Cognitive function estimation device and program

Abstract

To estimate cognitive function with high accuracy. [Solution] The cognitive function estimation device 10's physical data acquisition unit 102 acquires physical data indicating the subject's age and gender. The odor selection unit 104 uses the correlation / average information database 100 to select odor components corresponding to the subject's physical data. The olfactory data collection unit 101 releases the odor components selected by the odor selection unit 104 to the subject and collects an olfactory score indicating the subject's olfactory ability. The cognitive function evaluation unit 108 acquires the average value of olfactory scores in groups to which the subject belongs, for example, by age and gender, and derives an evaluation value of the subject's cognitive function. The presentation unit 110 presents information regarding cognitive function. In this way, the cognitive function estimation device 10 estimates cognitive function using the correlation / average information database 100.

Description

Technical Field

[0006] , , ,

[0004] , , , , , , , , , , ,

[0005]

[0001] The present disclosure relates to a cognitive function estimation device and a program.

Background Art

[0002] Conventionally, techniques related to the measurement of olfactory ability have been known. In Patent Document 1, a plurality of images including correct answers are displayed, odor components are released, the selection of a subject is received, and the olfactory age is derived based on whether the correct answer and the selection result match. Further, in Non-Patent Document 1, a risk of dementia onset is determined using 10 odor cards closely related to brain function that present odors extracted from the correlation with MMSE (cognitive function examination).

Prior Art Documents

Non-Patent Documents

[0003]

Non-Patent Document 1

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] In the technique of Patent Document 1 described above, odor components are released, and the olfactory age of the subject is derived from the response. The technique of Patent Document 1 is a technique based on the medical knowledge that olfactory ability declines with aging, and there is insufficient consideration regarding the risk of dementia such as a decline in cognitive function occurring with respect to olfactory ability. There is room for improvement when estimating cognitive function.

[0006] Furthermore, in the technology described in Non-Patent Document 1, since the same 10 types of scent cards are used for each of the multiple subjects, it is difficult to improve the accuracy of estimating the cognitive ability of subjects whose correlation with scents changes depending on their age and gender, and there is room for improvement when estimating cognitive function.

[0007] This disclosure is made in consideration of the above facts and aims to provide a cognitive function estimation device and program that can estimate cognitive function with high accuracy through a simple olfactory test. [Means for solving the problem]

[0008] A first aspect of this disclosure is a cognitive function estimation device comprising: an acquisition unit that acquires identification information including the age of a subject for identifying a subject for whom cognitive function is to be estimated; an odor component setting unit that sets odor components having a correlation with the subject based on the identification information including the age for identifying a subject who has undergone a predetermined cognitive function test, based on correlation information including information showing the correlation between odor components and cognitive function tests, and the correlation between the odor components and the test values ​​of the cognitive function tests; and an estimation unit that estimates an evaluation value of cognitive function for the subject based on the results of the odor component test by the subject's sense of smell when the set odor components are released and the test values ​​of the cognitive function tests in the correlation information.

[0009] The second embodiment is a cognitive function estimation device of the first embodiment, wherein the correlation information derives the correlation between the type of odor component and the test value of the odor component for multiple subjects, classifies them into age groups based on predetermined age boundaries, and tests for odor components are performed according to the classified age groups. The average of the results is derived as the test value for the odor component.

[0010] A third embodiment is a cognitive function estimation device according to the first or second embodiment, further comprising a presentation unit that presents an evaluation value of cognitive function for the subject estimated by the estimation unit, wherein the estimation unit presents recommendation information recommending that the subject undergo the cognitive function test if the evaluation value of cognitive function exceeds a predetermined threshold.

[0011] The fourth embodiment is a cognitive function estimation device according to any one of the first to third embodiments, wherein the identification information includes at least one piece of information: physical data including the age and gender of the subject, smoking data including whether or not they have a history of smoking, stay data indicating their location and duration of stay, and education involvement time data indicating their involvement in education.

[0012] The fifth embodiment is a cognitive function estimation device according to the fourth embodiment, wherein the gender of the physical data indicates one of two types, a first gender and a second gender, and the odor component setting unit sets odor components including vanilla, rag odor, yellow peach, and socks for the first gender, and sets odor components including caramel and socks for the second gender.

[0013] The sixth aspect is a program that causes a computer to function as an acquisition unit that acquires identification information, including the age of a subject, for identifying a subject whose cognitive function is to be estimated; an odor component setting unit that sets odor components that have a correlation with the subject, based on the identification information, including the age, for identifying a subject who has undergone a predetermined cognitive function test, and correlation information, which includes information showing the correlation between odor components and cognitive function tests, and the correlation between the odor components and the test values ​​of the cognitive function tests; and an estimation unit that estimates an evaluation value of cognitive function for the subject, based on the results of the odor component test by the subject's sense of smell when the set odor components are released and the test values ​​of the cognitive function tests in the correlation information.

[0014] The seventh embodiment is a cognitive function estimation device comprising: an odor component emission unit that emits each of a plurality of odor components to a subject, with the concentration of each odor component changing in multiple stages; a score setting unit that sets a score corresponding to the concentration indicated when the subject's answer indicating which odor component was emitted from the odor component emission unit is correct; and an estimation unit that estimates a cognitive function evaluation value for the subject based on the sum of the scores set by the score setting unit for the plurality of odor components.

[0015] The eighth aspect is the cognitive function estimation device of the seventh aspect, wherein the odor components are the respective odor components of vanilla, mint, yellow peach, and socks.

[0016] The ninth aspect is the cognitive function estimation device of the seventh aspect, The estimation unit uses constants a and b, Cognitive function assessment value = a × sum of scores + b The cognitive function evaluation value based on the formula represented by is estimated as the aforementioned cognitive function.

[0017] The tenth embodiment is a cognitive function estimation device according to any one of the seventh to ninth embodiments, in which the score is set to a higher value as the concentration of correct answers decreases.

[0018] The 11th embodiment is a cognitive function estimation device of the 7th embodiment, further comprising a scent component setting unit that sets a group of scent components that have a correlation with the cognitive function test and have fewer scent components than the number of scent components, based on a correspondence relationship between the test values ​​of the cognitive function test performed on a predetermined cognitive function test on a plurality of subjects and the scores of each of the multiple scent components released to the plurality of subjects, and the scent component release unit uses the group of scent components set by the scent component setting unit as the plurality of scent components. And release.

[0019] Aspect 12 is the cognitive function estimation device according to Aspect 11, wherein the odor component setting unit analyzes the distribution of subjects whose test values of the cognitive function test exceed a predetermined test value and subjects whose test values are below the predetermined test value with respect to the total value of the plurality of odor components for the plurality of subjects, and sets a plurality of odor components in a combination that exceeds a predetermined correlation value regarding the discrimination power evaluation value as the odor component group.

[0020] Aspect 13 is the cognitive function estimation device according to Aspect 7, wherein the estimation unit estimates an evaluation value of the cognitive function for the subject based on the correlation relationship between the total value of the scores of the odor components for a plurality of subjects and the ratio of subjects whose test values of the cognitive function test, which has been performed in advance on the plurality of subjects, exceed a predetermined test value and subjects whose test values are below the predetermined test value.

[0021] Aspect 14 is the cognitive function estimation device according to any one of Aspects 7 to 13, further comprising a presentation unit that presents an evaluation value of the cognitive function for the subject estimated by the estimation unit, and the estimation unit presents recommendation information recommending that the subject take the cognitive function test when the evaluation value of the cognitive function exceeds a predetermined threshold value.

[0022] Aspect 15 is a program for causing a computer to execute a process of estimating the cognitive function of a subject, the program causing the computer to function as a score setting unit that sets a score corresponding to a concentration indicating a correct answer when an answer of the subject indicating which odor component is the odor component emitted from an odor component emission unit that emits each of a plurality of odor components for the subject by changing the concentration of each odor component in a plurality of steps is correct, and an estimation unit that estimates an evaluation value of the cognitive function for the subject based on the total value of the scores of the plurality of odor components set by the score setting unit.

[0023] The 16th aspect is a cognitive function estimation device including an acquisition unit that acquires identification information including physical data indicating the age and gender of the subject for estimating the cognitive function and smoking data indicating the presence or absence of smoking experience, an odor component emission unit that emits each of a plurality of odor components to the subject by changing the concentration of each odor component in a plurality of levels, a score setting unit that sets a score corresponding to the concentration indicating the correct answer when the answer of the subject indicating which odor component is the odor component emitted from the odor component emission unit is correct, and an estimation unit that estimates an evaluation value of the cognitive function for the subject based on the identification information acquired by the acquisition unit and the total value of the scores for the plurality of odor components set by the score setting unit.

[0024] The 17th aspect is the cognitive function estimation device according to the 16th aspect, wherein the estimation unit estimates an evaluation value of the cognitive function for the subject based on a correlation relationship between the identification information, the type of the odor component and the test value of the odor component, and the test value of the cognitive function test for a plurality of subjects.

[0025] The 18th aspect is the cognitive function estimation device according to the 17th aspect, wherein the estimation unit sets A, B, C, D as constants, sets pa as (A × total value of scores), sets pb as (B × age), sets pc as (C × gender), sets pd as (D × smoking history), and uses the estimation formula represented by p = 1 / [1 + exp{-(pa + pb + pc + pd)}] to derive an evaluation value of the cognitive function.

[0026] The 19th embodiment is a cognitive function estimation device of the 16th embodiment, further comprising a scent component setting unit that sets a group of scent components that have a correlation with the cognitive function test and have fewer scent components than the number of scent components, based on a correspondence relationship that associates the test values ​​of the cognitive function test performed on a predetermined cognitive function test on a plurality of subjects with the scores of each of the plurality of scent components released to the plurality of subjects, and the scent component release unit releases the group of scent components set by the scent component setting unit as the plurality of scent components.

[0027] The 20th embodiment is a cognitive function estimation device according to the 19th embodiment, wherein the odor component setting unit sets a group of odor components as the odor component group, in which the information regarding the discriminative evaluation value obtained by analyzing the distribution of subjects whose cognitive function test results exceed a predetermined test value and subjects whose cognitive function test results are below a predetermined test value exceeds a predetermined correlation value, based on the total value of the multiple odor components for the multiple subjects.

[0028] The 21st embodiment is a cognitive function estimation device according to the 16th embodiment, wherein the odor component emitted from the odor component emission unit includes at least a set of components consisting of a first odor component predetermined as an odor component that gives the subject a feeling of freshness, and a second odor component predetermined as an odor component that gives the subject a feeling of discomfort.

[0029] The 22nd embodiment is a cognitive function estimation device of the 21st embodiment, wherein the first odor component is mint, and the second odor component is an odor component that smells like a dirty rag or an odor component that smells like socks.

