JP6782006B2 - Brain information analysis device and brain health index calculation device - Google Patents

Description

The present invention relates to a brain information analysis device and a brain health index calculation device.
More specifically, biological information is collected using brain image data output by a device that performs three-dimensional imaging of the human brain non-invasively, such as MRI (magnetic resonance imaging). The present invention relates to a brain information analysis device that performs analysis and a brain health index calculation device that estimates a value that serves as a health index of a subject using a feature amount output by the brain information analysis device.

Originally, MRI was a non-invasive tool for discovering and diagnosing visceral diseases including the brain. However, in recent years, it has been permanently performed to photograph the human brain with MRI and acquire three-dimensional brain image data, so that the brain image data of various people are gradually accumulated. It has become. Therefore, in recent years, brain scientists have come to search not only for brain diseases but also for the correlation between the information obtained from brain image data and human health.
Hereinafter, the information obtained from the brain image data in the present specification is collectively referred to as brain information. For example, the amount of gray matter in a specific part of the brain obtained as a result of image analysis of brain image data, the anisotropy of nerve fibers in a specific part of the brain, and the like.

In Patent Document 1, a gray matter image of a brain cross section is created from an captured MRI image, and the gray matter image is compared with a gray matter image of a healthy person to determine atrophy of a specific part of the brain. The technical contents for diagnosing signs such as illness are disclosed.

Japanese Unexamined Patent Publication No. 2005-230456

There are various methods for acquiring information from brain image data for each part of the brain. The information obtained by any brain information acquisition method is very detailed. However, this detailed information, that is, the large number of variables, does not fit into the so-called big data analysis method that statistically infers the overall tendency of data.
MRI equipment is expensive, and although the number of MRI equipment introduced in Japan is relatively large in international comparison, the number of MRI equipment introduced is not large compared to other medical equipment. Moreover, its operating cost is not cheap. Therefore, it is not easy to accumulate brain image data by MRI. Therefore, at present, the population parameter of the data that is the basis of statistical analysis is not very large. On the other hand, the number of information obtained from brain image data, that is, variables is large. This causes a phenomenon called overfitting, in which the statistical analysis method gives correct estimation results for known values, but does not give correct estimation results for unknown values.
In addition, a large number of variables indicates that the index of judgment becomes complicated. For this reason, the current detailed brain analysis information can only be understood by brain science experts, and is not suitable for use by the general public as an index of brain health.

An object of the present invention is to provide a brain information analysis device and a brain health index calculation device that can solve such a problem, convert brain analysis information into a value that is easy to handle, and use it as a health index.

In order to solve the above problems, the brain information analysis apparatus of the present invention includes an gray white mass calculation unit that obtains a plurality of gray white mass data from the brain image data of a subject who has acquired the brain image data, and an average value of the plurality of gray white mass data. The average value calculation unit that outputs GM-BHQ, which is the calculated value, the subject ID that uniquely identifies the subject who acquired the brain image data, the BHQ table that registers the GM-BHQ in association with each other, and the subject ID. Multivariate analysis is performed after linking the health check table, the BHQ table, and the health check table, which register the health check results of the subjects who have acquired the brain image data, with the subject ID, and for any subject. It includes a multivariate analysis processing unit that outputs a feature amount for estimating GM-BHQ.

INDUSTRIAL APPLICABILITY According to the present invention, it is possible to provide a brain information analysis device and a brain health index calculation device that can convert brain analysis information into a value that is easy to handle and use it as an index of health.
Issues, configurations and effects other than those described above will be clarified by the description of the following embodiments.

It is the schematic of the brain information analysis apparatus which concerns on embodiment of this invention, and is the schematic diagram of the brain health index calculation apparatus using the coefficient of the multiple regression equation calculated by the brain information analysis apparatus. It is a block diagram which shows the hardware structure of the brain information analysis apparatus. It is a block diagram which shows the software function of the brain information analysis apparatus. It is a block diagram which shows the software function of the brain health index arithmetic unit. It is a graph which shows the relationship between GM-BHQ and age calculated from the brain image data of a subject, and the graph which shows the relationship between FA-BHQ and age.

