)) via the Internet.

[0038] In a system 1 having a first input module 10, a second input module 20, and an estimation module 30 configured to estimate the dementia state of a subject, in step 51, the estimation module 30 obtains a first evaluation index X1 from the first input module 10, and in step 52, the estimation module 30 obtains a second evaluation index X2 via the second input module 20. Further, in step 53, the estimation module 30 estimates the dementia state of the subject based on an evaluation value obtained by a first evaluation function having the first evaluation index X1 and the second evaluation index X2 as variables, for example, the evaluation function f1a or f1b described above.

[0039] In step 53, the estimation module 30 may perform a process 54 of estimating the morbidity state of the first causative disease. Also, when the subject is included in a group that has ingested at least one of pharmaceuticals, food and drink products, and supplements, the estimation module 30 may perform a process 58 of evaluating the effect of the ingested substance on dementia.

[0040] In step 51, the estimation module 30 may obtain, via the first input module 10, the value statistically evaluated by the first evaluation unit 11 for the medical image as the first evaluation index X1, or may obtain the value evaluated by the second evaluation unit 12 for the medical image as the first evaluation index X1. As shown in the values Xa to Xf, the first evaluation index X1 may be obtained that reflects both the value statistically evaluated for the medical image and the value when evaluating the medical image of the subject using the machine-learned model (the first model).

[0041] Also, in step 52, the estimation module 30 may obtain an evaluation of clinical information including a cognitive ability test as the second evaluation index X2. Further, in steps 51 and 52, the estimation module 30 may obtain the first evaluation index X1 regarding the discrimination of the first causative disease and the second evaluation index X2 regarding the discrimination of the first causative disease, and in step 53, may estimate the morbidity state of the first causative disease by the first evaluation function.

[0042] In step 53, when the evaluation value by the evaluation function exceeds the first threshold value, the estimation module 30 may perform a process of estimating the first causative disease. An example of the evaluation function is, as described above, the process 55 of performing odds determination and the process 56 of performing probability determination using a logistic regression model. Hereinafter, the process 55 of performing odds determination will be further described.

[0043] FIG. 5 shows an example of the distribution (number of incorrect answers) of the delayed reproduction scores of the cognitive ability test (ADAS-Jcog) obtained by the second input module 20 for each disease group. Specifically, it shows the number of incorrect answers in delayed reproduction for subjects with AD as the causative disease and subjects with DLB as the causative disease.

[0044] Figure 6 shows an odds table created from the distribution of the delayed reproduction scores of the cognitive ability test (ADAS-Jcog) for each disease group. At the same time, the first input module 10 obtains the output value of the activation function (softmax function) of the deep learning discrimination model with the brain image of each subject as the input. In the estimation module 30, the softmax value obtained from the first input module 10 is used as the first evaluation index X1, and the disease odds score with the result score of the clinical information (the above cognitive ability test) of the same subject (patient) as the input is used as the second evaluation index X2. By multiplying these indices X1 and X2 by the first evaluation function f1a, an evaluation value S indicating that the causative disease is AD is obtained.

[0045] Figure 7 shows the first evaluation index X1, the second evaluation index X2, and the evaluation value S obtained as described above. When the evaluation value S to be evaluated is equal to or greater than the cut-off value (threshold value, 50% in this example), it is possible to determine whether the brain disease, in this case AD, is present or not. In this example, as a result, the discrimination accuracy, which was 72% for the image alone, increased to 83% by jointly discriminating the disease using the first evaluation index X1 and the second evaluation index X2.

[0046] Note that the odds table may be constructed by, for example, taking the ratio of the proportion of the cognitive ability scores of the subjects in the disease group used as learning data to that of the control group. In this example, the MRI image and the delayed reproduction score of ADAS-Jcog were used for the determination of Alzheimer's disease and dementia with Lewy bodies, but it is not limited to this.

[0047] Figure 8 shows the first evaluation index X1, the second evaluation index X2, and the evaluation value py by logistic regression obtained as described above. When the evaluation value py to be evaluated is equal to or greater than the cut-off value (threshold value, 50% in this example), it is possible to determine whether the brain disease, in this case AD, is present or not. In this example, as a result, the discrimination accuracy, which was 72% for the image alone, increased to 83% by jointly discriminating the disease using the first evaluation index X1 and the second evaluation index X2.

[0048] Note that, in this example, the input to the logistic regression model was evaluated using MRI images and the delayed recall score of ADAS-Jcog for the determination of Alzheimer's disease and Lewy body dementia, but it is not limited to this.

[0049] Figure 9 shows the evaluation value py obtained by inputting, into logistic regression, the values obtained by calculating the atrophy degrees for three different brain regions for the first evaluation indices X1, X2, and X3 as Z scores, together with the second evaluation index X4. When the evaluated evaluation value py is equal to or higher than the cut-off value (threshold value, 50% in this example), it is possible to determine whether or not it is a brain disease, AD in this example. In this example, as a result, the discrimination accuracy, which was 56% for the image alone, increased to 78% by jointly discriminating the disease using the first evaluation indices X1 to X3 and the second evaluation index X4.

[0050] Note that, in this example, the input to the logistic regression model was evaluated using Z scores and the delayed recall score of ADAS-Jcog for the determination of Alzheimer's disease and Lewy body dementia, but it is not limited to this.