[0030] The 23rd embodiment is a program for causing a computer to function as an acquisition unit that acquires identification information including physical data indicating the age and gender of a subject for identifying a subject whose cognitive function is to be estimated, and smoking data indicating whether or not the subject has a history of smoking; a component emission control unit that controls an odor component emission unit that emits each of a plurality of odor components to the subject, changing the concentration of each odor component in multiple stages; a score setting unit that sets a score corresponding to the concentration indicated when the subject's answer indicating which odor component was emitted from the odor component emission unit is correct; and an estimation unit that estimates an evaluation value of the cognitive function of the subject based on the identification information acquired by the acquisition unit and the sum of the scores set by the score setting unit for the plurality of odor components. [Effects of the Invention]

[0031] According to this disclosure, the effect is that cognitive function can be estimated with high accuracy by a simple olfactory test. [Brief explanation of the drawing]

[0032] [Figure 1] This block diagram shows the overall schematic configuration of the cognitive function estimation system according to the first embodiment. [Figure 2] This block diagram shows the hardware configuration of the cognitive function estimation device according to the first embodiment. [Figure 3] This is a block diagram showing an example of the functional configuration in the learning phase of the cognitive function estimation device according to the first embodiment. [Figure 4] This is a flowchart showing the learning process flow of the cognitive function estimation device according to the first embodiment. [Figure 5] This figure shows an example of a selection screen according to the first embodiment. [Figure 6] This figure shows an example of an olfactory score according to the first embodiment. [Figure 7] This figure shows an example of the correlation value regarding odor components with respect to the cognitive function test according to the first embodiment. [Figure 8]This figure shows an example of the correlation value regarding odor components with other cognitive function tests according to the first embodiment. [Figure 9] This figure shows an example of the average olfactory score according to the first embodiment. [Figure 10] This block diagram shows an example of the functional configuration in the estimation phase of the cognitive function estimation device according to the first embodiment. [Figure 11] A flowchart showing the estimation process flow of the cognitive function estimation device according to the first embodiment is provided. [Figure 12] This block diagram shows the overall schematic configuration of the cognitive function estimation system according to the second embodiment. [Figure 13] This is a block diagram showing the functional configuration in the learning phase of the cognitive function estimation device according to the second embodiment. [Figure 14] This flowchart shows the learning process flow of the cognitive function estimation device according to the second embodiment. [Figure 15] This figure shows an example of a data table that associates olfactory scores for each odor component with the results of cognitive function tests according to the second embodiment. [Figure 16] This figure shows an example of data illustrating the evaluation indicators according to the second embodiment. [Figure 17] This figure shows an example of the correspondence between olfactory scores and the number of people according to the second embodiment. [Figure 18] This figure shows an example of the relationship between the odor test results and the cognitive function test results according to the second embodiment. [Figure 19] This is a block diagram showing the functional configuration in the estimation phase of the cognitive function estimation device according to the second embodiment. [Figure 20] This flowchart shows the estimation process flow of the cognitive function estimation device according to the second embodiment. [Figure 21] This block diagram shows the overall schematic configuration of the cognitive function estimation system according to the third embodiment. [Figure 22] This is a block diagram showing an example of the functional configuration of a cognitive function estimation device according to the third embodiment. [Figure 23]A flowchart illustrating the learning process flow of the cognitive function estimation device according to the third embodiment is provided. [Figure 24] This figure shows an example of an olfactory score according to the third embodiment. [Figure 25] This figure shows an example of an ROC curve according to the third embodiment. [Figure 26] This figure shows an example of the correlation value regarding odor components for a cognitive function test according to the third embodiment. [Figure 27] This figure shows an example of the correlation value regarding odor components for a cognitive function test according to the third embodiment. [Figure 28] This block diagram shows another example of the functional configuration of the cognitive function estimation device according to the third embodiment. [Figure 29] A flowchart showing the estimation process flow of the cognitive function estimation device according to the third embodiment is provided. [Modes for carrying out the invention]

[0033] Hereinafter, an example of an embodiment for carrying out the technology of this disclosure will be described in detail with reference to the drawings. Components and processes that perform the same operation, action, or function are given the same reference numerals throughout the drawings, and redundant explanations may be omitted. Each drawing is only a schematic representation to the extent that the technology of this disclosure can be fully understood. Therefore, the technology of this disclosure is not limited to the illustrated examples. Furthermore, in this embodiment, explanations of configurations not directly related to the present invention or well-known configurations may be omitted.

[0034] [First Embodiment]

[0035] <Configuration of the Cognitive Function Estimation System> Figure 1 is a block diagram showing the overall schematic configuration of the cognitive function estimation system. As shown in Figure 1, the cognitive function estimation system according to this embodiment includes a cognitive function estimation device 10 that estimates cognitive function from information obtained by the emission of odor components by the odor component emission device 20. The cognitive function estimation device 10 also estimates the cognitive function of a subject using a correlation and average information database 100 that stores identification information including the age of the subject who underwent a cognitive function test, the types of odor components that correlate with the cognitive function test, and correlation information showing the correlation between the test values ​​of the odor components.

[0036] The identification information is data including age, etc., used to identify individuals whose cognitive function is estimated through a simple olfactory test, and physical data including age and gender of the individuals whose cognitive function is estimated can be applied. In this embodiment, the case in which physical data is applied as identification information will be described. The identification information is not limited to physical data including age and gender, and may include at least one piece of information such as smoking data including whether or not the individual has smoked, accommodation data indicating the location and duration of stay where the individual resided, etc., and educational participation time data in which the individual was involved in education. The data on the time spent in education by the subjects mentioned above includes, for example, the period during which the subject received education (e.g., number of years) and the period during which the subject was present in an educational setting (e.g., number of years). The period during which the subject was present in an educational setting includes the period during which the subject led the education (e.g., the period of guidance as a teacher or leader) and the period during which the subject acted as a leader during education.

[0037] Referring to Figure 2, the electrical configuration of the cognitive function estimation device 10 according to this embodiment will be described. Figure 2 is a block diagram showing an example of the hardware configuration of the cognitive function estimation device 10. As shown in Figure 2, the cognitive function estimation device 10 includes a computer main unit 10X. The computer main unit 10X includes a CPU (Central Processing Unit) 11, RAM (Random Access Memory) 12, ROM (Read Only Memory) 13, and an input / output interface (I / O) 14. The CPU 11, RAM 12, ROM 13, and I / O 14 are connected to a bus 15 so that they can communicate with each other. An input unit 16, a display unit 17, a communication unit 18, and a storage unit 19 are connected to the I / O 14.

[0038] The CPU 11 is a central processing unit that executes various programs and controls various parts. Specifically, the CPU 11 reads a program from the ROM 13 or memory unit 19 and executes the program using the RAM 12 as a working area. The CPU 11 controls each of the above components and performs various calculations according to the program stored in the ROM 13 or memory unit 19. In this embodiment, the ROM 13 or memory unit 19 stores an estimation processing program for executing cognitive function estimation processing and a learning processing program for constructing a correlation / average information database 100 used for cognitive function estimation processing. In this embodiment, the case in which the estimation processing program and the learning processing program are stored in the ROM 13 will be described as an example.

[0039] Furthermore, ROM 13 stores various programs and data. RAM 12 temporarily stores programs or data as a working area. The storage unit 19 functions as storage and is composed of an HDD (Hard Disk Drive) or SSD (Solid State Drive), and stores various programs, including the operating system, and various data.

[0040] The input unit 16 includes a pointing device such as a mouse, a keyboard, and an audio input device, and is used for various types of input.

[0041] The display unit 17 is, for example, a liquid crystal display or a speaker, and displays or plays various types of information. The display unit 17 may also function as an input unit 16 by employing a touch panel system.

[0042] The communication unit 18 is an interface for communicating with other devices such as the odor component emission device 20, and interfaces conforming to standards such as Ethernet®, FDDI, Wi-Fi®, and LTE are used.

[0043] Next, the functional configuration of the cognitive function estimation device 10 will be described. In this embodiment, the cognitive function estimation device 10 exhibits different functional configurations in a learning phase in which a correlation and average information database 100 used for cognitive function estimation processing is constructed, and in an estimation phase in which cognitive function estimation processing is performed using the constructed correlation and average information database 100.

[0044] (Learning Phase) First, let's describe the cognitive function estimation device 10, which functions as a learning phase. Figure 3 is a block diagram showing the functional configuration in the learning phase of the cognitive function estimation device 10, where a correlation and average information database 100 is constructed.

[0045] The cognitive function estimation device 10 has a function to construct a correlation and average information database 100 as a learning phase by executing a learning processing program. Specifically, the learning processing program has a function to collect olfactory data, a function to acquire physical data, a function to acquire cognitive function test results, a function to derive correlation and average information, and a registration function. In this learning phase, the cognitive function estimation device 10 registers identification information including the subject's age, the types of odor components that correlate with a predetermined cognitive function test, and correlation information showing the correlation between the odor components and the test values ​​of the cognitive function test in the correlation and average information database 100.

[0046] As shown in Figure 3, during the learning phase, the CPU 11 of the cognitive function estimation device 10 functions as an olfactory data acquisition unit 101, a physical data acquisition unit 102, a cognitive function test result acquisition unit 103, a correlation / average information derivation unit 105, and a registration unit 106.

[0047] Figure 4 is a flowchart showing the flow of the learning process during the learning phase of the cognitive function estimation device 10. During the learning phase, the CPU 11 of the cognitive function estimation device 10 executes a learning processing program, thereby realizing the learning process shown in Figure 4. As will be described later, during the learning phase, information from the olfactory data acquisition unit 101, the physical data acquisition unit 102, and the cognitive function test result acquisition unit 103 is acquired for each subject.

[0048] In the cognitive function estimation device 10, when the construction process of the correlation / average information database 100 is started as the learning phase, the learning processing program shown in Figure 4 is executed. First, in step S100, the physical data acquisition unit 102 receives and acquires physical data that indicates information related to the subject's biological system. The physical data includes at least information indicating the subject's age and may further include information indicating gender. The physical data may also include genetic information (APoE). In this embodiment, the case in which the physical data includes at least the subject's age and gender will be described as an example. The physical data acquired by the physical data acquisition unit 102 may use information input in the input unit 16, or it may use information obtained through communication with other devices via the communication unit 18. The physical data acquired by the physical data acquisition unit 102 is then passed to the correlation / average information derivation unit 105.

[0049] Next, in step S102, the cognitive function test result acquisition unit 103 acquires information indicating the results of predetermined cognitive function tests such as MMSE and MoCA-J administered to the subject. The information showing the results is passed to the correlation / average information derivation unit 105. The information acquired by the cognitive function test result acquisition unit 103 may be the information input in the input unit 16, or it may be the information obtained through communication with other devices via the communication unit 18. MMSE is a dementia test known as the Mini-Mental State Examination. MoCA-J is a dementia test known as the Japanese version of MoCA (Montreal Cognitive Assessment).

[0050] Once the acquisition of the aforementioned physical data and test results is complete, the olfactory data collection unit 101 releases odor components to the subject. The olfactory data collection unit 101 then collects data (olfactory data) indicating the subject's olfactory ability to respond to the released odor components.

[0051] Specifically, in step S104, the olfactory data collection unit 101 controls the display unit 17 to display a selection of images related to a smell, including an image showing a predetermined odor component, with that odor component being the correct answer.

[0052] In other words, the olfactory data collection unit 101 displays to the subject images of options for which a predetermined odor component is the correct answer, and images of options for other odor components. The odor components used are those related to cognitive function. For example, vanilla, mint, rag odor, sock odor, caramel, yellow peach, etc., can be used as odor components.