FIG. 1A is a schematic view of a brain information analysis device 101 according to an embodiment of the present invention.
The brain information analysis device 101 is a personal computer or a server having a large-capacity non-volatile storage and a predetermined computing power.
The brain information analysis device 101 describes (1) an MRI image file group 103 obtained by photographing the subject's brain with the MRI device 102, and (2) the results of the health diagnosis received by the subject. Data 104 and (3) living environment questionnaire result data 105, which describes the results of the subjects answering a predetermined living environment questionnaire, are input and stored as a database. The MRI image file group 103, the health examination result data 104, and the living environment questionnaire result data 105 are all linked by a subject ID that uniquely identifies the subject 106. In particular, in order to store the MRI image file group 103, the brain information analysis device 101 requires a large-capacity non-volatile storage.

The MRI image file group 103 is a plurality of image files in which the MRI apparatus 102 photographs the brain of the subject 106. A plurality of image files obtained by virtually slicing the brain of the subject 106 in a sliced state can be obtained from the crown to the neck. The brain of the subject 106 is analyzed three-dimensionally and spatially using the plurality of image files.

The health diagnosis result data 104 is a collection of information on the diagnosis results obtained in a normal health diagnosis. Subject 106's age, height, weight, eyesight, hearing, body fat percentage, BMI (Body Mass Index), triglyceride level in blood, γGTP (γ-glutamyltransferase), etc. once a year It is information on the diagnosis result of the health examination to be carried out. These data are objective measurement results obtained from the subject 106 by mechanical or chemical means, with no room for subject 106's subjectivity. In addition to the health examination, biometric data such as body weight and blood pressure measured daily from the subject 106 and life log data such as the number of steps and the amount of activity can be included in the health examination result data 104.

The living environment questionnaire result data 105 includes various questionnaires that are asked at the time of the health examination and answered by the subject 106, as well as a questionnaire for estimating the health index of the neuropsychiatric system. For example, a question is set for the following items, and the subject 106 is asked to answer.
(1) Awareness of hierarchical belonging and socio-economic conditions including financial instability (2) Awareness of health and security including a sense of satisfaction and improvement of life (3) Personal values, Post-materialism (the idea of prioritizing spiritual and mental well-being (richness) over material and economic wealth) and epicrossism (the principle of living in the present rather than preparing for the future)
That is, these data are different from the health diagnosis result data 104, and are information such as a questionnaire result in which the subjectivity of the subject 106 is mixed.

The brain information analysis device 101 first acquires brain information from the MRI image file group 103 by using a predetermined image analysis process. Next, the average value of this brain information is calculated. Then, the brain information analysis device 101 performs multiple regression analysis using the average value of the calculated brain information as the objective variable, the health examination result data 104 and the living environment questionnaire result data 105 as explanatory variables, and calculates the multiple regression equation coefficient 107. To do.

FIG. 1B is a schematic view of a brain health index calculation device 111 using the multiple regression equation coefficient 107 calculated by the brain information analysis device 101.
The brain health index calculation device 111 can be realized by a general personal computer. It can also be realized by portable wireless terminals such as smartphones, which have become widespread in recent years, and health appliances such as body fat scales that use one-chip microcomputers.
When the health diagnosis result data 104 and the living environment questionnaire result data 105 of any subject 106 are input to the brain health index calculation device 111, the multiple regression equation used by the brain information analysis device 101 and the brain information analysis device Using the multiple regression equation coefficient 107 calculated by 101, the average value of the brain information of the subject 106 is estimated as the diagnosis result 112.
That is, the average value of brain information is a value that is an index of human health. The average value of brain information will be described in detail in FIG. 3 and subsequent sections.