[0051] Figure 10 shows the case of determining the Negative / Positive of the image-based evaluation and the cognitive ability test score for a certain disease or its degree. By determining the Negative / Positive of the image-based evaluation and the cognitive ability test score for a certain disease or its degree, the following four categories can be obtained. Category A: Image (Positive) and cognitive ability test (Positive) Category B: Image (Negative) and cognitive ability test (Positive) Category C: Image (Positive) and cognitive ability test (Negative) Category D: Image (Negative) and cognitive ability test (Negative)

[0052] There is a method of determining a disease by using a cut-off for an evaluation index of a brain image. For example, for Alzheimer's disease patients and subjects with Lewy body dementia, the degree of blood flow reduction (CIScore) in a specific part of the occipital lobe is evaluated, and a method of discrimination using a cut-off value is known. In addition to setting a cut-off, a method of determining an attribute may be provided according to the degree of magnitude of the evaluation index of the brain image separately from the cut-off. Regarding the evaluation of brain diseases using the scores of cognitive tests, it is known that, for example, healthy subjects, mild cognitive impairment, and Alzheimer's disease patients can be discriminated by the cut-off values of CDR and MMSE. By using the evaluation values obtained by the above-described dementia evaluation system 1 and evaluation method, it becomes possible to more accurately perform the subdivision of the categories D and D' of the category concept diagram 10. Also, by using the evaluation of a single brain image, the results score of clinical information (for example, a cognitive test), and the type and progression classification of brain diseases evaluated by integrating both, it is possible to categorize the subject and determine the subject for evaluating the drug efficacy, and further use it as information for performing an evaluation analysis for each category.

[0053] This system 1 may include a module that outputs the estimation (evaluation) of the above-described estimation module 30 and / or the process leading to the estimation, the first evaluation index X1, the second evaluation index X2, and other information xi to an output medium including a smartphone, a PC, a tablet, and paper. The output form may be characters, figures, images, information encrypted in a QR code (registered trademark), etc., or information indicating the access destination of the information.

[0054] Also, this system 1 may include a module that classifies a subject into a predetermined category based on the estimation of the estimation module 30 and / or the process leading to the estimation, the first evaluation index X1, the second evaluation index X2, and other information xi. In drug discovery research and stratified medicine, it is known that patients belonging to a certain disease are classified into several subgroups using biomarkers, and treatments and evaluations suitable for each subgroup are performed. The estimation evaluation of the estimation module 30 can also be used as a stratification marker.

[0055] (Second Embodiment) In the second embodiment, as the first evaluation function, an equation (f1b) that converts log(p / (1 - p)) to a probability using a sigmoid function was used. That is, as the evaluation value, the disease probability p output from the logistic regression model was used. In the second embodiment, a configuration using ridge regression instead of logistic regression will be described. In ridge regression, the output evaluation value is a scalar, not a probability.

[0056] When the estimation module 30 uses at least one of the first evaluation index X1 and the second evaluation index X2 as the explanatory variable xi, and the weight coefficient of each explanatory variable xi is wi, the estimated value ŷ output as the evaluation value can be calculated by equation (1) as the first evaluation function. Note that x0 = 1 and w0 is the intercept.

[0057]

Equation

[0058] Let the disease labels to be separated be -1 and 1. For example, let label -1 represent a healthy person and label 1 represent dementia. Classification can be performed based on whether ŷ is greater than or less than 0 for the unknown data xi.

[0059] The weight coefficient wi in equation (1) can be obtained by solving an optimization problem that minimizes the loss function F shown in equation (2). In equation (2), k = 1,..., n is the number of data samples, yk is the measured value. β is a parameter that can be set in advance and determines the magnitude of the influence of the regularization term represented by the square of the L2 norm of wi.

[0060]

Equation

[0061] As described above, the estimation module 30 can estimate the state of a brain disorder including dementia of a subject based on an evaluation value obtained by a first evaluation function using a first evaluation index and a second evaluation index as explanatory variables. Here, the first evaluation function is represented by a linear combination of explanatory variables with weight coefficients corresponding to the respective explanatory variables predicted by ridge regression using learning data, as shown in Equation (1). Note that the explanatory variable may be only one of the first evaluation index and the second evaluation index.

[0062] By using ridge regression, it is possible to solve the problem of multicollinearity in which the calculation of the estimated value becomes unstable when there are many explanatory variables, for example, when the explanatory variables are correlated with each other.

[0063] (Third Embodiment) In the third embodiment, in order to evaluate the risk of a degenerative brain disease, a system for evaluating each brain region (region of interest) of a brain image (MR, SPECT, PET, etc.) of a subject and a system for evaluating the risk of a brain disease targeting the whole brain based on the evaluation value of a characteristic region of the subject will be described. Note that the MR image is also referred to as an MRI image. The MRI image includes, for example, a T1-weighted image, a T2-weighted image, a diffusion-weighted image, a FLAIR image, a diffusion tensor image, a QSM image, a pseudo-PET image, a pseudo-SPECT image, and the like.

[0064] Brain diseases include dementia (including AD, DLB, frontotemporal lobar degeneration (FTLD), normal pressure hydrocephalus (NPH), etc.), brain tumors, mental disorders (also referred to as mental illnesses, including schizophrenia, epilepsy, mood disorders, substance use disorders, higher-order function disorders, etc.), Parkinson's disease, Asperger's syndrome, attention deficit hyperactivity disorder (ADHD), sleep disorders, pediatric diseases, ischemic brain disorders, mood disorders (including depression, etc.). Brain disorders include, as diseases related to the brain, dementia and multiple sclerosis, etc., and as diseases related to amyloid-β, for example, mild cognitive impairment (MCI), mild cognitive impairment due to Alzheimer's disease (MCIdue to AD), prodromal Alzheimer's disease, preclinical AD (preclinical AD), Parkinson's disease, multiple sclerosis, insomnia, sleep disorders, decline in cognitive function, cognitive dysfunction, neurodegenerative diseases such as amyloid-positive / negative related diseases, etc.