[0053] Figure 5 shows a screen displaying images of items corresponding to the scent components—vanilla, mint, a rag, socks, caramel, and yellow peach—as options. Note that the options are not limited to images of each item; text, videos, actual photographs, or anything else used to measure cognitive function are acceptable. Although not shown in the illustration, an option for "I don't know" may also be included.

[0054] Furthermore, in step S104, the olfactory data collection unit 101 controls the release of odor components to the subject by the odor component release device 20, gradually changing the amount. Specifically, the olfactory data collection unit 101 notifies the odor component release device 20 of a release instruction that includes the odor component to be released and a concentration level indicating the amount of odor component to be released. Here, the odor component to be released is the correct odor component. The concentration level is changed in stages. For example, the release amount is gradually increased at predetermined intervals from the start of release. In this embodiment, the case where the concentration level is increased at predetermined intervals will be explained as an example. In this embodiment, the case where there are three concentration levels (low / medium / high) will be explained as an example.

[0055] The olfactory data collection unit 101 first notifies the odor component release device 20 of the odor component to be released and an instruction to release it at a low concentration level. After a predetermined time has elapsed, the olfactory data collection unit 101 notifies the odor component release device 20 of the odor component to be released and an instruction to release it at a medium concentration level, one step higher. After another predetermined time has elapsed, the olfactory data collection unit 101 notifies the odor component release device 20 of the odor component to be released and an instruction to release it at a high concentration level, one step higher. Finally, after another predetermined time has elapsed, the olfactory data collection unit 101 notifies the odor component release device 20 of an instruction to end the release.

[0056] Furthermore, the olfactory data collection unit 101 repeats the same process when the correct odor component is changed. It should be noted that the display unit 17 may display the options after the olfactory data collection unit 101 releases the odor component; therefore, it is not necessarily required that the odor be released after the display unit 17 displays the options.

[0057] In the next step S106, the olfactory data collection unit 101 will determine the release conditions of the notified odor components. The system derives and records an olfactory score based on the subject's sense of smell, indicating the selection result of the correct image selected by the subject, corresponding to the type and concentration of the odor. The olfactory score derived by the olfactory data collection unit 101 is passed to the correlation and average information derivation unit 105.

[0058] In step S107, the correlation and average information derivation unit 105 records the above-mentioned data, namely physical data, cognitive function test results, and olfactory scores for each odor component, in association with each other.

[0059] Figure 6 shows an example of data including the olfactory score derived by the olfactory data collection unit 101. The olfactory score is derived so that the score increases as the degree of correctness increases for each odor component, when the subject is asked to select an image corresponding to the released odor component. In the example shown in Figure 6, the olfactory score is "3 points" when the subject correctly identifies the released odor component at a low concentration, "2 points" when the subject is incorrect at a low concentration but correct at a medium concentration, "1 point" when the subject is incorrect at low and medium concentrations but correct at a high concentration, and "0 points" when the subject is incorrect at low, medium, and high concentrations. The MMSE and MoCA-J test values ​​shown as cognitive function tests are examples where 30 points is the maximum score.

[0060] Furthermore, whether the image selected by the subject as described above is correct can be determined from the number indicating the image entered by the subject, or it can be automatically determined from the position of the image as seen by the subject. The position of the image as seen by the subject can be determined from the image displayed on the display unit 17 corresponding to the position of the detected gaze point data, which is detected by a detection device such as a camera (not shown).

[0061] The above process is executed until the termination condition indicating completion for a predetermined number of subjects is met (positive judgment in step S108). That is, if the CPU 11 determines that the process for the predetermined number of subjects is incomplete, it makes a negative judgment in step S108 and returns to step S100. On the other hand, if it makes a positive judgment in step S108, it proceeds to step S110.

[0062] In step S110, the correlation and average information derivation unit 105 classifies the data stored in step S107 according to predetermined classification conditions. Specifically, for each cognitive function test, i.e., MMSE and MoCA-J, subjects are classified according to classification conditions based on age and sex. In this embodiment, the classification is performed for each MMSE and MoCA-J using two age classification conditions: under 65 years old and 65 years old and over, and a combination of these and two sex classification conditions: male as the primary sex and female as the secondary sex.

[0063] Next, the correlation and average information derivation unit 105 quantifies the probability of correlation between each odor component and each dementia test for each classified group in step S112. Specifically, the p-value of correlation is applied as data showing the correlation relationship to quantify the probability of correlation between each odor component and each dementia test. The p-value of correlation (0-1) indicates whether the hypothesis "there is no correlation" is significantly different in statistical hypothesis testing (the proportion of cases where the hypothesis is incorrectly rejected), and if the p-value is below the significance level, it is concluded that "the hypothesis 'there is no correlation' is incorrect." (Reference: Bivariate correlation test, t-test, Introduction to Statistics URL)<http: / / www.tamagaki.com / math / Statistics605.html> ).

[0064] The significance level is expressed as Cp, and values ​​such as 10% (Cp=0.1), 5% (Cp=0.05), and 1% (Cp=0.01) are applied. (Reference technical literature: URL <http: / / www1.meijigakuin.ac.jp / ~iwamura / class / seminar_10_2012 / statis tics_chapter5.pdf.>

[0065] Figure 7 shows an example of the p-values ​​of the correlation between each odor component and a dementia test related to MMSE, according to the classification criteria. Figure 8 shows an example of the p-values ​​of the correlation between each odor component and a dementia test related to MoCA-J, according to the classification criteria. Figures 7 and 8 show the correspondence between odor components with a p-value of less than 0.1 (p<0.1).

[0066] Furthermore, in step S112, the correlation and average information derivation unit 105 derives a first average olfactory score and a second average olfactory score for each group, for odors correlated with each dementia test. The first average olfactory score represents the average olfactory score of subjects whose cognitive function test values ​​are below a predetermined threshold. Subjects whose cognitive function test values ​​are below a predetermined threshold are an example of subjects with MCI (mild cognitive impairment). The second average olfactory score represents the average olfactory score of subjects whose cognitive function test values ​​exceed a predetermined threshold. Subjects whose cognitive function test values ​​exceed a predetermined threshold are an example of subjects who do not fall under the category of MCI (mild cognitive impairment) (so-called healthy individuals). Therefore, the correlation and average information derivation unit 105 classifies subjects into those whose test values ​​are below a predetermined threshold and other subjects, and derives the average olfactory score for each classified subject. In other words, participants are classified into age groups based on a predetermined age (65 years old), and the average value of the test results for odor components is derived as the test value for odor components for each classified age group.

[0067] Figure 9 shows examples of the first average olfactory score (average olfactory score for MCI) and the second average olfactory score (average olfactory score for healthy individuals) that correlate with the MMSE and MoCA-J cognitive function tests for each classification condition.

[0068] Next, in step S114, the registration unit 106 registers data (Figures 7 and 8) showing the p-value of the correlation for each odor component according to the classification conditions described above into the correlation and average information database 100. The registration unit 106 also registers data (Figure 9) showing the relationship between the first average olfactory score and the second average olfactory score, which are correlated with the cognitive function tests for each of the classification conditions described above, into the correlation and average information database 100.

[0069] Therefore, the cognitive function estimation device 10 constructs the correlation and average information database 100 in the learning phase as described above.

[0070] (Estimated phase) Next, we will describe the cognitive function estimation device 10, which functions as an estimation phase. Figure 10 is a block diagram showing the functional configuration in the estimation phase, where cognitive function estimation processing is performed using a pre-constructed correlation and average information database 100. The cognitive function estimation device 10 has the function of estimating the cognitive function of the subject (target) being tested using a pre-constructed correlation and average information database 100, as an estimation phase, by executing an estimation processing program. Specifically, the estimation processing program has an olfactory data collection function, a physical data acquisition function, an odor selection function, a correlation and average information extraction function, a cognitive function evaluation function, and a presentation function. In this estimation phase, the cognitive function estimation device 10 selects odor components that have a correlation with identification information including the subject's age and gender based on the correlation and average information database 100, and estimates cognitive function from the olfactory data obtained from those odor components.

[0071] As shown in Figure 10, in the estimation phase, the CPU 11 of the cognitive function estimation device 10 functions as an olfactory data acquisition unit 101, a physical data acquisition unit 102, a scent selection unit 104, an extraction unit 107 that extracts data from the correlation / average information database 100, a cognitive function evaluation unit 108, and a presentation unit 110. The physical data acquisition unit 102 is an example of an acquisition unit in this disclosure, the scent selection unit 104 is an example of a scent component setting unit in this disclosure, and the cognitive function evaluation unit 108 is an example of an estimation unit in this disclosure. This is an example of a fixed section, and the presentation section 110 is an example of a presentation section of this disclosure.

[0072] Figure 11 is a flowchart showing the flow of the estimation process in the estimation phase of the cognitive function estimation device 10. In the estimation phase, the CPU 11 of the cognitive function estimation device 10 executes an estimation processing program, thereby realizing the estimation process shown in Figure 11.

[0073] In the cognitive function estimation device 10, when the estimation phase for estimating the subject's cognitive function is initiated, the estimation processing program shown in Figure 11 is executed. First, in step S200, the physical data acquisition unit 102 receives and acquires physical data that indicates information related to the subject's biological system. Here, the physical data acquired includes data indicating the subject's age and gender. The physical data acquired by the physical data acquisition unit 102 may use information input in the input unit 16, or it may use information obtained through communication with other devices via the communication unit 18. The physical data acquired by the physical data acquisition unit 102 is then passed to the odor selection unit 104 and the cognitive function evaluation unit 108.

[0074] Next, in step S202, the odor selection unit 104 uses the registered correlation and average information database 100 to select odor components corresponding to the subject's physical data. Specifically, when the odor selection unit 104 receives physical data from the physical data acquisition unit 102, it requests the extraction unit 107 to extract data from the registered correlation and average information database 100 that shows odor components corresponding to the physical data and that have a correlation with the subject's age and gender. In response to the request from the odor selection unit 104, the extraction unit 107 provides data (Figures 7 and 8) showing the p-value of the correlation for each odor component according to the classification conditions corresponding to the physical data. The odor selection unit 104 uses the data from the extraction unit 107 to select odor components corresponding to the subject's physical data. The information showing the odor components selected by the odor selection unit 104 is passed to the olfactory data collection unit 101.

[0075] In other words, the odor selection unit 104 selects odor components that correlate with cognitive function tests in groups corresponding to the subject's age and gender, based on data obtained by accessing the registered correlation and average information database 100. For example, if the odor selection unit 104 is given physical data indicating that the subject is a woman aged 65 or older, it will select the odor components of "caramel" (Figure 7) and "socks" (Figure 8).

[0076] The olfactory data collection unit 101 releases odor components to the subject according to the information indicating the odor components provided by the odor selection unit 104. The olfactory data collection unit 101 then collects data indicating the subject's olfactory ability to the released odor components.

[0077] Specifically, in step S204, the olfactory data collection unit 101 controls the display unit 17 to display multiple options of images related to the odor, including an image representing the odor component, as the correct answer for each odor component received from the odor selection unit 104.

