JP7366368B2 - Information processing method, information processing device, program, recording medium - Google Patents

Description

Application of Article 30, Paragraph 2 of the Patent Act (1) Meeting name/34th Japan Geriatric Psychiatry Society Conference Date/June 8, 2019 (2) Meeting name/The Alzheimer's Association International Conference 2019 (AAIC2019) Publication date: July 16, 2019 (3) Publication address: http://dementia. umin. jp/program38. pdf Publication date/October 15, 2019 (Reiwa 1) (4) Meeting name/38th Japanese Dementia Society Academic Meeting Date/November 8, 2019 (Reiwa 1) (5) Publication Name of publication/Sankei Shimbun Date of publication/December 14, 2019 (6) Name of publication/Nikkan Kogyo Shimbun Date of publication/January 22, 2020 (7) Publication address/https://www. meijiyasuda. co. jp/profile/news/release/2019/pdf/20200124_01. pdf Publication date/January 24, 2020 (8) Distribution material/pamphlet "MY Health Report" Distribution location/Meiji Yasuda Life Insurance Company Central Tokyo Headquarters, etc. Distribution date/2020 ( January 24, 2020 (9) Broadcast program / TV Asahi Super J Channel Broadcast date / January 24, 2020 (10) Publication name / Evening Fuji Publication date / 2020 ( February 4, 2020 (11) Broadcast program/NHK News Good Morning Japan Broadcast date/February 12, 2020 (12) Distribution materials/Pamphlet "Dementia Care MCI Plus" Distribution location/ Meiji Yasuda Life Insurance Company Central Tokyo Headquarters, etc. Date of distribution: February 14, 2020 (13) Distribution materials/Usage guide “Easy Brain Check User Guide February 2020” Distribution location: Meiji Yasuda Mutual Life Insurance Company Tokyo Headquarters, etc. Distribution date: February 26, 2020 (Reiwa 2)

The present invention relates to an information processing method, an information processing device, a program, and a recording medium for supporting diagnosis of mild cognitive dysfunction.

Mild Cognitive Impairment (hereinafter referred to as MCI) refers to a condition that cannot be called dementia but is not intellectually normal. It is known that not all MCI patients progress to dementia, and that a certain percentage of patients recover with cognitive training and the latest medical technology. Early detection of MCI is extremely important as it may lead to fundamental treatment for the patient.

Conventionally, the diagnosis of MCI has generally been carried out by a specialist in a hospital over a full day, and the burden on the patient is large, as the patient's cerebrospinal fluid must be collected. Therefore, there is a need for a new approach that can easily detect signs of cognitive decline with less burden on patients. In Patent Document 1, it is determined whether a subject has MCI based on whether the subject's answers to questions are correct or incorrect.

JP 2014-8329 Publication

However, the above-mentioned determination method performs a screening test using a conventionally known testing method. The questions used in this test are randomly selected without considering the responses of subjects who have taken screening tests in the past. Therefore, there is a problem that the accuracy of determining MCI cannot be sufficiently increased, and patients with MCI are not detected.

The present invention has been made in view of the above points, and one aspect of the present disclosure is an information processing method executed by a processor that executes instructions stored in a storage unit, the method comprising:
A step of displaying the text of a test question determined to have a statistically significant difference in order to determine the user's cognitive level, the cognitive level being between a healthy person level and a mild cognitive impairment (MCI) level. , a step including a dementia level;
Accepting the user's answer to the problem statement,
This is an information processing method that determines whether or not the user has MCI by inputting feature amounts calculated based on the user's answer results into a learned model.

Another aspect of the present disclosure is an information processing method executed by a processor that executes instructions stored in a storage unit, the method comprising:
The target variable is a correct label indicating whether the subject is healthy, has MCI, or has dementia,
The model uses machine learning to calculate the features as a score calculated based on the correctness of each subject's answer to the question in the answer results obtained from multiple subjects, and the response time until receiving the response to the question from the subject. to generate a trained model,
It is an information processing method.

Another aspect of the present disclosure is an information processing device including a storage unit that stores instructions and a processor,
The processor executes instructions stored in the storage unit, thereby
In order to determine the user's cognitive level, the text of the test question determined to have a statistically significant difference is displayed, and the cognitive level includes a healthy level, an MCI level, and a dementia level. ,
Accepting the user's answer to the problem statement,
Inputting feature amounts calculated based on the user's answer results into a trained model to determine whether the user is MCI;
It is an information processing device.