FIG. 2 is a block diagram showing a hardware configuration of the brain information analysis device 101.
In the brain information analysis device 101, which is a general personal computer or server, a CPU 201, a ROM 202, a RAM 203, a display unit 204 such as a liquid crystal display, an operation unit 205 such as a keyboard and a mouse, and a non-volatile storage 206 such as a hard disk device are connected to a bus 207. It is connected. In addition to this, the bus 207 receives the MRI image file group 103, the health examination result data 104, and the living environment questionnaire result data 105, and registers the serial port 208 in the database formed in the non-volatile storage 206. And NIC (Network Interface Card) 209 are connected. The non-volatile storage 206 stores an OS, a program for operating a personal computer or a server as a brain information analysis device 101, and various databases described later in FIG.
When the brain information analysis device 101 is a server, the NIC 209 is indispensable, but the display unit 204 and the operation unit 205 are not always necessary, and the serial port 208 is not necessary. In that case, a terminal that operates the server via the network may be prepared separately.
The block diagram of the brain information analysis device 101 shown in FIG. 2 is almost the same as that of the brain health index calculation device 111.

FIG. 3 is a block diagram showing a software function of the brain information analysis device 101.
As described with reference to FIG. 1A, the brain information analysis device 101 receives the MRI image file group 103, the health examination result data 104, and the living environment questionnaire result data 105.
The gray matter mass calculation unit 301 reads the MRI image file group 103 and calculates the gray matter mass data group 302 indicating the amount of gray matter in a specific part of the brain. The gray mass data group 302 is a collection of gray mass data at 116 sites in the brain.
The nerve fiber anisotropy calculation unit 303 reads the MRI image file group 103 and calculates the nerve fiber anisotropy data group 304 showing the index value of the anisotropy of the nerve fibers constituting the white matter in a specific part of the brain. .. The nerve fiber anisotropy data group 304 is a collection of index values of the anisotropy of nerve fibers constituting the white matter at 48 sites in the brain.

The gray white mass data group 302 calculated by the gray white mass calculation unit 301 and the nerve fiber anisotropy data group 304 calculated by the nerve fiber anisotropy calculation unit 303 are registered in the brain information table 306 together with the subject ID 305 of the subject 106. To.
The brain information table 306 comprises a subject ID field, 116 gray-white mass data fields, and 48 nerve fiber anisotropy data fields. That is, the gray-white mass data group 302 and the nerve fiber anisotropy data group 304 are associated with the subject ID 305 of the subject 106 and registered in the brain information table 306.
Since the MRI image file group 103 is a collection of valuable image data, it is also associated with the subject ID 305 and stored in the large-capacity non-volatile storage 206. In the embodiment of the present invention, the MRI image file group 103 will not be used in the following description, and therefore the illustration and details will be omitted.

The 116 gray-white mass data group 302 and the 48 nerve fiber anisotropy data group 304 stored in the brain information table 306 are input to the mean value calculation unit 307, respectively.
The average value calculation unit 307 calculates the average value of 116 gray-white mass data groups 302. Hereinafter, the average value of the gray matter mass data group 302 will be referred to as GM-BHQ (Grey-Matter Brain Healthcare Quotient). GM-BHQ is a single scalar value.
Similarly, the average value calculation unit 307 calculates the average value of the 48 nerve fiber anisotropy data groups 304. Hereinafter, the average value of the nerve fiber anisotropy data group 304 is referred to as FA-BHQ (fractional anisotropy Brain Healthcare Quotient). FA-BHQ, like GM-BHQ, has a single scalar value.

The GM-BHQ and FA-BHQ calculated by the average value calculation unit 307 are registered in the BHQ table 308 together with the subject ID 305 of the subject 106.
The BHQ table 308 includes a subject ID field, a GM-BHQ field, and an FA-BHQ field. That is, the GM-BHQ and the FA-BHQ are associated with the subject ID 305 of the subject 106 and registered in the BHQ table 308.
These GM-BHQ and FA-BHQ are objective variables in the multiple regression equation described later.

On the other hand, the health diagnosis result data 104 and the living environment questionnaire result data 105 are registered in the health diagnosis table 309 together with the subject ID 305 of the subject 106.
The health examination table 309 includes a subject ID field, a health examination result data field for each item, and a living environment questionnaire result data field for each item. That is, the health diagnosis result data 104 and the living environment questionnaire result data 105 are associated with the subject ID 305 of the subject 106 and registered in the health diagnosis table 309.
These health diagnosis result data 104 and living environment questionnaire result data 105 serve as explanatory variables in the multiple regression equation described later.