[0078] Furthermore, in step S204, the olfactory data collection unit 101 controls the release of odor components to the subject in the odor component release device 20, changing the amount in stages, for example, at three concentration levels (low / medium / high). Specifically, the olfactory data collection unit 101 notifies the odor component release device 20 of a release instruction that includes the odor component to be released and the concentration level indicating the amount of odor component to be released. Here, the odor component to be released is the correct odor component.

[0079] Next, in step S206, the olfactory data collection unit 101 derives and records an olfactory score based on the subject's sense of smell, indicating the selection result of the correct image selected by the subject, corresponding to the release conditions (type and concentration) of the odor components. The olfactory score derived by the olfactory data collection unit 101 is passed to the cognitive function evaluation unit 108.

[0080] In step S208, the cognitive function assessment unit 108 obtains the average olfactory score for the classification corresponding to the subject. Specifically, the cognitive function assessment unit 108 requests the extraction unit 107 to extract data showing the average olfactory score for the group corresponding to the classification based on the subject's physical data. In response to the request from the cognitive function assessment unit 108, the extraction unit 107 provides data (Figure 9) showing the first average olfactory score and the second average olfactory score for each dementia test corresponding to the group corresponding to the classification condition based on physical data. The cognitive function assessment unit 108 then obtains data (Figure 9) showing the first average olfactory score and the second average olfactory score registered in the correlation / average information database 100 from the extraction unit 107.

[0081] Next, in step S210, the cognitive function evaluation unit 108 uses the average value of the acquired olfactory scores to derive an evaluation value for the subject's cognitive function. The index derived by the cognitive function evaluation unit 108 is passed to the presentation unit 110. The subject's cognitive function evaluation value can be derived as an index SR that indicates the evaluation value using the following calculation formula. The evaluation value derived by the calculation formula, i.e., the index SR, is calculated such that as it approaches "0", the degree to which cognitive function is likely to be declining (for example, the risk of dementia) decreases.

[0082] SR = (Sc - Sa) / (Sc - Sn) However, Sn represents the first average olfactory score (average olfactory score for MCI), Sc represents the second average olfactory score (average olfactory score for healthy individuals), and Sa represents the olfactory score of the subject being estimated.

[0083] For example, for a subject whose physical data indicates they are female and 65 years of age or older, the olfactory score for the "caramel" scent component will be the average olfactory score from the MMSE cognitive function test, labeled as "MMSE Judgment" in Figure 9. Similarly, for the "socks" scent component, the average olfactory score from the MoCA-J cognitive function test, labeled as "MoCA-J Judgment" in Figure 9, will be applied.

[0084] Then, in step S212, the presentation unit 110 presents information related to cognitive function. For example, the presentation unit 110 performs display control to display the index SR on the display unit 17 as information related to the cognitive function test. The information related to the cognitive function test may include data indicating whether or not there is a risk of cognitive decline, and data regarding the degree (magnitude) of the risk of cognitive decline. For data indicating whether or not there is a risk of cognitive decline, it is possible to determine that there is no risk of cognitive decline if the index SR is below a predetermined threshold, and that there is a risk of cognitive decline if the index SR exceeds a predetermined threshold. For data regarding the degree (magnitude) of the risk of cognitive decline, it is sufficient to ensure that the degree (magnitude) of the risk of cognitive decline increases as the index SR increases.

[0085] Furthermore, the presentation unit 110 can present cognitive function tests recommended based on the subject's olfactory score as information related to cognitive function testing. In order to present cognitive function tests recommended based on the subject's olfactory score, the presentation unit 110 only needs to obtain information indicating the recommended cognitive function tests (recommendation information) from the cognitive function evaluation unit 108. That is, the cognitive function evaluation unit 108 can derive cognitive function tests corresponding to the odor components indicated by the subject's olfactory score as recommended cognitive function tests and notify the presentation unit 110. For example, if a woman under 65 years of age has a high index SR for the release of the odor component of "socks" (above a predetermined threshold), the correlation / average information database 100 can be referenced to present the subject with MoCA-J cognitive function tests that correlate with the relevant odor component (in this case, the odor component of "socks") as recommendation information recommending that she take the desired cognitive function test.

[0086] Therefore, in the estimation phase, the cognitive function estimation device 10 estimates cognitive function using the correlation and average information database 100 as described above.

[0087] As described above, the cognitive function device according to this embodiment releases multiple odor components at varying concentrations, allows subjects to select the released odor components, and correlates the olfactory score based on correct and incorrect selections with the results of multiple dementia tests. Furthermore, subjects are classified by age and sex, and the likelihood of correlation between each odor component and each cognitive function test is quantified for each classified group. In addition, for odor components correlated with each dementia test, the average olfactory score of subjects at risk of cognitive decline (e.g., subjects with MCI: mild cognitive impairment) and the average olfactory score of subjects at risk of cognitive decline less than those at risk (e.g., healthy individuals) are derived and stored as a correlation / average information database. By using this correlation / average information database, it becomes possible to derive an evaluation value (an index indicating the risk of cognitive decline) by deriving the olfactory score based on the release of odor components.

[0088] Furthermore, for subjects who may have impaired cognitive function, it becomes possible to present recommended cognitive function tests by using the correlation and average information database as described above.

[0089] Thus, olfactory testing allows for accurate and simple estimation of the risk of cognitive decline, such as dementia risk, based on olfactory scores. Furthermore, since the estimation of cognitive function can be performed simply by the subject reporting their olfactory perception of released odor components, the burden on the subject can be minimized.

[0090] In this embodiment, the case in which the cognitive function estimation device 10 constructs the correlation and average information database 100 (learning phase) has been described, but learning data learned by other devices may be acquired as the already constructed correlation and average information database 100.

[0091] Furthermore, the technology disclosed herein is not limited to the embodiments described above, and various modifications and applications are possible without departing from the spirit of this invention.

[0092] For example, the above embodiment described a case where the amount of odor components released is increased in stages, but it is not limited to this. A configuration in which the amount of odor components released is decreased in stages is also possible.

[0093] Furthermore, when multiple scents are released, the order of the questions, which represent pairs of the correct answer and the answer choices, may also be considered when measuring olfactory ability. This is because accuracy can be improved by considering the correlation between the order of each question and the similarity of the scent components of the correct answer.

[0094] [Second Embodiment] Next, a second embodiment of the present disclosure will be described. Since the second embodiment has the same configuration as the first embodiment, the same reference numerals are used for the same parts, and detailed descriptions will be omitted. The differences will be described instead.

[0095] In the first embodiment, cognitive function was evaluated for the subject (target) using the estimated age and sex of the subject. In the second embodiment, cognitive function was evaluated for the subject without using the subject's age and sex.

[0096] Specifically, the cognitive function estimation system according to the second embodiment comprises an odor component emission unit, a score setting unit, and an estimation unit. The odor component emission unit emits each of several odor components to the subject, with the concentration of each odor component changing in multiple stages. The score setting unit determines whether the subject's answer indicating which odor component was emitted from the odor component emission unit is correct. If this occurs, a score corresponding to the concentration that showed the correct answer is set. The estimation unit estimates the cognitive function evaluation value for the subject based on the sum of the scores for the multiple odor components set by the score setting unit.

[0097] According to the cognitive function estimation system of the second embodiment, cognitive function can be estimated with high accuracy by olfactory testing using certain odor components.

[0098] <Configuration of the Cognitive Function Estimation System> Figure 12 is a block diagram showing the overall schematic configuration of the cognitive function estimation system according to this embodiment. As shown in Figure 12, in the cognitive function estimation system according to this embodiment, the cognitive function estimation device 10A is equipped with a correlation / probability information database 100A instead of the correlation / average information database 100 shown in Figure 1. The cognitive function estimation device 10A estimates cognitive function using the correlation / probability information database 100A. The correlation / probability information database 100A stores correlation information showing the correlation between the results of a cognitive function test performed on a subject and the test results of odor components determined from the combination of odor components determined from the cognitive function test results. The odor component emission device 20 is an example of an odor component emission unit in this disclosure.

[0099] In this embodiment, physical data, including age and gender, of the subjects and the individuals whose cognitive function is to be estimated is not required.

[0100] The electrical configuration of the cognitive function estimation device 10A according to this embodiment is the same as that of the cognitive function estimation device 10 described above, so a description will be omitted.

[0101] Next, the functional configuration of the cognitive function estimation device 10A will be described. In this embodiment, the cognitive function estimation device 10A exhibits different functional configurations in a learning phase in which a correlation / probability information database 100A used for cognitive function estimation processing is constructed, and in an estimation phase in which cognitive function estimation processing is performed using the constructed correlation / probability information database 100A.

[0102] (Learning Phase) First, let's describe the cognitive function estimation device 10A, which functions as a learning phase. Figure 13 is a block diagram showing the functional configuration in the learning phase of the cognitive function estimation device 10A, where the correlation and probability information database 100A is constructed.

[0103] The cognitive function estimation device 10A has a function to construct a correlation / probability information database 100A as a learning phase by executing a learning processing program. Specifically, the learning processing program has an olfactory data collection function, a cognitive function test result acquisition function, and a registration function, excluding the physical data acquisition function and correlation / average information derivation function described above, and newly has a correlation / probability information derivation function. In this learning phase, the cognitive function estimation device 10A registers in the correlation / probability information database 100A combinations of odor components that correlate with a predetermined cognitive function test, and correlation information showing the correlation between the test results of the odor components and the test results of the cognitive function test.

[0104] As shown in Figure 13, during the learning phase, the CPU 11 of the cognitive function estimation device 10A functions as an olfactory data acquisition unit 101, a cognitive function test result acquisition unit 103, a correlation / probability information derivation unit 115, and a registration unit 116.

[0105] Figure 14 is a flowchart illustrating the flow of the learning process during the learning phase of the cognitive function estimation device 10A. During the learning phase, the CPU 11 of the cognitive function estimation device 10A executes a learning program, thereby realizing the learning process shown in Figure 14. Further details will be discussed later. During the learning phase, information from the olfactory data collection unit 101 and the cognitive function test result acquisition unit 103 is acquired for each subject.

[0106] In the cognitive function estimation device 10A, when the construction process of the correlation / probability information database 100A is started as the learning phase, the learning processing program shown in Figure 14 is executed. First, similar to step S102 described above, in step S120, the cognitive function test result acquisition unit 103 acquires information indicating the results of a predetermined cognitive function test performed on subjects of the target group. In this embodiment, the cognitive function test is described as an example of a dementia screening test performed by a licensed physician, and the MoCA-J described above is applied. In step S120, the MoCA-J acquired by the cognitive function test result acquisition unit 103 is... The information showing the results of the cognitive function test is passed to the correlation / probability information derivation unit 115.

[0107] Once the acquisition of information indicating the test results is complete, in step S122, similar to step S104 described above, the olfactory data acquisition unit 101 releases odor components to the subject and collects data indicating the subject's olfactory ability to the released odor components. That is, the olfactory data acquisition unit 101 displays to the subject an image of an option in which a predetermined odor component is the correct answer, and images of options for other odor components (Figure 5), and controls the odor component release device 20 to release the correct odor component in a gradually changing amount.