[0065] Figure 11 is a schematic diagram showing a first example of the configuration of the system of the third embodiment. The system (also referred to as the "evaluation system" or "dementia evaluation system") includes a first input module 10 and an estimation module 40. The first input module 10 has the same configuration as in the case of the first embodiment. The estimation module 40 includes an evaluation value calculation unit 41 and an output unit 42. The first input module 10 can access the necessary information by referring to the subject DB60. The subject DB60 may be the database 19 of the first embodiment. The estimation module 40 can access the necessary information by referring to the healthy subject DB61 and the brain disease patient DB62 respectively.

[0066] The first input module 10 includes a first evaluation unit 11 and a second evaluation unit 12. The first evaluation unit 11 calculates a first evaluation index X1 obtained by statistically evaluating medical images of each part (region of interest) of the subject's brain. The second evaluation unit 12 outputs the first evaluation index X1 of each part (region of interest) of the subject's brain using a trained model that has been machine-learned so as to output the first evaluation index X1 of each part (region of interest) of the subject's brain based on the medical images. Note that only one of the first evaluation unit 11 and the second evaluation unit 12 may be used, or both may be used. The first input module 10 outputs the first evaluation index X1 of each part (region of interest) of the subject's brain to the estimation module 40. Hereinafter, as the first evaluation index X1, the Z-score value in the anatomical standard space of the gray matter volume value of the region of interest will be described, but the first evaluation index X1 is not limited to the Z-score.

[0067] The Z-score of a specific part of the brain can be obtained for each pixel of the part by the following formula. Z-score = (Pixel value of the part of the subject's brain - Average value of the part of healthy subjects) / (Standard deviation of the part of healthy subjects) The Z-score of a part can be obtained as the average value of the positive Z-scores for each pixel of the part. The Z-score represents the degree to which the pixel value of the part of the subject deviates from the pixel value of the part of the healthy subject's brain. In the case of an MR image, the larger the value of the Z-score, the more atrophy is indicated compared to healthy subjects. Examples of the part (region of interest) of the brain include, but are not limited to, the diencephalon, superior parietal lobe, inferior parietal lobe, globus pallidus, cerebellum, paracentral lobule, hippocampus, parahippocampal gyrus, precentral gyrus, lateral ventricle, tonsil, entorhinal cortex, brainstem, etc.

[0068] The first input module 10 can calculate Extent and Ratio for each part. Extent represents the ratio of the number of voxels with a Z-score of 2 or more within the part to the total number of voxels within the part. When the Z-score is 2 or more, the value is more than twice the standard deviation from the average value of the pixel values, and a statistically significant difference is recognized. Ratio is the value obtained by dividing the average Z-score within the part by the average Z-score of the whole brain. The first input module 10 outputs the calculated Extent and Ratio to the estimation module 40.

[0069] The output unit 42 has a function as an output part and outputs display data for display on a display device (not shown). The display device may be incorporated within the system or may be an external device of the system.

[0070] FIG. 12 is a schematic diagram showing an example of the evaluation result of the hippocampal region. FIG. 12 schematically shows an image when the brain of the subject is viewed from the front, and may be different from the actual image. In the figure, the patterned parts are the left and right hippocampi respectively, and are displayed in a map form (evaluation index map) with the color or pattern changed according to the value of the first evaluation index (Z-score). In the example of FIG. 12, the Z-score is 0.7, the Extent is 20%, and the Ratio is 1.5 times. Note that the Z-score is calculated with each of the left and right hippocampal regions as one region. Also, the numerical values shown in FIG. 12 are for convenience and may be different from the actual ones.

[0071] FIG. 13 is a schematic diagram showing an example of the evaluation result of the middle temporal gyrus region. FIG. 13 schematically shows an image of a cross-section of the middle temporal gyrus of the subject, and may be different from the actual image. In the figure, the patterned parts are the left and right middle temporal gyri respectively, and are displayed in a map form with the color or pattern changed according to the value of the first evaluation index (Z-score). In the example of FIG. 13, the Z-score is 0.41, the Extent is 0%, and the Ratio is 0 times. Note that the Z-score is calculated with each of the left and right middle temporal gyrus regions as one region. Also, the numerical values shown in FIG. 13 are for convenience and may be different from the actual ones.

[0072] For parts other than the hippocampal region and the middle temporal gyrus region, they can be similarly displayed. By providing evaluation results such as those in FIGS. 12 and 13 to doctors or the like, doctors or the like can use them as materials for diagnosing various brain diseases including dementia of the subject.

[0073] The evaluation value calculation unit 41 has a function as a calculation unit, and calculates the region evaluation value (also referred to as the "evaluation value of the part") of each of a plurality of parts (regions of interest) of the subject's brain based on the first evaluation index of each of the plurality of parts (regions of interest) of the subject's brain and the weight coefficient corresponding to each first evaluation index.

[0074] FIG. 14 is a schematic diagram showing an example of the evaluation value of each part. As shown in FIG. 14, let the index of the part be j and the number of parts be m. Each part can be represented by j = 1, 2,..., m. Let the evaluation index of each part j be xdj and the weight coefficient of the evaluation coefficient xdj be wdj. The evaluation value Ej of each part can be represented by E1 = wd1·xd1, E2 = wd2·xd2,..., Em = wdm·xdm.

[0075] The output unit 42 can output the evaluation value of the part calculated by the evaluation value calculation unit 41. The output unit 42 may output display data for arranging and displaying the evaluation values of each of a plurality of parts of the subject's brain in a predetermined order (for example, in descending order of the evaluation value). When the evaluation value calculation unit 41 receives a selection of a required subject from among a plurality of subjects, it can calculate the evaluation values of each of a plurality of parts of the selected subject's brain.