The multiple regression analysis processing unit 310 reads the objective variable GM-BHQ from the BHQ table 308, and the explanatory variables health diagnosis result data 104 and living environment questionnaire result data 105 from the health diagnosis table 309, respectively, with the subject ID 305. After associating, the coefficient of the multiple regression equation in GM-BHQ is calculated, and the multiple regression equation coefficient 311 for GM-BHQ, which is the multiple regression equation coefficient for estimating GM-BHQ, is output.
Similarly, the multiple regression analysis processing unit 310 reads the FA-BHQ objective variable from the BHQ table 308, and the health diagnosis result data 104 and the living environment questionnaire result data 105, which are explanatory variables, from the health check table 309, respectively. After associating with the subject ID 305, the coefficient of the multiple regression equation in FA-BHQ is calculated, and the multiple regression equation coefficient 312 for FA-BHQ, which is the multiple regression equation coefficient for estimating FA-BHQ, is output.
These multiple regression equation coefficients 311 for GM-BHQ and multiple regression equation coefficients 312 for FA-BHQ are the multiple regression equation coefficients 107 described in FIGS. 1A and 1B.

FIG. 4 is a block diagram showing a software function of the brain health index arithmetic unit 111.
When the multiple regression calculation processing unit 401 reads the health diagnosis result data 104 of any subject 106 and the living environment questionnaire result data 105, it is for GM-BHQ output by the brain information analysis device 101 described in FIG. Using the multiple regression equation coefficient 311, the multiple regression equation used by the multiple regression analysis processing unit 310 of the brain information analysis device 101 is calculated, and the GM-BHQ estimated value 402 of the subject 106 is output.
Similarly, when the multiple regression calculation processing unit 401 reads the health diagnosis result data 104 of any subject 106 and the living environment questionnaire result data 105, the FA output by the brain information analysis device 101 described in FIG. 3 -Using the multiple regression equation coefficient 312 for BHQ, the multiple regression equation used in the multiple regression analysis processing unit 310 of the brain information analysis device 101 is calculated, and the FA-BHQ estimated value 403 of the subject 106 is output.
These GM-BHQ estimated value 402 and FA-BHQ estimated value 403 are the diagnostic results 112 described in FIG. 1B.

In FIG. 1, the diagnosis result 112 is written for easy understanding, but in reality, the “estimated brain health index” that estimates the brain health index of the subject 106 is more accurate. If the health diagnosis data and the questionnaire result are input to the brain health index calculation device 111, the brain health state of the subject 106 can be estimated without using the MRI device 102.

Of course, the health examination result data 104 read by the multiple regression arithmetic processing unit 401 and the living environment questionnaire result data 105 have the same contents as the health examination and questionnaire read by the brain information analysis device 101 and received by the subjects 106. There is a need.
That is, the main arithmetic processing of the brain health index arithmetic unit 111 is only the calculation of the multiple regression equation which is a linear function. There is no large amount of data processing or the like in this arithmetic processing. As is well known, since the calculation of the multiple regression equation can be realized by four arithmetic operations, even a computer resource called a low-resource device having poor computing power such as a one-chip microcomputer can be sufficiently put into practical use.

For example, the brain health index calculation device 111 according to the embodiment of the present invention is used as a body composition meter capable of measuring body weight, body fat percentage, visceral fat level, subcutaneous fat percentage, basal metabolism, skeletal muscle percentage, BMI, body age, and the like. Think about applying. It is possible to register the date of birth and height of the user in the body composition monitor. However, it is troublesome to register the values obtained by the blood test in the health examination, and it is not realistic because the measurement date and time are different from the various values obtained by the body composition meter used every day. Therefore, the brain information analysis device 101 calculates the GM-BHQ multiple regression equation coefficient 311 and the FA-BHQ multiple regression equation coefficient 312 using only the information that can be acquired by the body composition meter. By using the GM-BHQ multiple regression equation coefficient 311 for the body composition meter and the FA-BHQ multiple regression equation coefficient 312, the body composition meter can also function as the brain health index calculation device 111. Although the estimation accuracy is lower than when all the information such as the results of the health examination and the living environment questionnaire is available, the body composition monitor uses the GM-BHQ estimated value 402 and the FA-BHQ estimated value 403 as significant health indicators. It becomes possible to output.