Further, another aspect of the present disclosure is a program executed by a processor, comprising:
When the processor executes the program,
In order to determine the user's cognitive level, the text of the test question determined to have a statistically significant difference is displayed, and the cognitive level includes a healthy level, an MCI level, and a dementia level. ,
Accepting the user's answer to the problem statement,
Inputting feature amounts calculated based on the user's answer results into a trained model to determine whether the user is MCI;
It is a program.

Another aspect of the present disclosure is a recording medium storing a program executed by a processor,
By the processor executing a program recorded in the memory,
In order to determine the user's cognitive level, the text of the test question determined to have a statistically significant difference is displayed, and the cognitive level includes a healthy level, an MCI level, and a dementia level. ,
Accepting the user's answer to the problem statement,
Inputting feature amounts calculated based on the user's answer results into a trained model to determine whether the user is MCI;
It is a recording medium.

1 is an MCI diagnosis support system according to an embodiment of the present disclosure. FIG. 2 is a block diagram showing the functional configuration of the server shown in FIG. 1. FIG. The operation flow (first phase) of the server shown in FIG. 2 is illustrated. This is an example of a histogram showing the feature amount (average score) for each group of questions. This is an example of a histogram showing the feature amount (average response time) for each group of questions. FIG. 2 is a block diagram showing the functional configuration of the user terminal shown in FIG. 1. FIG. The operation flow (second phase) of the user terminal shown in FIG. 5 is illustrated. The details of the steps for level determination shown in FIG. 6 are shown. The screen of a user terminal for a screening test is illustrated. The screen of a user terminal for a screening test is illustrated. The screen of a user terminal for a screening test is illustrated. The screen of a user terminal for a screening test is illustrated. The screen of a user terminal for a screening test is illustrated. The screen of a user terminal for a screening test is illustrated. The screen of a user terminal for a screening test is illustrated. The screen of a user terminal for a screening test is illustrated. The screen of a user terminal for a screening test is illustrated. The screen of a user terminal for a screening test is illustrated. The screen of a user terminal for a screening test is illustrated. The screen of a user terminal for a screening test is illustrated. The screen of a user terminal for a screening test is illustrated. The screen of a user terminal for a screening test is illustrated. The screen of a user terminal for a screening test is illustrated. The screen of a user terminal for a screening test is illustrated. The screen of a user terminal for a screening test is illustrated.

[Description of embodiments of the present disclosure]
First, the contents of the embodiments of the present disclosure will be listed and explained. One embodiment of the present disclosure includes the following configuration.

(Item 1) An information processing method executed by a processor that executes instructions stored in a storage unit,
A step of displaying the text of a test question determined to have a statistically significant difference in order to determine the user's cognitive level, the cognitive level being between a healthy person level and a mild cognitive impairment (MCI) level. , a step including a dementia level;
Accepting the user's answer to the problem statement,
An information processing method that determines whether or not the user has MCI by inputting feature amounts calculated based on the user's answer results into a learned model.

(Item 2) The information processing method described in Item 1,
The determined test questions are divided into multiple items,
Each item consists of selecting a memorized word at a later time, selecting a letter, moving the user's terminal as instructed, answering the date and time orientation, answering a memorized number, and recognizing a plane figure. An information processing method, wherein one of the plurality of questions included in the item is the question determined to be statistically significant.

(Item 3) The information processing method described in Item 2, further comprising:
The scores of the healthy group, the MCI group, and the dementia group were obtained from the answers of the plurality of subjects to the question sentences of the plurality of questions, and the correlation between the obtained scores and the cognitive level was determined. the step of determining that a certain problem is a problem with the statistically significant difference, and when the score of the MCI group is between the score of the healthy group and the score of the dementia group, the cognitive An information processing method that determines that the problem has a correlation with the level.

(Item 4) The information processing method described in Item 2, further comprising:
From the results of the test subjects' answers to the question sentences of the plurality of questions, the response times of the healthy group, the MCI group, and the dementia group were obtained, and the relationship between the obtained response times and the cognitive level was calculated. The step of determining a question with a high correlation as a question with the statistically significant difference, wherein the response time of the MCI group is between the response time of the healthy group and the response time of the dementia group. An information processing method in which it is determined that the question has a high correlation with the cognitive level.

(Item 5) The information processing method according to any one of items 1 to 4, wherein the feature amount is a score calculated based on the correctness of the answer obtained from the user's answer to the question statement. and a response time until receiving a reaction from the user to the question statement.