[0076] FIG. 15 is a schematic diagram showing a first example of a site comparison screen. The site comparison screen has a region for displaying a list of subjects and a region for displaying the evaluation values of the sites for each subject's site (ROI). In the subject list, the IDs, names, etc. of a plurality of subjects are listed, and the target subject can be selected. For example, a doctor or the like can select a subject to be diagnosed from the list. In the example of FIG. 15, the subject OOOX surrounded by the dashed line is selected. The evaluation values of the subjects' sites can be arranged and displayed in descending order of the evaluation values. In the example of FIG. 15, since the evaluation value of the diencephalon is 0.5, which is the largest value, it is displayed at the top. Hereinafter, the ROI and the evaluation value of the site are displayed in association with each other in descending order of the evaluation value. Thereby, when a subject suffers from a specific brain disease, it can be used to determine which part of the brain is contributing. Note that the numerical values illustrated in FIG. 15 are for convenience and may be different from the actual values.

[0077] On the screen shown in FIG. 15, by appropriately selecting a subject, a plurality of subjects can be dynamically switched, and the evaluation values of the sites can be displayed.

[0078] FIG. 16 is a schematic diagram showing a second example of a site comparison screen. The difference from the first example illustrated in FIG. 15 is that the evaluation values of each site of a healthy subject and the evaluation values of each site of a specific brain disease patient are displayed. The estimation module 40 has a function as a healthy subject evaluation value acquisition unit and can acquire the evaluation values of a plurality of sites of a healthy subject (for example, the average of the evaluation values of a large number of healthy subjects) from the healthy subject DB 61. Further, the estimation module 40 has a function as a brain disease patient evaluation value acquisition unit and can acquire the evaluation values of a plurality of sites of a brain disease patient suffering from a specific brain disease (for example, the average of the evaluation values of a large number of brain disease patients suffering from a specific brain disease) from the brain disease patient DB 62. The estimation module 40 can selectively acquire the evaluation values of a plurality of sites of a specific brain disease patient from among a plurality of types of brain diseases.

[0079] In FIG. 16, the configuration is to display the evaluation values of each part of both healthy subjects and brain disease patients. However, the evaluation values of each part of either the healthy subjects or the brain disease patients may be displayed. That is, the output unit 42 can output display data for displaying the evaluation values of each of a plurality of parts of the subject and the healthy subjects in a display mode that enables comparison. Also, the output unit 42 can output display data for displaying the evaluation values of each of a plurality of parts of the subject and the brain disease patients in a display mode that enables comparison. Although not shown, on the comparison screen of the parts illustrated in FIG. 16, a plurality of types of brain diseases are listed, and the required brain disease can be selected from the listed brain diseases. Each time a brain disease is selected, the evaluation values of each part of the brain disease patient suffering from the selected brain disease may be displayed.

[0080] As shown in FIG. 16, by displaying the evaluation value of each part of the subject and the evaluation value of each part of at least one of the healthy subject and the brain disease patient in a display mode that enables comparison, the state of the brain disease of the subject can be judged compared with the healthy subject or compared with the brain disease patient.

[0081] FIG. 17 is a schematic diagram showing a third example of the part comparison screen. The example of FIG. 17 shows that it is possible to estimate the brain disease of the subject based on the evaluation value for each part of the subject. In a subject having the core symptoms of dementia, there is a suggestive feature of the imaging finding that "in Lewy body dementia (DLB), medial temporal lobe atrophy is relatively mild compared to Alzheimer's disease (AD)". Based on the evaluation value related to the atrophy for each part of the subject, the above-mentioned suggestive feature can be judged. Here, the medial temporal lobe is a gray matter region including the parahippocampal gyrus and the hippocampus. As shown in FIG. 17, in the case of subject A, the atrophy evaluation of the hippocampus appears at the top. However, compared with the evaluation value of the brain disease patient, the evaluation value of subject A is small, and it can be judged that the atrophy state is mild. On the other hand, in the case of subject B, it can be judged that the evaluation values of the hippocampus and the parahippocampal gyrus are large and the atrophy has progressed. In such a case, it can be estimated that subject A has Lewy body dementia, and it can be estimated that subject B has Alzheimer's disease.

[0082] Next, an evaluation of the disease risk for the entire brain of the subject will be described based on the evaluation values for each part of the subject's brain.

[0083] The estimation module 40 has a function as an estimation unit, and estimates the state of brain disorders including dementia of the subject based on the whole-brain evaluation value obtained by a second evaluation function using, as variables, the first evaluation indices for each of a plurality of parts of the subject's brain. Specifically, the evaluation value calculation unit 41 can calculate the whole-brain evaluation value using the second evaluation function.

[0084] The second evaluation function can be represented by Expression (3).

[0085]

Equation

Equation

[0086] In Expression (3), EA is the predicted value of the whole-brain evaluation value, xdj is the evaluation index of part j, wdj is the weight coefficient of the evaluation coefficient xdj, m is the number of parts, and ε is a constant for evaluating the error.

[0087] The weight coefficient wdj in Expression (3) can be obtained by solving an optimization problem that minimizes the loss function L shown in Expression (4). In Expression (4), i = 1, …, n is the number of samples of the learning data, E is the measured value of the whole-brain evaluation value, and α is a parameter that can be set in advance and determines the magnitude of the influence of the regularization term represented by the square of the L2 norm of wdj. That is, the second evaluation function represented by Expression (3) is represented by a linear combination of the first evaluation indices xdj with the weight coefficients wdj corresponding to the respective first evaluation indices xdj predicted by ridge regression using the learning data as coefficients.