In addition, the inventors performed multiple regression analysis using the GM-BHQ and FA-BHQ of the subject 106 registered in the brain information analysis device 101 as explanatory variables and the age of the subject 106 as the objective variable. Good results were also obtained. From this, when the brain image data of a certain subject 106 is obtained, it becomes possible to estimate the "age of the brain" by multiple regression analysis.

FIG. 5 is a graph showing the relationship between GM-BHQ and age calculated from the brain image data of the subject 106, and a graph showing the relationship between FA-BHQ and age. In the graph, the horizontal axis is age, and the vertical axis is GM-BHQ and FA-BHQ. These graphs are the results of analyzing the brain image data of a large number of subjects 106, which have led the inventors to complete the brain information analysis device 101 and the brain health index calculation device 111 according to the embodiment of the present invention.
The subject 106, which is the basis of all the data, is 144 healthy subjects consisting of 80 males and 64 females, aged 26 to 69 years. The average age is 48.5 years and the standard deviation of age is 8.05.
The research relating to the present invention was approved by the ethics committees of RIKEN, Kyoto University, and the University of Tokyo, and was carried out in accordance with the rules and guidelines of these research institutes. Written consent was obtained from all participants prior to participation.
All MRI image data were collected at RIKEN, Kyoto University and the University of Tokyo using a 3T Siemens scanner with a 32-channel head array coil.

In order to acquire the gray matter mass data that is the basis of GM-BHQ, a three-dimensional magnetization ready high-speed gradient echo (MPRAGE) image was acquired using structural MRI (sMRI). Only gray matter was extracted from the image data by specifying the pixel density range, and standardized by DARTEL (diffeomorphic anatomical registration through exponentiated Lie algebra). Then, after the obtained image was smoothed by integration, the individual difference in the volume of the brain was canceled by dividing by the intracranial volume of the subject 106, and the final gray-white mass data was acquired.

In order to acquire the neurofiber anisotropy data that is the basis of FA-BHQ, the diffusion tensor data that is the basis of diffusion tensor imaging (DTI) is used as SE-EPI (Spin-Echo Echo-Planner Imaging). Was obtained using. FA (Fractional anisotropy: an index showing the degree of anisotropy of diffusion) was obtained from DTI, and final nerve fiber anisotropy data was obtained.

Looking at the graph of GM-BHQ, it can be seen that age and GM-BHQ have a relatively steep negative proportional relationship. The correlation coefficient R of GM-BHQ was 0.61. The broken line L501 in the graph is a line of a linear function composed of a multiple regression equation by multiple regression analysis. The slope of the broken line L501 is −0.6, and the y-intercept is 131. This line corresponds to the GM-BHQ estimate 402 for age.
Looking at the graph of FA-BHQ, it can be seen that age and FA-BHQ have a relatively loose negative proportional relationship. The correlation coefficient R of FA-BHQ was 0.42. The broken line L502 in the graph is a line of a linear function composed of a multiple regression equation by multiple regression analysis. The slope of the broken line L502 is −0.2, and the y-intercept is 109. This line corresponds to the FA-BHQ estimate 403 for age.
As can be seen from the graphs in FIG. 5, GM-BHQ and FA-BHQ have a significant correlation with the age of the subject 106.
Roughly speaking, it can be seen from this graph that the gray mass of the brain decreases with age in humans. In addition, it can be seen from this graph that the density of nerve fibers contained in the white matter of the brain decreases with age.