[0108] Next, in step S124, similar to step S106 described above, the olfactory data collection unit 101 derives and records an olfactory score based on the subject's sense of smell, indicating the selection result of the correct image selected by the subject, corresponding to the release conditions (type and concentration) of the odor components. The olfactory score derived by the olfactory data collection unit 101 is passed to the correlation and probability information derivation unit 115.

[0109] In step S126, similar to step S107 described above, the correlation / probability information derivation unit 115 records the above-mentioned data, namely the results of the cognitive function test and the olfactory score, in association with each other.

[0110] Figure 15 shows an example of data that associates olfactory data, including olfactory scores derived by the olfactory data collection unit 101, with data from cognitive function tests. Figure 15 is an example of a data table in which data such as olfactory scores for each odor component and cognitive function test results are associated with an identification number that identifies the subject.

[0111] Next, similar to step S108 described above, the above process is performed until the termination condition indicating completion for a predetermined number of subjects is met (affirmative judgment in step S128).

[0112] In step S130, the correlation / probability information derivation unit 115 uses the odor test results (olfactory data including olfactory score) and the cognitive function test results (MoCA-J data) to derive combinations of odor components to be used in the estimation phase described later.

[0113] The process of deriving combinations of odor components is carried out by executing the following selection procedure.

[0114] In the first selection step, data obtained from multiple subjects through the aforementioned tests, namely MoCA-J data representing cognitive function test results and olfactory score data representing olfactory function test results, are acquired. In this embodiment, olfactory scores based on the release of six types of scent components—vanilla, rag odor, caramel, mint, socks, and yellow peach—were applied (see Figure 15).

[0115] In the second selection step, the above data (olfactory data and MoCA-J data) is classified into individuals suspected of having MCI and individuals who do not have MCI (so-called healthy individuals).

[0116] Incidentally, a predetermined threshold is set for the cognitive function test values ​​(MoCA-J data) to determine whether a person is "suspected of having MCI (mild cognitive impairment)," based on statistical and empirical perspectives. Therefore, subjects whose cognitive function test values ​​(MoCA-J data) exceed the threshold are designated as so-called healthy individuals (hereinafter referred to as CN individuals) who do not fall under the category of MCI. On the other hand, subjects whose cognitive function test values ​​are below the threshold are designated as subjects suspected of having MCI (hereinafter referred to as MCI individuals). Accordingly, the correlation / probability information derivation unit 115 classifies individuals into MCI individuals whose MoCA-J data is below the threshold and CN individuals whose data exceeds the threshold. For example, if the MoCA-J data is distributed within the range of values ​​from 0 to 30, the threshold is set to 25, and subjects with values ​​of 25 or less are determined to be MCI individuals.

[0117] In the third selection step, for any combination of odor components from among multiple odor components (hereinafter referred to as "odor component groups"), an evaluation index is derived to assess the correlation with cognitive function test results for each different odor component group.

[0118] In this embodiment, for each odor component group, the distribution of MCI individuals and CN individuals with respect to the total olfactory score is analyzed, and an evaluation index is derived. Receiver Operating Characteristic (ROC) analysis is applied to the distribution analysis. The evaluation index is derived by calculating the AUC (Area Under the Curve), which represents the discriminative evaluation value obtained from this ROC analysis. That is, in the third selection step, ROC analysis is performed on the distribution of MCI individuals and CN individuals with respect to the total olfactory score of multiple odor components for each odor component group, and the evaluation index AUC is derived. Note that since ROC analysis is a known technique, a detailed explanation is omitted.

[0119] In the fourth selection step, a group of odor components with an evaluation index (AUC) that meets predetermined evaluation conditions is derived as a combination of odor components to be used in the estimation phase described later. The evaluation conditions can be selected by extracting any of the odor component groups whose evaluation index (AUC) exceeds a predetermined threshold (correlation value). In this case, the odor component group with the maximum evaluation index (AUC) can be selected as the most preferable combination of odor components for cognitive function testing.

[0120] Figure 16 shows an example of data illustrating the evaluation index. Figure 16 is an example of an evaluation table in which the correspondence between odor component groups and evaluation indexes showing the AUC values ​​in the results of ROC analysis is arranged in descending order of AUC value. The column for odor component combinations is indicated by an identification number that identifies the odor component. Specifically, "1" represents vanilla, "2" represents rag odor, "3" represents caramel, "4" represents mint, "5" represents socks, and "6" represents yellow peach.

[0121] In the example shown in Figure 16, the odor component group consisting of the four scents "1" vanilla, "4" mint, "5" socks, and "6" yellow peach yielded the highest AUC (0.649 in Figure 16). This indicates that the odor component group consisting of the four scents "1" vanilla, "4" mint, "5" socks, and "6" yellow peach is the combination of odor components that has the strongest correlation with the test values ​​of the cognitive function test. It should also be noted that by setting a threshold (correlation value) of 0.645, the above four scent components, which ranked 1st, can be designated as the odor component group. The threshold (correlation value) can be set as appropriate, and if multiple odor component groups are extracted, the component group with the highest value may be selected as described above, or an odor component with fewer combinations than other component groups may be selected.

[0122] Furthermore, the derivation of the evaluation metrics described above is not limited to deriving AUC by performing ROC analysis. For example, other analytical methods that analyze the distribution of MCI individuals and CN individuals can be applied. This is also acceptable. If other analytical methods are applied, an evaluation method should be used to evaluate the analysis results using the other analytical methods.

[0123] In this way, the correlation and probability information derivation unit 115 uses data from the odor test results and the cognitive function test results to derive combinations of odor components (in this case, vanilla, mint, socks, and yellow peach).

[0124] Next, in step S132 of Figure 14, the correlation / probability information derivation unit 115 derives data showing the correlation between the odor test results (olfactory data including olfactory score) and the cognitive function test results (MoCA-J data).

[0125] The process of deriving the data showing the above correlation is carried out by executing the following derivation procedure.

[0126] In the first derivation procedure, multiple subjects are classified into MCI and CN, and the number of each group of subjects for olfactory data in each odor component group is derived.

[0127] First, the correlation and probability information derivation unit 115 uses the olfactory data and the cognitive function test results data (Figure 15) to obtain data that associates the MoCA-J data, which is the result of the cognitive function test, with the olfactory score, which indicates the result of the olfactory function test. Here, the olfactory score, which indicates the result of the olfactory function test, is determined by applying the olfactory scores for the four types of odor components derived by the selection procedure described above: vanilla, mint, socks, and yellow peach.

[0128] Next, the correlation and probability information derivation unit 115 calculates the total olfactory score (hereinafter referred to as the total score) for each of the multiple subjects, and classifies them into MCI and CN using MoCA-J data for each total score, and calculates the number of people in each classification.

[0129] Figure 17 shows an example of data illustrating the relationship between the total score and the number of individuals with mild cognitive impairment (MCI) and those with chronic cognitive impairment (CN). In the example shown in Figure 17, the number of CN individuals tends to increase as the total score increases, while the number of MCI individuals is random.

[0130] In the second derivation procedure, for each total olfactory score, the ratio of the number of MCI individuals to the number of CN individuals (hereinafter referred to as the MCI rate) is calculated based on the following formula. (MCI rate) = (Number of people with MCI) / {(Number of people with MCI) + (Number of people with CN)}

[0131] In the third derivation step, the correlation between the total olfactory score and the probability of being MCI is derived. The MCI rate can be considered as the probability that a subject is likely to be MCI, relative to the total olfactory score based on the odor component group. Therefore, in this embodiment, the correlation between the total olfactory score and the probability of being MCI is derived as the probability of being MCI, using the test results from a large number of subjects.

[0132] Figure 18 shows an example of the relationship between odor test results and cognitive function test results, illustrating the two-dimensional relationship between the total odor test score and the MCI rate. As shown in Figure 18, the MCI rate decreases as the total score increases. The third derivation step derives the trend in the relationship shown in Figure 18.

[0133] Specifically, the MCI rate is weighted by the sample size (total number of subjects) for each value of the MCI rate, and the total MCI rate is then analyzed using a simple linear regression. As a result of this analysis, the correlation between the total score and the probability of being MCI can be expressed, for example, by the regression equation shown below. (MCI rate) = -0.04275 × (Total score) + 0.6703

[0134] In simple linear regression analysis, the relationship between the total score and the MCI rate (Figure 18) is examined by generating the same number of samples as the number of people used to calculate the MCI rate. For example, if there are 32 people with a total score of 0 who have MCI and 18 people with CN, the MCI rate is 0.64 (=32 / (32+18)). Since there are 50 people in total with a total score of 0, 50 samples are generated with an MCI rate of 0.64 for the result of a total score of 0, and simple linear regression is performed. In other words, since the MCI rate calculated with a small number of people may be greatly affected by outliers, the MCI rate calculated with a small number of people is disregarded, and the MCI rate calculated with a larger number of people is given more weight in the regression.

[0135] The regression equation above can be replaced with the following general equation. f(x) = a·x + b f(x) represents the MCI rate, and x represents the total score. The coefficients a and b are constants obtained through analysis.

[0136] In this way, the correlation and probability information derivation unit 115 derives the information related to the regression equation described above as data showing the correlation between the odor test results and the cognitive function test results.

[0137] Next, in step S134 shown in Figure 14, the registration unit 116 registers data showing the combinations of odor components described above into the correlation and probability information database 100A. The registration unit 116 also registers data showing the correlation between the odor test results and the cognitive function test results into the correlation and probability information database 100A.

[0138] As described above, the cognitive function estimation device 10A constructs the correlation and probability information database 100A during the learning phase described above.

[0139] (Estimated phase) Next, we will describe the cognitive function estimation device 10A, which functions as an estimation phase. Figure 19 is a block diagram showing the functional configuration in the estimation phase when cognitive function estimation processing is performed using the constructed correlation and probability information database 100A.

[0140] The cognitive function estimation device 10A has the function of estimating the cognitive function of the subject (target) being tested using a pre-constructed correlation and probability information database 100A, as part of the estimation phase, by executing an estimation processing program. Specifically, the estimation processing program has an olfactory data collection function, an odor selection function, an extraction function for correlation and probability information, a cognitive function evaluation function, and a presentation function. In this estimation phase, the cognitive function estimation device 10A selects odor components that have a correlation relationship based on the correlation and probability information database 100A, and estimates cognitive function from the olfactory data obtained from those odor components.

[0141] As shown in Figure 19, in the estimation phase, the CPU 11 of the cognitive function estimation device 10A functions as an olfactory data collection unit 101, a scent selection unit 104, an extraction unit 117 that extracts data from the correlation / probability information database 100A, a cognitive function evaluation unit 118, and a presentation unit 120. The olfactory data collection unit 101 is an example of the score setting unit of this disclosure, the scent selection unit 104 is an example of the scent component setting unit of this disclosure, the cognitive function evaluation unit 118 is an example of the estimation unit of this disclosure, and the presentation unit 120 is an example of the presentation unit of this disclosure.

[0142] Figure 20 is a flowchart showing the flow of the estimation process in the estimation phase of the cognitive function estimation device 10A. In the estimation phase, the CPU 11 of the cognitive function estimation device 10A executes an estimation processing program, thereby realizing the estimation process shown in Figure 20.