[0088] Let the labels of each class to be separated be -1 and 1. For example, assume that label -1 represents a healthy person and label 1 represents a person with dementia. Classification can be performed based on whether the whole-brain evaluation value EA is greater than or less than 0 for an unknown evaluation index xdj. Also, when there are three or more classes to be separated (for example, three classes are a healthy person, brain disease B1, and brain disease B2), the brain diseases can be classified by a majority vote for all possible two-class combinations. Specifically, for the three combinations of a healthy person and brain disease B1, brain disease B1 and brain disease B2, and brain disease B2 and a healthy person, if brain disease B1 is classified 2 times, brain disease B2 is classified 1 time, and the healthy person is classified 0 times, it can be classified as brain disease B1 with the most number of times.

[0089] As described above, by using the ridge regression model, the problem of multicollinearity can be solved. This will be explained below.

[0090] FIG. 18 is a schematic diagram showing an example of the correlation of Z scores between each part of the left and right brains. j is an index indicating the part. Each part of the right brain is represented by j = 1 to 51. Each part of the left brain is represented by j = 52 to 102. Here, the part j of the right brain corresponds to the part (j + 51) of the left brain. For example, if the part j of the right brain is the hippocampus, the part (j + 51) of the left brain is also the hippocampus. In FIG. 18, the straight line indicated by the broken line represents the part with a large correlation of Z scores. That is, the same parts on the left and right tend to have a large correlation of Z scores.

[0091] If a function based on a logistic regression model is used as the second evaluation function, variables in the logistic regression model will be correlated with other variables, resulting in the problem of multicollinearity and unstable calculation of estimated values. Therefore, by using a ridge regression model represented by Equation (3), a regularization term is added, so the problem of multicollinearity can be solved. Specifically, when a function based on a logistic regression model is used, at the same part of the right and left brains, there are many parts where one of the right and left brains has a positive weight coefficient and the other has a negative weight coefficient, and the whole-brain evaluation value cannot be accurately calculated. Originally, all weight coefficients are expected to be positive coefficients. By using a function based on a ridge regression model, the number of parts with negative weight coefficients can be significantly reduced, and for example, the whole-brain evaluation value can be obtained with an accuracy of about 82%.

[0092] Next, a method for further improving the estimation accuracy of the whole-brain evaluation value when using a ridge regression model will be described.

[0093] FIG. 19 is a schematic diagram showing an example of the aggregation method of the left and right brains. As shown in FIG. 19, the part of the right brain is represented by j, and the same part of the left brain as the right brain is represented by (j + 51). The first aggregation method is a method of aggregating the right brain part j and the left brain part (j + 51) into one region by unifying them to the right brain index j. If the number of parts in each of the right and left brains is 51 and the number of parts in the whole brain is 102, the number of parts becomes 51 from 102 by the first aggregation method. By ridge regression using the first aggregation method, the estimation accuracy of the whole-brain evaluation value could be about 96%. Note that the number of parts in the whole brain is not limited to 102, and other numerical values may also be used.

[0094] The second aggregation method is a method of unifying the average of the weight coefficient of the right brain part j and the weight coefficient of the left brain part (j + 51) to the right brain part index. If the number of parts in each of the right and left brains is 51 and the number of parts in the whole brain is 102, the number of parts becomes 51 from 102 by the second aggregation method. By ridge regression using the second aggregation method, the estimation accuracy of the whole-brain evaluation value could be about 91%.

[0095] The third aggregation method is a method of unifying the value of the larger of the weight coefficients of the right brain region j and the weight coefficient of the left brain region (j + 51) to the right brain region index. Assuming that the number of parts in each of the right brain and the left brain is 51 and the total number of parts in the brain is 102, the number of parts is reduced from 102 to 51 by the third aggregation method. By ridge regression using the third aggregation method, the estimation accuracy of the whole brain evaluation value could be set to about 82%.

[0096] Next, a specific example of the whole brain evaluation value EA calculated by the evaluation value calculation unit 41 will be described.

[0097] FIG. 20 is a schematic diagram showing a specific example of the whole brain evaluation value. As described above, the whole brain evaluation value EA can be obtained by the formula (3). When the above-described aggregation method is used, the number of parts j can be, for example, 51, and the number of part evaluation values is also 51. In FIG. 20, for simplicity, the meaning of the whole brain evaluation value is explained on a two-dimensional plane by setting the number of parts to 2. As shown in FIG. 20, let the part evaluation values be E1 and E2. The distance from the class division line (plane) that divides between classes (healthy subjects, brain disease patients) on the two-dimensional plane is the whole brain evaluation value. For example, in the case of the subject S1, -d1 represents the whole brain evaluation value, and in the case of the subject S2, d2 represents the whole brain evaluation value. Here, the healthy subject side is assigned a negative sign, and the brain disease patient side is assigned a positive sign. By using the whole brain evaluation value of the subject, as shown below, it is possible to visually represent whether the subject is close to a healthy subject or a brain disease patient.

[0098] FIG. 21 is a schematic diagram showing a first example of the whole brain evaluation screen. The whole brain evaluation screen has a region for displaying a list of subjects and a region for displaying the whole brain evaluation value of the subject. In the subject list, the IDs, names, etc. of a plurality of subjects are listed, and the target subject can be selected. For example, a doctor or the like can select a subject to be diagnosed from the list. In the example of FIG. 21, the subject OOOO surrounded by the broken line is selected.