Conventionally, in brain science, 116 gray-white mass data and 48 nerve fiber anisotropy data have been individually studied. However, if the number of samples of the underlying brain image data is small and the number of variables is so large, it is not suitable for the statistical analysis method.
The inventors have tried and errored a method of reducing the number of these huge variables. As a result, by making it a simple mean value, the variability generated from many underlying variables is well leveled, and it can be used as easy-to-use information indicating human physical and mental health. understood.
Simple multiple regression analysis became available by converting 116 gray-white mass data and 48 nerve fiber anisotropy data as mean values, each of which is a single scalar value.

The brain information analysis device 101 and the brain health index calculation device 111 according to the embodiment of the present invention employ multiple regression analysis as multivariate analysis, but as is well known, multivariate analysis is not limited to multiple regression analysis. Various multivariate analyzes such as logistic regression, principal component analysis, Bayesian inference, and support vector machine can be applied. However, it is necessary to use the same multivariate analysis adopted by the brain information analysis device 101 and the multivariate estimation adopted by the brain health index calculation device 111. For example, in the case of Bayesian estimation, it is necessary for the brain health index calculation device 111 to have the data set created by the brain information analysis device 101.
That is, the superordinate concept of the multiple regression analysis processing unit 310 in the brain information analysis device 101 is the multivariate analysis processing unit, and the superordinate concept of the multiple regression equation coefficient output by the multivariate analysis processing unit is the feature quantity, which is a brain health index. The superordinate concept of the multiple regression arithmetic processing unit 401 in the arithmetic unit 111 is the multivariate estimation processing unit.

In the embodiment of the present invention, the brain information analysis device 101 and the brain health index calculation device 111 have been described.
The brain information analysis device 101 acquires 116 gray-white mass data and 48 nerve fiber anisotropy data from the MRI image file group 103 of the subject 106. Then, using GM-BHQ for which the average value of the grayish-white mass data was calculated and FA-BHQ for which the average value of the nerve fiber anisotropy data was calculated as objective variables, the results of the health examination and the living environment questionnaire of the subject 106 were used as explanatory variables. As a result, multiple regression analysis is performed and the coefficient is calculated. The brain health index calculation device 111 can estimate GM-BHQ and FA-BHQ, which are indicators of health, by using this coefficient.

Although the embodiments of the present invention have been described above, the present invention is not limited to the above embodiments, and other modifications and applications are provided as long as they do not deviate from the gist of the present invention described in the claims. including.
For example, the above-described embodiment describes in detail and concretely the configurations of the apparatus and the system in order to explain the present invention in an easy-to-understand manner, and is not necessarily limited to those including all the described configurations. Further, it is possible to replace a part of the configuration of one embodiment with the configuration of another embodiment, and further, it is possible to add the configuration of another embodiment to the configuration of one embodiment. Further, it is also possible to add / delete / replace a part of the configuration of each embodiment with another configuration.
In addition, each of the above configurations, functions, processing units, and the like may be realized by hardware, for example, by designing a part or all of them with an integrated circuit. In addition, each of the above configurations, functions, and the like may be realized by software for the processor to interpret and execute a program that realizes each function. Information such as programs, tables, and files that realize each function should be stored in memory, volatile or non-volatile storage such as hard disks and SSDs (Solid State Drives), or recording media such as IC cards and optical disks. Can be done.
In addition, control lines and information lines are shown as necessary for explanation, and not all control lines and information lines are shown in the product. In practice, it can be considered that almost all configurations are interconnected.

101 ... Brain information analysis device, 102 ... MRI device, 103 ... MRI image file group, 104 ... Health diagnosis result data, 105 ... Living environment questionnaire result data, 106 ... Subject, 107 ... Multiple regression coefficient, 111 ... Brain health index Arithmetic logic unit, 112 ... Diagnosis result, 201 ... CPU, 202 ... ROM, 203 ... RAM, 204 ... Display unit, 205 ... Operation unit, 206 ... Non-volatile storage, 207 ... Bus, 208 ... Serial port, 209 ... NIC, 301 ... Gray white mass calculation unit, 302 ... Gray white mass data group, 303 ... Neurofiber anisotropy calculation unit, 304 ... Neurofiber anisotropy data group, 305 ... Subject ID, 306 ... Brain information table, 307 ... Average value calculation unit , 308 ... BHQ table, 309 ... Health diagnosis table, 310 ... Multiple regression analysis processing unit, 311 ... Multiple regression equation coefficient for GM-BHQ, 312 ... Multiple regression equation coefficient for FA-BHQ, 401 ... Multiple regression arithmetic processing unit , 402 ... GM-BHQ estimated value, 403 ... FA-BHQ estimated value