[0143] In the cognitive function estimation device 10A, when the estimation phase begins, which is the estimation phase for estimating the cognitive function of the subject (target) being tested, the estimation processing program shown in Figure 20 is executed. First, in step S220, the odor selection unit 104 obtains data indicating combinations of multiple odor components (odor component groups) from the correlation / probability information database 100A. Specifically, the odor selection unit 104 requests the extraction unit 117 to extract data indicating combinations of multiple odor components (odor component groups) from the correlation / probability information database 100A. In response to the request from the odor selection unit 104, the extraction unit 107 passes data indicating combinations of multiple odor components (odor component groups).

[0144] Next, in step S222, the odor selection unit 104 selects the multiple odor components (odor component group) acquired in step S220 as the multiple odor components (odor component group) to be used for testing. The odor selection unit 104 passes information indicating the selected multiple odor components to the olfactory data collection unit 101. Here, the odor selection unit 104 selects four types of odor components: vanilla, mint, socks, and yellow peach.

[0145] In step S224, the olfactory data collection unit 101 releases the odor component to the subject according to the information indicating the odor component provided by the odor selection unit 104, similar to step S204 described above. The olfactory data collection unit 101 then collects data indicating the subject's olfactory ability to the released odor component.

[0146] Next, in step S226, similar to step S206 described above, the olfactory data collection unit 101 derives and records an olfactory score based on the subject's sense of smell, indicating the selection result of the correct image selected by the subject, corresponding to the release conditions (type and concentration) of the odor components. The olfactory score derived by the olfactory data collection unit 101 is passed to the cognitive function evaluation unit 118.

[0147] In step S228, the cognitive function evaluation unit 118 acquires data showing the correlation between the olfactory score and the cognitive function test results. Specifically, the cognitive function evaluation unit 118 requests the extraction unit 117 to extract data showing the regression equation described above. In response to the request from the cognitive function evaluation unit 118, the extraction unit 117 provides the data showing the regression equation. The cognitive function evaluation unit 118 then acquires the data showing the regression equation registered in the correlation / probability information database 100A from the extraction unit 117.

[0148] Next, in step S230, the cognitive function evaluation unit 118 first sums the olfactory scores for each odor component derived by the olfactory data collection unit 101 to derive a total score. The total score may also be derived and obtained by the olfactory data collection unit 101. Next, using the total score and the regression equation, the cognitive function evaluation value for the subject, i.e., the MCI rate (probability), is derived. The index derived by the cognitive function evaluation unit 118 is passed to the presentation unit 120.

[0149] Then, in step S232, the presentation unit 120 presents information related to cognitive function. For example, the presentation unit 120 performs display control to display the value of the index MCI rate (probability) on the display unit 17 as information related to the cognitive function test. The data displayed on the display unit 17 can indicate that there is no risk of cognitive decline if the index is below a predetermined threshold, and that there is a risk of cognitive decline if the index SR exceeds a predetermined threshold. The data regarding the degree (magnitude) of the risk of cognitive decline should be such that the degree (magnitude) of the risk of cognitive decline increases as the value of the index MCI rate increases. Alternatively, the MCI rate may be displayed as a probability.

[0150] Therefore, in the estimation phase, the cognitive function estimation device 10A estimates cognitive function using the correlation and probability information database 100A as described above.

[0151] As described above, the cognitive function device according to this embodiment releases multiple odor components at varying concentrations, allows the subject to select the released odor component, and correlates the olfactory score based on correct and incorrect selections with the results of the cognitive function test. Furthermore, using the sum of the olfactory scores (total score) for multiple odor components correlated with each cognitive function test, a regression equation showing the correlation between the odor test results and the cognitive function test results is derived and stored as a correlation / probability information database. By using this correlation / probability information database to derive the sum of the olfactory scores based on the release of odor components related to cognitive function, it becomes possible to derive an evaluation value (an index indicating a potential decline in cognitive function).

[0152] [Third Embodiment] Next, a third embodiment of the present disclosure will be described. Since the third embodiment has the same configuration as the first embodiment, the same reference numerals are used for the same parts, and detailed descriptions will be omitted. The differences will be described instead.

[0153] In the first embodiment, cognitive function was evaluated for a subject (target) using their age and sex. In the third embodiment, cognitive function was further evaluated for the subject using their smoking history. That is, in this embodiment, the identification information includes physical data indicating age and sex, and smoking data indicating whether or not they have smoked as an example of smoking history.

[0154] <Configuration of the Cognitive Function Estimation System> Figure 21 is a block diagram showing the overall schematic configuration of the cognitive function estimation system according to the third embodiment. The cognitive function estimation system according to this embodiment further uses the smoking data of the subject (target) to be estimated and includes a cognitive function estimation device 10B that estimates cognitive function from information obtained by the emission of odor components by the odor component emission device 20. That is, the cognitive function estimation device 10B estimates cognitive function using a correlation / regression information database 100B that stores information relating identification information including the age, sex and smoking history of the subject who underwent the cognitive function test, the types of odor components that correlate with the cognitive function test, and the test values ​​of said odor components.

[0155] Furthermore, the smoking data included in the identification information can include data on whether or not the person has smoked before. If the person has smoked for a specified period or for a specified number of cigarettes or more, they may be classified as having smoking experience, and if the period or number of cigarettes is less than the specified period, they may be classified as not having smoking experience.

[0156] In the cognitive function estimation device 10B, the ROM 13 or memory unit 19 (Figure 2) stores an estimation processing program for executing the cognitive function estimation process, which will be described in detail later, and a learning processing program for constructing the correlation / regression information database 100B used for the cognitive function estimation process.

[0157] The electrical configuration of the cognitive function estimation device 10B is the same as that of the cognitive function estimation device 10 described above, so its explanation will be omitted.

[0158] Next, the functional configuration of the cognitive function estimation device 10B will be described. The cognitive function estimation device 10B exhibits different functional configurations in the learning phase, which constructs the correlation / regression information database 100B used for cognitive function estimation processing, and the estimation phase, which performs cognitive function estimation processing using the constructed correlation / regression information database 100B.

[0159] (Learning Phase) First, let's describe the cognitive function estimation device 10B, which functions as a learning phase. Figure 22 is a block diagram showing the functional configuration in the learning phase of the cognitive function estimation device 10B, where the correlation / regression information database 100B is constructed.

[0160] The cognitive function estimation device 10B has the function of constructing a correlation / regression information database 100B as a learning phase by executing a learning processing program. Specifically, the learning processing program has an olfactory data collection function, a data acquisition function, a cognitive function test result acquisition function, a correlation / regression information derivation function, and a registration function. In this learning phase, the cognitive function estimation device 10B registers identification information including the subject's age, gender, and smoking history, the types of odor components that correlate with a predetermined cognitive function test, and correlation information showing the correlation between the odor components and the cognitive function test results in the correlation / regression information database 100B.

[0161] As shown in Figure 23, during the learning phase, the CPU 11 of the cognitive function estimation device 10B functions as an olfactory data collection unit 101, a data acquisition unit 122, a cognitive function test result acquisition unit 103, a correlation / regression information derivation unit 125, and a registration unit 126.

[0162] Figure 23 is a flowchart showing the flow of the learning process during the learning phase of the cognitive function estimation device 10B. During the learning phase, the CPU 11 of the cognitive function estimation device 10B executes a learning process program, thereby realizing the learning process shown in Figure 24. As will be described later, during the learning phase, information from the olfactory data collection unit 101, the data acquisition unit 122, and the cognitive function test result acquisition unit 103 is acquired for each subject.

[0163] In the cognitive function estimation device 10B, when the construction process of the correlation / regression information database 100B is started as the learning phase, the learning processing program shown in Figure 24 is executed. First, in step S140, the data acquisition unit 122 receives and acquires physical data, including information indicating the subject's age and gender, and smoking data. The data acquired by the data acquisition unit 122 is passed to the correlation / regression information derivation unit 125.

[0164] In step S142, similar to step S102, the cognitive function test result acquisition unit 103 acquires information indicating the results of predetermined cognitive function tests, such as the MMSE and MoCA-J, administered to the subject. The information indicating the results of predetermined cognitive function tests acquired by the cognitive function test result acquisition unit 103 is passed to the correlation / regression information derivation unit 125.

[0165] In step S144, similar to step S104, the olfactory data collection unit 101, which collects data related to the olfactory test, controls the display unit 17 to display multiple options showing images related to odors, including an image showing the odor component, with a predetermined odor component as the correct answer. The odor components can be vanilla, mint, rag odor, sock odor, caramel, yellow peach, etc., as in the example described above.

[0166] In step S146, similar to step S106 above, the olfactory data collection unit 101 derives and records an olfactory score based on the subject's sense of smell, indicating the selection result of the correct image selected by the subject, corresponding to the release conditions (type and concentration) of the notified odor components. The olfactory score derived by the olfactory data collection unit 101 is passed to the correlation / regression information derivation unit 125.

[0167] In step S147, the correlation / regression information derivation unit 125 records the aforementioned data, namely physical data, smoking data, cognitive function test results, and olfactory score, in association with each other.

[0168] Figure 24 shows an example of data including the olfactory score derived by the olfactory data collection unit 101. Similar to the example shown in Figure 6, the olfactory score is calculated by having the subject select an image corresponding to the released odor component, with the score increasing as the degree of correctness of the selection for each odor component increases. This is derived from the above. Furthermore, the cognitive function test values ​​shown are examples based on a maximum score of 30 points, as mentioned above.

[0169] The above process is executed until the termination condition indicating completion for a predetermined number of subjects is met (positive judgment in step S148). That is, if the CPU 11 determines that the process for the predetermined number of subjects is incomplete, it makes a negative judgment in step S148 and returns to step S140. On the other hand, if it makes a positive judgment in step S148, it moves the process to step S150.

[0170] In step S150, the correlation / regression information derivation unit 125 classifies the data stored in step S147 according to predetermined classification conditions. In this embodiment, the predetermined classification conditions are discrimination conditions for determining whether a subject who has undergone a cognitive function test is an MCI (Mild Cognitive Impairment) person or a CN (Computational Nervous System) person. In cognitive function tests, thresholds for determining whether a test value indicates an MCI person or a CN person are set in advance. This is because the criteria for determining whether a test value indicates an MCI person or a CN person are thought to differ depending on the type of cognitive function test.

[0171] For example, in the MoCA-J, a threshold of 25 points is pre-set for test scores distributed in the range of 0 to 30 points. Therefore, subjects with a MoCA-J score of 25 points or less are more likely to be MCI (Mild Cognitive Impairment), while subjects with a score above 25 points are less likely to be MCI and more likely to be CN (Computational Nervous System). Note that the 25-point threshold set in the MoCA-J is just an example and is not limited to this score. On the other hand, in the MMSE (Mini-Mental Health Examination), a threshold of 27 points is pre-set for test scores distributed in the range of 0 to 30 points. Therefore, subjects with an MMSE score of 27 points or less are more likely to be MCI, while subjects with a score above 27 points are less likely to be MCI and more likely to be CN. Note that the 27-point threshold set in the MMSE is also just an example and is not limited to this score.