[0099] The estimation module 40 has a function as a whole-brain evaluation value acquisition unit, and can acquire the whole-brain evaluation value of a healthy person from the healthy person DB 61 and the whole-brain evaluation value regarding the required brain disease of a brain disease patient from the brain disease patient DB 62. The whole-brain evaluation value of the subject can be displayed in a display mode comparable to the whole-brain evaluation values of healthy persons and brain disease patients (for example, the average value of the whole-brain evaluation values). In the example of FIG. 21, one end of the horizontal bar graph represents a healthy person, the other end represents a brain disease patient, and the subject is represented by a position on the horizontal bar graph. Thereby, it is possible to visually represent whether the subject is close to a healthy person or a brain disease patient.

[0100] On the screen shown in FIG. 21, by appropriately selecting a subject, a plurality of subjects can be dynamically switched to display the whole-brain evaluation value. Also, although not shown, on the whole-brain evaluation screen exemplified in FIG. 21, a plurality of types of brain diseases are listed, and the required brain disease can be selected from the listed brain diseases. Each time a brain disease is selected, a horizontal bar graph showing brain disease patients suffering from the selected brain disease at the other end may be displayed.

[0101] FIG. 22 is a schematic diagram showing a second example of the whole-brain evaluation screen. As shown in FIG. 22, the whole-brain evaluation value of the subject, the whole-brain evaluation values of healthy persons and brain disease patients can be represented on a radar chart. The axes of a plurality of brain diseases are arranged in a regular polygon shape from the center. In the example of FIG. 22, three brain diseases, brain disease 1, 2, and 3, are represented by an equilateral triangular radar chart. The position of the healthy person on the radar chart is indicated by a broken line, and the position of the subject is represented by a solid line. Thereby, for each brain disease, it is possible to visually represent whether the subject is close to a healthy person or a brain disease patient.

[0102] In the first example of the system exemplified in FIG. 11, the first evaluation index X1 was used, but in addition to the first evaluation index, the second evaluation index X2 may be used.

[0103] FIG. 23 is a schematic diagram showing a third example of the whole brain evaluation screen. As shown in FIG. 23, based on the whole brain evaluation value of the subject, the possibility of each brain disease (in the example of FIG. 23, brain diseases 1, 2, 3, and 4) is visually represented by the size of a circle like a bubble chart. For example, for each brain disease, as the whole brain evaluation value of the subject approaches the average of the whole brain evaluation values of healthy subjects, the size of the circle becomes smaller. Conversely, as the whole brain evaluation value of the subject approaches the average of the whole brain evaluation values of brain disease patients, the size of the circle becomes larger. In the example of FIG. 23, it can be seen that subject OOOO is highly likely to be a brain disease patient with brain disease 1. Also, it can be seen that subject OOOO may have or cannot be denied the possibility of brain diseases 2 and 3. Furthermore, for brain disease 4, it can be seen that subject OOOO is, for example, at the healthy subject level. Note that the types of brain diseases are not limited to four as shown in FIG. 23.

[0104] FIG. 24 is a schematic diagram showing a fourth example of the whole brain evaluation screen. In FIG. 24, the symbols A to F in each cell can represent the type of brain disease or the part of the brain. The strength of the association of each brain disease or each part can be expressed by the numbers in the table. The numbers may be the whole brain evaluation values of the subject. Also, as shown in FIG. 24, the strength of the association of each brain disease or each part can also be represented by the patterns or colors attached to each cell. In the example of FIG. 24, it can be seen that subject OOOO has a strong association with brain disease (or part) C.

[0105] FIG. 25 is a schematic diagram showing a fifth example of the whole brain evaluation screen. As shown in FIG. 25, based on the whole brain evaluation value of the subject, the possibilities of each brain disease (in the example of FIG. 23, brain diseases 1 and 2) are visually represented in a matrix chart. It is divided into four regions S1 to S4 by a two-axis matrix. Region S1 indicates that the possibility of brain disease 2 is high and the possibility of brain disease 1 is low (that is, it can be determined that it is brain disease 2). Region S2 indicates that the possibility of brain disease 2 is high and the possibility of brain disease 1 is also high (that is, it can be determined that it is both brain diseases 1 and 2). Region S3 indicates that the possibility of brain disease 2 is low and the possibility of brain disease 1 is also low (that is, it can be determined that it is neither brain disease 1 nor 2). Region S4 indicates that the possibility of brain disease 2 is low and the possibility of brain disease 1 is high (that is, it can be determined that it is brain disease 1). In the example of FIG. 25, the subject OOOO can be determined to have both brain diseases 1 and 2.

[0106] FIG. 26 is a schematic diagram showing a second example of the configuration of the system according to the third embodiment. The second example is different from the second example in that it includes a second input module 20. Since the second input module 20 is the same as that in the first embodiment shown in FIG. 1, the description thereof is omitted. The estimation module 40 can estimate the state of a brain disorder including dementia of the subject based on the whole brain evaluation value obtained by a second evaluation function using, as variables, the first evaluation index and the second evaluation index X2 for each of a plurality of parts of the subject's brain. Specifically, the evaluation value calculation unit 41 can calculate the whole brain evaluation value using the second evaluation function. The second evaluation function can be represented by the above-mentioned formula (3), and the evaluation index xdj includes the first evaluation index and the second evaluation index X2.

[0107] FIG. 27A and FIG. 27B are schematic diagrams showing examples of ROC curves. FIG. 27A shows an ROC (Receiver Operating Characteristic) curve when differentiating between AD and DLB using only the first evaluation index (Z-score of 51 sites), and FIG. 27B is an ROC curve when differentiating using the second evaluation index (delayed reproduction score in the ADASCog cognitive test) in addition to the first evaluation index. In FIGS. 27A and 27B, the vertical axis represents the true positive rate (TPR), and the horizontal axis represents the FRP (false positive rate). The ROC curve is obtained by calculating and plotting the true positive rate and the false positive rate for each cut-off point that distinguishes normal from abnormal in AD diagnosis and DLB diagnosis. The area under the plotted graph is called the AUC (Area Under the Curve), and the AUC can take values from 0 to 1. The closer the AUC is to 1, the higher the discrimination accuracy. In FIG. 27A, AUC = 0.62, and in FIG. 27B, AUC = 0.74. FIG. 27B has a higher AUC. From the example of FIG. 27, it can be seen that the discrimination accuracy is improved by adding the delayed reproduction score that evaluates the decline in long-term memory (more pronounced in AD cases).