Claims (6)

And gray mass calculation unit for obtaining a plurality of gray mass data from the brain image data of a subject acquired brain image data,
An average value calculation unit that outputs GM-BHQ, which is a value obtained by calculating the average value of a plurality of the gray and white mass data, and
A subject ID that uniquely identifies the subject from whom the brain image data has been acquired, a BHQ table that associates and registers the GM-BHQ, and a BHQ table.
A health diagnosis table for registering the subject ID and the health diagnosis result of the subject who acquired the brain image data in association with each other.
A multivariate analysis processing unit that links the BHQ table and the health examination table with the subject ID, performs multivariate analysis, and outputs a feature amount for estimating the GM-BHQ for an arbitrary subject. A brain information analyzer equipped with. In addition
; And a nerve fiber anisotropic calculating portion for obtaining a plurality of nerve fibers anisotropic data from the brain image data of a subject acquired the brain image data,
The average value calculation unit also outputs FA-BHQ, which is a value obtained by calculating the average value of a plurality of the nerve fiber anisotropy data.
In the BHQ table, the FA-BHQ is also registered in association with the subject ID.
The multivariate analysis processing unit also outputs a feature amount for estimating the FA-BHQ.
The brain information analysis device according to claim 1. The multivariate analysis processing unit performs multiple regression analysis on the multivariate analysis, and has a multiple regression equation coefficient for GM-BHQ for estimating the GM-BHQ and a FA-BHQ for estimating the FA-BHQ. Outputs multiple regression coefficients,
The brain information analysis device according to claim 2. The subject is uniquely composed of a gray white mass calculation unit that obtains a plurality of gray white mass data from the brain image data of the subject, and an average value calculation unit that outputs a GM-BHQ that is a value obtained by calculating the average value of the plurality of gray white mass data. A BHQ table for registering the subject ID to be identified and the GM-BHQ in association with each other, a health examination table for registering the subject ID and the health examination result of the subject in association with each other, the BHQ table and the health examination table. The feature output by a brain information analysis device including a multivariate analysis processing unit that performs multivariate analysis after associating with the subject ID and outputs a feature amount for estimating the GM-BHQ. A brain health index calculation device including a multivariate estimation processing unit that estimates the health diagnosis result of an arbitrary subject based on multivariate analysis and outputs a GM-BHQ estimated value using the amount. The brain information analysis device further
It is equipped with a nerve fiber anisotropy calculation unit that obtains a plurality of nerve fiber anisotropy data from the brain image data of the subject.
The average value calculation unit also outputs FA-BHQ, which is a value obtained by calculating the average value of a plurality of the nerve fiber anisotropy data.
In the BHQ table, the FA-BHQ is also registered in association with the subject ID.
The multivariate analysis processing unit also outputs a feature amount for estimating the FA-BHQ.
The multivariate estimation processing unit estimates the health diagnosis result of the arbitrary subject based on the multivariate analysis, and also outputs the FA-BHQ estimated value.
The brain health index calculation device according to claim 4. The multivariate analysis processing unit performs multiple regression analysis on the multivariate analysis, and has a multiple regression equation coefficient for GM-BHQ for estimating the GM-BHQ and a FA-BHQ for estimating the FA-BHQ. It outputs multiple regression equation coefficients and outputs multiple regression equation coefficients.
The multivariate estimation processing unit calculates the multiple regression equation based on the multiple regression equation coefficient for GM-BHQ and the multiple regression equation coefficient for FA-BHQ on the health diagnosis result of the arbitrary subject, and GM- It outputs the BHQ estimated value and the FA-BHQ estimated value.
The brain health index calculation device according to claim 5.

Images