(Item 6) An information processing method executed by a processor that executes instructions stored in a storage unit,
The target variable is a correct label indicating whether the subject is healthy, has MCI, or has dementia,
The model uses machine learning to calculate the features as a score calculated based on the correctness of each subject's answer to the question in the answer results obtained from multiple subjects, and the response time until receiving the response to the question from the subject. to generate a trained model,
Information processing method.

(Item 7) The information processing method according to claim 6, wherein the feature amount further includes the user's age and the user's gender.

(Item 8) An information processing device comprising a storage unit that stores instructions and a processor,
The processor executes instructions stored in the storage unit, thereby
In order to determine the user's cognitive level, the text of the test question determined to have a statistically significant difference is displayed, and the cognitive level includes a healthy level, an MCI level, and a dementia level. ,
Accepting the user's answer to the problem statement,
Inputting feature amounts calculated based on the user's answer results into a trained model to determine whether the user is MCI;
Information processing device.

(Item 9) A program executed by a processor,
When the processor executes the program,
In order to determine the user's cognitive level, the text of the test question determined to have a statistically significant difference is displayed, and the cognitive level includes a healthy level, an MCI level, and a dementia level. ,
Accepting the user's answer to the problem statement,
Inputting feature amounts calculated based on the user's answer results into a trained model to determine whether the user is MCI;
program.

(Item 10) A recording medium storing a program executed by a processor,
By the processor executing a program recorded in the memory,
In order to determine the user's cognitive level, the text of the test question determined to have a statistically significant difference is displayed, and the cognitive level includes a healthy level, an MCI level, and a dementia level. ,
Accepting the user's answer to the problem statement,
Inputting feature amounts calculated based on the user's answer results into a trained model to determine whether the user is MCI;
recoding media.

[Details of embodiments of the present disclosure]
Embodiments of the present disclosure will be described below with reference to the drawings. In the drawings, the same or similar elements are denoted by the same or similar reference numerals, and redundant description of the same or similar elements may be omitted in the description of each embodiment. Furthermore, features shown in each embodiment can be applied to other embodiments as long as they do not contradict each other. However, embodiments of the present disclosure are not necessarily limited to such aspects. It will be apparent to those skilled in the art that the embodiments of the present disclosure may take on various aspects within the scope of the claims.

According to the present disclosure, a plurality of subjects are divided into a group of healthy subjects without impaired cognitive function, an MCI group with MCI, and a dementia group with dementia, and each group is given a test using a user terminal. implement. Based on the results of this test, questions that have a significant correlation with MCI are selected. Questions that have a significant correlation with MCI are those that, when the same question is asked between a healthy group and an MCI group, or between an MCI group and a dementia group, there is a significant correlation between the response results (response time and scores) between each group. This is a problem where differences emerge. Next, a learned model that can determine the MCI is generated from the past answer results. Furthermore, using the generated trained model, a screening test is performed on new users to determine whether they have MCI.

Embodiments of the present invention will be described below with reference to the drawings. In FIG. 1, an MCI diagnosis support system 10 includes a user terminal 100 and a server 200 that supports MCI diagnosis. The server 200 selects questions that have a significant correlation with MCI, and stores the results of MCI screening tests conducted on subjects based on the selected questions. The user terminal 100 and the server 200 are connected to be able to communicate with each other via a communication network 12 such as the Internet or a public communication network.

The user terminal 100 is a terminal as an information processing device used by a user undergoing a screening test, and is, for example, a tablet terminal or a smartphone having a touch panel 106. Hereinafter, the user terminal 100 will be explained as a smartphone. The user terminal 100 may be installed in a medical facility, or may be a terminal owned by the user himself/herself. The user terminal 100 is configured to download an application for an MCI screening test from a server 200 that supports MCI diagnosis, provide the MCI screening test, and send the response results to the server 200. can do. As illustrated, the user terminal 100 includes a processor 102, a memory 104 as a recording medium, a touch panel 106, an acceleration sensor 108 that detects movement of the user terminal 100, and a communication interface (communication IF) 110 as hardware elements. and a speaker 114. These elements are communicably connected to each other via a bus (not shown).