[0172] Therefore, in step S150, as the first step, subjects are classified according to the type of cognitive function test, MMSE and MoCA-J. That is, the subjects' identification information and olfactory score data are classified into MCI and CN.

[0173] Next, in step S152, the correlation / regression information derivation unit 125 uses the classified data to derive data showing the subject's data, odor components, and the correlation between the odor components and cognitive function tests. Step S152 includes a process of analyzing the data distribution to derive combinations of odor components according to the type of cognitive function test.

[0174] First, using the data classified as described above, the data distribution is analyzed to derive combinations of odor components corresponding to the type of cognitive function test. These odor component combinations are determined using the AUC obtained from the analysis using a regression model. In this embodiment, a logistic regression model is applied as the regression model.

[0175] Specifically, logistic regression is performed on the total score for multiple different combinations of odor components using the subject's data, and the AUC is derived. Logistic regression is a well-known technique, so a detailed explanation is omitted, but it is an analytical method that predicts the probability that a variable that takes either 0 or 1 is 1 using multiple explanatory variables. In this embodiment, for logistic regression, the explanatory variables for the subject's data are data indicating age, sex, and smoking history from the subject's identification information and smoking data, and the subject's olfactory score, while the dependent variable is a binary variable where MCI individuals are "1" and CN individuals are "0". Here, an analysis is performed to predict the probability px of a subject being MCI (i.e., a subject whose binary variable is 1) using multiple explanatory variables, based on the variable indicating whether the subject is MCI or CN (i.e., a binary variable that takes either 0 or 1).

[0176] The logistic regression model can be represented by the following estimation formula. px = 1 / [1 + exp{-(pa + pb + pc + pd)}] In the above formula, When A, B, C, and D are constants, pa represents (A × total score), pb represents (B × age), pc indicates (C × gender), pd indicates (D × smoking history).

[0177] The constants A, B, C, and D mentioned above are the coefficients of logistic regression. These constants, A, B, C, and D, can be estimated using the Maximum Likelihood Estimation (MLE) method. While a detailed explanation of the Maximum Likelihood Estimation method is omitted as it is a well-known technique, it is a method for deriving values ​​for parameters (the coefficients of logistic regression) that are optimized to maximize the probability obtained from the sample data, i.e., the data of the subjects.

[0178] Therefore, using data indicating age, sex, and smoking history based on the subject's identification information (physical data and smoking data), and the subject's olfactory score, constants A, B, C, and D are derived such that the probability px for subjects with MCI is high, and the probability px for subjects with CN is low. These constants A, B, C, and D are derived for each of the cognitive function tests, MMSE and MoCA-J. Furthermore, these constants A, B, C, and D are derived for each of all possible combinations of one or more of the six types of odor components mentioned above. The derived constants A, B, C, and D are passed to the registration unit 126 in association with the type of cognitive function test and the combination of odor components, and registered in the correlation / regression information database 100B.

[0179] Next, we will evaluate the estimation of cognitive function based on combinations of odor components. For this evaluation, we will use the AUC obtained from the ROC analysis described above as the evaluation index.

[0180] The estimation formula described above makes it possible to calculate the probability px that a subject is predicted to be MCI (Mild Cognitive Impairment). If subjects with a high predicted probability px are MCI, and subjects with a low probability px are CN (Chronic Neoplastic), then the prediction accuracy of the above estimation formula is high. It is possible to distinguish between MCI and CN by setting a threshold called a cutoff value. However, generally, setting a threshold that increases the accuracy of predicting subjects suspected of having MCI as MCI will decrease the accuracy of predicting subjects not suspected of having MCI as CN. On the other hand, setting a threshold that increases the accuracy of predicting subjects not suspected of having MCI as CN will decrease the accuracy of predicting subjects suspected of having MCI as MCI. Therefore, in this embodiment, instead of limiting to a specific threshold, an ROC curve is used to comprehensively evaluate the prediction accuracy. Specifically, the ROC curve in the ROC analysis is derived using the subject data, and the AUC is derived.

[0181] Figure 25 shows an example of an ROC curve. Figure 25 is a graph with sensitivity on the vertical axis and (1 - specificity) on the horizontal axis. Sensitivity represents the proportion of subjects suspected of having MCI who were predicted to have MCI. Specificity represents the proportion of subjects who were not MCI who were predicted to have CN. AUC corresponds to the area below the ROC curve in the graph.

[0182] As the AUC value increases, the prediction accuracy of predicting that a subject is MCI improves. Therefore, here, a group of odor components with an evaluation index (AUC) that fits predetermined evaluation conditions is derived as a combination of odor components to be used in the estimation phase described later. The evaluation conditions are determined by extracting any of the odor component groups whose evaluation index (AUC) exceeds a predetermined value. The objective is to determine the optimal combination of odor components for cognitive function testing. In this embodiment, the odor component group with the highest evaluation index (AUC) is selected.

[0183] Figures 26 and 27 show examples of data illustrating the evaluation index (AUC). Figure 26 is an example of an evaluation table for cognitive function test results using MoCA-J, where the AUC values ​​for each odor component group are arranged in descending order of AUC value. Figure 27 is an example of an evaluation table for cognitive function test results using MMSE, where the AUC values ​​for each odor component group are arranged in descending order of AUC value. In the "Odor Component Combination" column, the odor components are indicated by the identification numbers mentioned above.

[0184] As shown in Figure 26, the odor component group consisting of the two scents "4" mint and "5" socks yielded the highest AUC for the results of the cognitive function test using MoCA-J. Therefore, it can be understood that the odor component group consisting of the two scents "4" mint and "5" socks is the combination of odor components that has the strongest correlation with the test values ​​of the cognitive function test using MoCA-J. Furthermore, as shown in Figure 27, the odor component group consisting of the two scents "2" rag odor and "4" mint yielded the highest AUC for the results of the cognitive function test using MMSE. Therefore, it can be understood that the odor component group consisting of the two scents "2" rag odor and "4" mint is the combination of odor components that has the strongest correlation with the test values ​​of the cognitive function test using MMSE.

[0185] The mint scent component described above is an example of a first scent component predetermined to evoke a feeling of freshness in the subjects. The rag odor and sock odor components are examples of second scent components predetermined to evoke a feeling of unpleasantness in the subjects. Furthermore, the combinations of mint and rag odor components, as well as the mint and sock odor components, are examples of component sets between the first and second scent components.

[0186] In this way, the correlation / regression information derivation unit 125 uses the subject's data (age, sex, smoking history), the results of the odor component test, and the results of the cognitive function test to derive combinations of odor components (in this case, mint and socks, mint and rag odor). It also derives data showing the correlation between the subject's data, odor components, and the odor components and the cognitive function test results.

[0187] It should be noted that the logistic regression model described above is just one example of a generalized linear model, and this disclosure is not limited to using the logistic regression model as an analytical method; other analytical models may also be used. Furthermore, it goes without saying that the derivation of the evaluation index described above is not limited to deriving AUC by performing ROC analysis.

[0188] Next, in step S154, the registration unit 126 registers data showing combinations of odor components corresponding to the type of cognitive function test (for example, mint and socks, and mint and rag odor) in the correlation / regression information database 100B. The registration unit 126 also registers the data showing the correlation described above, i.e., the constants A, B, C, and D in the estimation formula above, in the correlation / regression information database 100B, associating them with the type of cognitive function test and the combination of odor components.

[0189] Therefore, the cognitive function estimation device 10 constructs the correlation / regression information database 100B in the learning phase as described above.

[0190] (Estimated phase) Next, we will describe the cognitive function estimation device 10B, which functions as the estimation phase. In this embodiment, for the sake of simplicity, the cognitive function test will be described as an example of a dementia screening test performed by a licensed physician, and its application to the estimation of the MoCA-J described above will be explained.

[0191] Figure 28 is a block diagram showing the functional configuration in the estimation phase, where cognitive function estimation processing is performed using the constructed correlation and regression information database 100B. The cognitive function estimation device 10B has the function of estimating the subject's cognitive function using a pre-constructed correlation and regression information database 100B as an estimation phase, by executing an estimation processing program. Specifically, the estimation processing program has an olfactory data collection function, a subject data acquisition function, an odor selection function, a correlation and regression information extraction function, a cognitive function evaluation function, and a presentation function. In this estimation phase, the cognitive function estimation device 10B selects odor components that have a correlation with the subject's identification information, including age, sex, and smoking history, and the type of cognitive function test, based on the correlation and regression information database 100B, and estimates cognitive function from the olfactory data obtained from these odor components.

[0192] As shown in Figure 28, during the estimation phase, the CPU 11 of the cognitive function estimation device 10B functions as an olfactory data collection unit 101, a data acquisition unit 122, a scent selection unit 124, an extraction unit 127 that extracts data from the correlation / regression information database 100B, a cognitive function evaluation unit 128, and a presentation unit 130.

[0193] Figure 29 is a flowchart showing the flow of the estimation process in the estimation phase of the cognitive function estimation device 10B. In the estimation phase, the CPU 11 of the cognitive function estimation device 10B executes an estimation processing program, thereby realizing the estimation process shown in Figure 29.

[0194] In the cognitive function estimation device 10B, when the estimation phase for estimating the subject's cognitive function is initiated, the estimation processing program shown in Figure 29 is executed. First, in step S240, the data acquisition unit 122 receives and acquires the subject's data. Here, the subject's data includes physical data indicating the subject's age and gender, and smoking data indicating their smoking history. The data acquired by the data acquisition unit 122 may be information entered in the input unit 16, or information obtained through communication with other devices via the communication unit 18. The data acquired by the data acquisition unit 122 is passed to the odor selection unit 124 and the cognitive function evaluation unit 128.

[0195] Next, in step S242, the odor selection unit 124 obtains data from the correlation / regression information database 100B that shows combinations of multiple odor components (odor component groups). Specifically, the odor selection unit 124 requests the extraction unit 117 to extract data showing combinations of multiple odor components (odor component groups) from the correlation / regression information database 100B. In response to the request from the odor selection unit 124, the extraction unit 127 passes the aforementioned registered data showing combinations of multiple odor components (odor component groups). Here, the odor selection unit 124 obtains data showing odor component groups that are combinations of two types of odor components: mint and socks.

[0196] Next, in step S244, the odor selection unit 124 selects the multiple odor components (odor component group) acquired in step S242 as the multiple odor components (odor component group) to be used for testing. The odor selection unit 124 passes information indicating the selected multiple odor components to the olfactory data collection unit 101. Here, two types of odor components, mint and socks, are selected.

[0197] In step S244, the olfactory data collection unit 101 releases the odor component to the subject according to the information indicating the odor component passed from the odor selection unit 124, similar to step S204 described above. The olfactory data collection unit 101 then collects data indicating the subject's olfactory ability to the released odor component.

[0198] Next, the olfactory data collection unit 101 proceeds in the same manner as in step S206 described above, in step S2 In step 46, the olfactory score of the odor, based on the subject's sense of smell, is derived and recorded, indicating the selection result of the correct image selected by the subject in accordance with the release conditions (type and concentration) of the odor components. The olfactory score derived by the olfactory data collection unit 101 is passed to the cognitive function evaluation unit 128. The olfactory data collection unit 101 is an example of a component release control unit.