[0108] Next, the operation of the system according to the third embodiment will be described.

[0109] FIG. 28 is a flowchart showing the procedure of the first example of the output process of the evaluation results of each part of the brain. The system acquires the image data of the subject's brain (S11) and sets the region of interest (part) of the subject (S12). The system calculates the evaluation index (for example, Z-score) for each region of interest (S13) and calculates Extent and Ratio for each region of interest (S14).

[0110] The system generates an evaluation index map for each region of interest based on the calculated evaluation index (S15) and outputs the subject's evaluation index map, evaluation index, Extent, and Ratio (S16). The system determines the presence or absence of other regions of interest (S17). If there are other regions of interest (YES in S17), the processing after step S12 is continued.

[0111] If there is no other region of interest (NO in S17), the system determines the presence or absence of other subjects (S18). If there are other subjects (YES in S18), the system continues the processing after step S11. If there are no other subjects (NO in S18), the processing ends.

[0112] Figure 29 is a flowchart showing the procedure of the second example of the output process of the evaluation results of each part of the brain. The system acquires image data of the subject's brain (S31) and sets the region of interest (part) of the subject (S32). The system calculates an evaluation index (for example, Z-score) for each region of interest (S33) and calculates a part evaluation value Ej for each region of interest of the subject (S34). The part evaluation value Ej can be calculated by Ej = wdj·xdj. Here, xdj is the evaluation index (for example, Z-score) of part j, and wdj is a weighting coefficient.

[0113] The system acquires the part evaluation values for each region of interest of healthy subjects from the healthy subject DB61 (S35) and acquires the part evaluation values for each region of interest of brain disease patients from the brain disease patient DB62 (S36). The system determines whether to display in the order of part evaluation values (for example, in descending order) (S37). If it is to be displayed in the order of part evaluation values (YES in S37), the system sorts the part evaluation values of the subject (S38) and performs the processing of step S39 described later.

[0114] If it is not to be displayed in the order of part evaluation values (NO in S37), the system determines whether to display the part evaluation values of the subject in comparison with healthy subjects and brain disease patients (S39). If it is to be displayed in comparison with healthy subjects and brain disease patients (YES in S39), the system outputs the part evaluation values of the subject, healthy subjects, and brain disease patients (S40) and performs the processing of step S42 described later.

[0115] If it is not to be displayed in comparison with healthy subjects and brain disease patients (NO in S39), the system outputs the part evaluation values of the subject (S41) and determines the presence or absence of other subjects (S42). If there are other subjects (YES in S42), the system continues the processing after step S31. If there are no other subjects (NO in S42), the processing ends.

[0116] Figure 30 is a flowchart showing the procedure of output processing of the whole-brain evaluation result. The system acquires image data of the subject's brain (S51) and sets the region of interest (part) of the subject (S52). The system calculates an evaluation index (e.g., Z-score) for each region of interest (S53) and calculates a site evaluation value Ej for each region of interest of the subject (S54). The site evaluation value Ej can be calculated by Ej = wdj·xdj. Here, xdj is the evaluation index (e.g., Z-score) of site j, and wdj is a weight coefficient.

[0117] The system calculates the whole-brain evaluation value EA (S55). The whole-brain evaluation value EA can be calculated by Equation (3). The system acquires the whole-brain evaluation value of healthy subjects from the healthy subject DB61 (S56) and acquires the whole-brain evaluation value of brain disease patients from the brain disease patient DB62 (S57). The system calculates the position of the subject's whole-brain evaluation value within the range of the whole-brain evaluation values of healthy subjects and brain disease patients respectively (S58). For example, if the average of the whole-brain evaluation values of healthy subjects is a, the average of the whole-brain evaluation values of brain disease patients is b, and the whole-brain evaluation value of the subject is c, the position can be determined according to where the value c is within the range from a to b.

[0118] The system outputs the subject's whole-brain evaluation value in a display mode (e.g., see Figure 21) that can be compared with the whole-brain evaluation values of healthy subjects and brain disease patients respectively (S59). The system determines the presence or absence of other subjects (S60). If there are other subjects (YES in S60), the system continues the processing after step S51. If there are no other subjects (NO in S60), the processing ends.

[0119] In the above-described third embodiment, the Z-score value in the anatomical standard space of the gray matter volume value of the region of interest was used as the first evaluation index X1. However, the first evaluation index is not limited to the Z-score. For example, physical quantities such as the blood flow volume of the region of interest and the accumulation amount of malignant proteins (e.g., amyloid-β, tau protein, etc.) in the region of interest may be used. When using such physical quantities, the ratio of the physical quantity in the region of interest exceeding a predetermined threshold (e.g., the ratio of the number of voxels exceeding the threshold to the total number of voxels in the region of interest) may be used for normalization. Thereby, the evaluation index can be made comparable between regions of interest regardless of the size of the region of interest.

[0120] Also, at least one of SUVR and BR may be used as the first evaluation index X1. In this case, in Equation (3), for example, SUVR ·SUVR, BR ·BR, at least one of each term may be added. Here, SUVR is the weight coefficient of the evaluation index SUVR, and BR is the weight coefficient of the evaluation index BR. Further, when using the evaluation index SUVR or the evaluation index BR, if their index values are made into standardized values from the distribution, etc., that a healthy subject DB has, then, similar to the case where each site is evaluated only by the Z-score, the importance when each subject is judged to be a patient can be evaluated. For example, the standardized value of SUVR, SUVR Z can be obtained by SUVR Z ={(average of healthy subjects of SUVR - SUVR of the subject) / standard deviation of healthy subjects of SUVR}.

Explanation of symbols

[0121] 1 Dementia evaluation system (system) 10 First input module 11 First evaluation unit 12 Second evaluation unit 20 Second input module 21, 22 Units 30 Estimation module 31 Odds determination unit 32 Probability determination unit 35 Disease onset estimation unit 36 Clinical evaluation unit (clinical evaluation function, function) 40 Estimation module 41 Evaluation value calculation unit 42 Output unit 60 Subject DB 61 Healthy person DB 62 Brain disease patient DB

Claims

1. A first input module configured to obtain a first evaluation index based on data related to the physical state of a subject's brain, a second input module configured to obtain a second evaluation index which is the odds x2 of a first causative disease based on data related to the function of the subject's brain, and a estimation module configured to estimate the state of a brain disorder including dementia of the subject based on an evaluation value obtained by a first evaluation function having the first evaluation index and the second evaluation index as variables.

2. In Claim 1, the first input module includes at least one of a first evaluation unit that obtains the first evaluation index by statistically evaluating a first type of medical image of at least a part of the region of interest of the subject's brain, and a second evaluation unit that obtains the first evaluation index by evaluating the medical image of the subject using a first model that has been machine-learned to evaluate a first disease based on the first type of medical image.

3. In Claim 1 or 2, the second input module includes a configuration for obtaining an evaluation of clinical information including a cognitive ability test as the second evaluation index.

4. In any one of Claims 1 to 3, the subject is included in a group that has ingested at least one of pharmaceuticals, food and drink products, and supplements, and the estimation module includes a function of evaluating the effect of the ingested substance on dementia.

5. In any one of Claims 1 to 3, the estimation module includes a function of estimating the morbidity state of a first causative disease.

6. In Claim 5, the first input module includes a configuration for obtaining the first evaluation index regarding the discrimination of a first causative disease, the second input module includes a configuration for obtaining a second evaluation index regarding the discrimination of the first causative disease, and the estimation module includes the first evaluation function for estimating the morbidity state of the first causative disease.

7. In Claim 6, the first input module includes a configuration for obtaining at least any one of the following values as the first evaluation index. a: The output softmax value of the activation function when estimating the first causative disease by a deep learning discrimination model that takes a brain image as an input b: Region of interest obtained by statistical processing of brain images. Softmax value of the activation function when estimating the first causative disease using the image obtained by filtering the brain image through a deep learning discrimination model. c: Volume value or blood flow of the region of interest obtained by statistical processing of brain images. d: Z-score value of the volume or blood flow evaluation of the region of interest obtained by statistical processing of brain images. e: Volume value or blood flow of the region of interest when estimating the first causative disease using a deep learning discrimination model with a brain image as input. f: Z-score value of the volume or blood flow evaluation of the region of interest when estimating the first causative disease using a deep learning discrimination model with a brain image as input.

8. In claim 6 or 7, The second input module includes a configuration for obtaining, as the second evaluation index, the result of a cognitive ability test suitable for the discrimination of the first causative disease. A system.

9. In any one of claims 6 to 8, The estimation module includes a first evaluation function for estimating the first causative disease when the evaluation value exceeds a first threshold. A system.

10. In any one of claims 6 to 9, The first evaluation index is the odds x1 of the first causative disease, and the estimation module includes a configuration for obtaining the evaluation value s by the following first evaluation function. A system. s = x1 × x2

11. In any one of claims 6 to 9, When the first evaluation index and the second evaluation index are set as xi, the estimation module includes a configuration for obtaining the probability of disease onset py* of the causative disease y* as the evaluation value by the following first evaluation function. A system. 【Number 1】 Set, wyi are the weight coefficients of the respective evaluation indices of each causative disease, and i is an integer.

12. In any one of claims 1 to 11, The estimation module includes a configuration for providing information as a stratification marker. A system.

13. In any one of claims 1 to 12, A system including a module for outputting the estimation and / or the process leading to the estimation of the estimation module, the first evaluation index, the second evaluation index, and other information to an output medium including a smartphone, a PC, a tablet, and paper.

14. In any one of claims 1 to 13, A system including a module for classifying a subject into a predetermined category based on the estimation and / or the process leading to the estimation of the estimation module, the first evaluation index, the second evaluation index, and other information.

15. A method for controlling a system, comprising: a first input module configured to obtain a first evaluation index based on data related to the physical state of a subject's brain; a second input module configured to obtain a second evaluation index which is twice the odds of a first causative disease based on data related to the function of the subject's brain; an estimation module configured to estimate the state of dementia of the subject; and wherein the control method comprises: the estimation module obtaining the first evaluation index and the second evaluation index via the first input module and the second input module; estimating the state of a brain disorder including dementia of the subject based on an evaluation value obtained by a first evaluation function having the first evaluation index and the second evaluation index as variables. A control method.

16. Obtaining a first evaluation index based on data related to the physical state of a subject's brain; Obtaining a second evaluation index which is twice the odds of a first causative disease based on data related to the function of the subject's brain; A computer program having instructions for a computer to execute estimating the state of a brain disorder including dementia of the subject based on an evaluation value obtained by a first evaluation function having the first evaluation index and the second evaluation index as variables.

Images