[0199] In step S248, the cognitive function evaluation unit 128 acquires data showing the correlation between the olfactory score and the cognitive function test results. Specifically, the cognitive function evaluation unit 128 requests the extraction unit 127 to provide the data showing the estimation formula described above. In response to the request from the cognitive function evaluation unit 128, the extraction unit 127 provides the estimation formula and the data showing the estimation formula. The cognitive function evaluation unit 128 then acquires the estimation formula registered in the correlation / regression information database 100B from the extraction unit 127, and data showing the constants of the estimation formula.

[0200] Next, in step S250, the cognitive function evaluation unit 128 derives a cognitive function evaluation value (probability px). First, it sums the olfactory scores for each odor component derived by the olfactory data collection unit 101 to derive a total score. Next, using data indicating the subject's age, sex, and smoking history, the total score, and the estimation formula, it derives a cognitive function evaluation value for the subject, i.e., the value of probability px. The evaluation value derived by the cognitive function evaluation unit 128 is passed to the presentation unit 130.

[0201] Then, in step S252, the presentation unit 130 presents information regarding cognitive function, similar to step S212 described above. For example, the presentation unit 110 performs display control to display the value of probability px, which is an evaluation value, on the display unit 17 as information regarding the cognitive function test. The data displayed on the display unit 17 can indicate that there is no risk of cognitive decline if the probability px is below a predetermined threshold, and that there is a risk of cognitive decline if the probability px exceeds a predetermined threshold. The data regarding the degree (magnitude) of the risk of cognitive decline should be such that the degree (magnitude) of the risk of cognitive decline increases as the value of probability px increases.

[0202] Therefore, in the estimation phase, the cognitive function estimation device 10B estimates the subject's cognitive function using the correlation / regression information database 100B as described above.

[0203] Furthermore, when performing estimation processing according to the type of cognitive function test, the extraction unit 127 should acquire each data of odor component combinations (odor component groups) associated with the type of cognitive function test. The olfactory data collection unit 101 should derive an olfactory score for each odor component combination (odor component group) for each type of cognitive function test, and pass this olfactory score to the cognitive function evaluation unit 128, associated with the type of cognitive function test. The cognitive function evaluation unit 128 should then derive each evaluation value (probability px) using an estimation formula, i.e., a constant, for each type of cognitive function test, and pass this evaluation value (probability px) to the presentation unit 130, associated with the type of cognitive function test.

[0204] As described above, the cognitive function device according to this embodiment selects odor components that correlate with cognitive function tests based on the correlation between data indicating the subject's age, sex, and smoking history, as well as the olfactory score of the odor components and the results of the cognitive function test. Furthermore, an estimation formula is derived as information indicating this correlation, and the estimation formula is registered in the correlation / regression information database. By using this estimation formula, it becomes possible to derive an evaluation value (probability) indicating the likelihood of cognitive decline based on data indicating the subject's age, sex, and smoking history, and the sum of the olfactory scores (total score) of odor components related to cognitive function.

[0205] [Other forms] Although the above has described the technology of this disclosure in detail with respect to specific embodiments, the technology of this disclosure is not limited to these embodiments, and various other embodiments are possible within the scope of the technology of this disclosure. It is possible to implement it in various forms. For example, the configuration of the cognitive function estimation device described in the above embodiment is just one example, and may be modified according to the situation without departing from the main purpose.

[0206] Furthermore, the program processing flow described in the above embodiment is just one example, and unnecessary steps may be deleted, new steps added, or the processing order rearranged, as long as it does not deviate from the main purpose.

[0207] Furthermore, although the above embodiment describes processing performed by executing a program stored in the memory unit, the program processing may also be implemented in hardware.

[0208] Furthermore, while the above embodiments describe a configuration in which the program is pre-stored (installed) in memory such as ROM, the invention is not limited to this. The program may be provided in the form of a recording medium such as a CD-ROM (Compact Disc Read Only Memory), DVD-ROM (Digital Versatile Disc Read Only Memory), or USB memory. Alternatively, the program may be provided in the form of a download from an external device via a network.

[0209] In the above embodiment, "processor" refers to a processor in a broad sense, including general-purpose processors (e.g., CPUs) and dedicated processors (e.g., GPUs: Graphics Processing Units, ASICs: Application Specific Integrated Circuits, FPGAs: Field Programmable Gate Arrays, programmable logic devices, etc.). Furthermore, the operation of the processor in the above embodiment may not be performed by a single processor, but may be performed by multiple processors located in physically separate locations working together. In addition, the order of each operation of the processor is not limited to the order described in the above embodiment and may be changed. [Explanation of Symbols]

[0210] 10. Cognitive function estimation device 10X Computer Unit 11 CPU 12 RAM 13 ROM 14 Input / Output Interfaces 15 bus 16 Input section 17 Display section 18 Communications Department 19 Memory section 20 Odor component release device 100 Correlation and Average Information Database 101 Olfactory Data Collection Department 102 Body Data Acquisition Unit 103 Cognitive Function Test Results Acquisition Department 104 Scent Selection Department 105 Correlation and Average Information Derivation Unit 106 Registration Department 107 Extraction part 108 Cognitive Function Assessment Department 110 Presentation section

Claims

1. A fragrance component release unit that releases each of several fragrance components to the target person, with the concentration of each fragrance component changing in multiple stages, A score setting unit sets a score corresponding to the concentration indicated by the correct answer when the subject's response indicating which odor component was released from the odor component release unit is correct. An estimation unit estimates an evaluation value of cognitive function for the subject based on the sum of the scores for the multiple odor components set by the score setting unit, A cognitive function estimation device equipped with the following features.

2. The aforementioned fragrance components are the respective fragrance components of vanilla, mint, yellow peach, and socks. The cognitive function estimation device according to claim 1.

3. The estimation unit uses constants a and b, Cognitive function assessment value = a × total score + b The cognitive function evaluation value based on the formula expressed is estimated as the aforementioned cognitive function. The cognitive function estimation device according to claim 1.

4. The aforementioned score is set to a higher value as the concentration of the correct answer decreases. The cognitive function estimation device according to claim 1.

5. The system further includes an odor component setting unit that sets an odor component group based on a correspondence between the test values ​​of a predetermined cognitive function test administered to multiple subjects and the scores of each odor component when each of the multiple odor components is released to the multiple subjects, and which has a correlation with the cognitive function test and consists of odor components with fewer components than the number of components in the multiple odor components. The odor component release unit releases the odor component group set in the odor component setting unit as the plurality of odor components. The cognitive function estimation device according to claim 1.

6. The odor component setting unit sets a group of odor components as the odor component group, where the information regarding the discriminative ability evaluation value obtained by analyzing the distribution of subjects whose cognitive function test results exceed a predetermined test value and subjects whose cognitive function test results are below a predetermined test value exceeds a predetermined correlation value, based on the total value of the multiple odor components for the multiple subjects. The cognitive function estimation device according to claim 5.

7. The estimation unit estimates a cognitive function evaluation value for the subject based on the correlation between the sum of the scores for the odor components for multiple subjects and the ratio of subjects whose cognitive function test results, obtained by administering a predetermined cognitive function test to multiple subjects, exceed a predetermined test value and those whose test results are below a predetermined test value. The cognitive function estimation device according to claim 1.

8. The system further includes a display unit that displays the cognitive function evaluation values ​​for the subject estimated by the estimation unit, The estimation unit provides recommendation information recommending that the user undergo a cognitive function test if the cognitive function evaluation value exceeds a predetermined threshold. The cognitive function estimation device according to claim 1.

9. A program that causes a computer to perform a process to estimate the cognitive function of a subject, The aforementioned computer, A score setting unit sets a score corresponding to the concentration indicated when the subject correctly identifies which of several odor components was released from an odor component release unit, which releases each of several odor components at multiple levels of varying concentrations. Estimation unit estimates cognitive function evaluation value for the subject based on the sum of the scores for the multiple odor components set by the score setting unit. A program designed to function as such.

10. An acquisition unit acquires identification information including physical data indicating the age and gender of a subject for identifying a subject for whom cognitive function is to be estimated, and smoking data indicating whether or not the subject has a history of smoking. A fragrance component release unit that releases each of several fragrance components to the target person, with the concentration of each fragrance component changing in multiple stages, A score setting unit sets a score corresponding to the concentration indicated by the correct answer when the subject's response indicating which odor component was released from the odor component release unit is correct. An estimation unit estimates an evaluation value of cognitive function for the subject based on the identification information acquired by the acquisition unit and the sum of the scores for the multiple odor components set by the score setting unit. A cognitive function estimation device equipped with the following features.

11. The estimation unit, For multiple subjects, an assessment value of cognitive function for each subject is estimated based on the correlation between the identification information, the type of odor component and the test value of that odor component, and the test value of the cognitive function test. The cognitive function estimation device according to claim 10.

12. The estimation unit, Let A, B, C, and D be constants. Let pa be (A × total score), Let pb be (B × age), Let PC be (C x gender), If PD is defined as (D × smoking history), p=1 / [1+exp{-(pa+pb+pc+pd)}] The estimation formula expressed by [formula] is used to derive an assessment value for cognitive function. The cognitive function estimation device according to claim 11.

13. The system further includes an odor component setting unit that sets an odor component group based on a correspondence between the test values ​​of a predetermined cognitive function test administered to multiple subjects and the scores of each odor component when each of the multiple odor components is released to the multiple subjects, and which has a correlation with the cognitive function test and consists of odor components with fewer components than the number of components in the multiple odor components. The odor component release unit releases the odor component group set in the odor component setting unit as the plurality of odor components. The cognitive function estimation device according to claim 10.

14. The odor component setting unit sets a group of odor components as the odor component group, where the information regarding the discriminative ability evaluation value obtained by analyzing the distribution of subjects whose cognitive function test results exceed a predetermined test value and subjects whose cognitive function test results are below a predetermined test value exceeds a predetermined correlation value, based on the total value of the multiple odor components for the multiple subjects. The cognitive function estimation device according to claim 13.

15. The odor components released from the odor component release unit include at least a set of components consisting of a first odor component predetermined as an odor component that the subject will find refreshing, and a second odor component predetermined as an odor component that the subject will find unpleasant. The cognitive function estimation device according to claim 10.

16. The first odor component is mint, and the second odor component is an odor component that smells like a dirty rag or a sock. The cognitive function estimation device according to claim 15.

17. Computers, An acquisition unit that acquires identification information including physical data indicating the age and gender of a subject for identifying a subject for whom cognitive function is to be estimated, and smoking data indicating whether or not the subject has a history of smoking. A component release control unit controls a component release unit that releases each of several odor components to the target person, with the concentration of each odor component changing in multiple stages, and A score setting unit sets a score corresponding to the concentration indicated by the correct answer when the subject's response indicating which odor component was released from the odor component release unit is correct. An estimation unit estimates an evaluation value of the cognitive function of the subject based on the identification information acquired by the acquisition unit and the sum of the scores for the multiple odor components set by the score setting unit. A program designed to function as such.

Images