The memory 104 stores at least an operating system. The operating system is a computer program for controlling the overall operation of user terminal 100. When each process of the computer test is executed on the user terminal 100, the memory 104 further stores a program for an MCI screening test. The test program is a computer program for causing the user terminal 100 to execute each process of the screening test. Memory 104 can also temporarily or permanently store data handled or generated in each process of a screening test. Specific examples of the memory 104 include main storage such as RAM (Random Access Memory), and auxiliary storage such as a hard disk, flash memory, optical disk, and ROM (Read Only Memory). Data related to various tests are downloaded from the server 200 and stored in the auxiliary storage, and expanded in the main storage to be used as appropriate when various programs are executed by the processor 102.

Note that when the test program is stored in the server 200 instead of the user terminal 100, the server 200 executes each process of the computer test, and the user terminal 100 simply displays the test image based on the signal received from the server 200. , may be configured to function as a browser that accepts user input on this test screen.

The processor 102 is configured to read the program stored in the memory 104 as described above and execute processing according to the program. As an example, when the processor 102 executes a test program stored in the memory 104, each function of the test described below is realized. The processor 102 includes a CPU (Central Processing Unit) and a G
Includes PU (Graphics Processing Unit).

The touch panel 106 is integrated with a display that provides visual information to the user. The touch panel 106 receives input from the user to operate the user terminal 100, and displays various screens on the display in response to the input. The screen is equipped with an operation function using a GUI (Graphical User Interface).

Communication interface (IF) 112 is a network interface for communicating with external server 200 via communication network 12. The communication interface 112 is configured to send and receive data to and from the server 200 according to a predetermined communication protocol including, for example, TCP/IP (Transmission Control Protocol/Internet Protocol).

The server 200 stores the results of screening tests conducted on multiple subjects in the past. The server 200 is a server that mainly selects questions that have a significant correlation with the MCI from past answer results, and further generates a trained model for determining the MCI from the answer results. Ru. The server 200 may be provided in the cloud, or may be provided at a local location such as a system user's building or site. Furthermore, the server 200 may be configured with one server or multiple servers. The server 200 includes at least a processor 202, a memory 204 including main memory and auxiliary memory, and a communication interface 206 as hardware elements. Detailed description of elements similar to those of the user terminal 100 already described will be omitted.

In the present disclosure, the MCI diagnosis support system 10 selects questions in which there is a statistically significant difference between each group from the response results for each of the healthy person group, MCI group, and dementia group. Next, based on the results of each group's answers to the selected questions, a trained model for determining whether it is MCI is generated (hereinafter referred to as the first phase).

Furthermore, the MCI diagnosis support system 10 presents the problem selected in the first phase to the new user, and determines whether the new user has MCI based on the trained model generated in the first phase (hereinafter referred to as , referred to as the second phase). Hereinafter, the first phase will be explained first, and then the second phase will be explained.

<First phase>
Here, we will explain the first phase from selecting questions with statistically significant differences in answer results to generating a trained model. FIG. 2 is a block diagram showing the functional configuration of the server 200 shown in FIG. 1. The server 200 includes a processing section 210 and a storage section 230. The processing unit 210 further includes an answer acquisition unit 212, a feature value calculation unit 214, a question selection unit 216, and a model generation unit 218. Storage unit 230 corresponds to memory 204 shown in FIG. The processing unit 210 and each unit 212 to 218 included in the processing unit 210 perform problem selection processing and learning according to the present disclosure, which is realized by the processor 202 shown in FIG. 1 reading and executing various programs in the memory 204. Represents the functionality realized by the completed model generation process.

The storage unit 230 stores a plurality of questions classified by item and the test subject's past answers to these questions as sample data. In the sample data, subjects are classified in advance into a healthy group, an MCI group, and a dementia group. The questions used in the screening test are, for example, questions based on the internationally used Mini-Mental State Examination (MMSE), etc., and include questions such as orientation regarding time and place, memorization of selecting several words repeatedly, It includes various items such as attention and calculation, selecting and playing some memorized words, understanding and carrying out given commands, and selecting missing shapes. The questions used in the screening test may further include questions of operating the user terminal 100 (for example, shaking the user terminal 100, turning over the user terminal 100, etc.) according to an instructed command. Various sensors (for example, acceleration sensors) acquire the behavior of the user terminal, and extract feature amounts from the acquired behavior of the user terminal. In addition, each item includes a plurality of questions expressed in different formats, for example, in different sentences. Subjects may also be associated with age and gender. The past answer results include, for example, the answer time until answering the presented question and the correctness of the answer.

The operation flow 300 (first phase) of the server 200 shown in FIG. 2 will be described below with reference to FIG. That is, the process from selecting the optimal question for the MCI screening test to generating a trained model for determining MCI will be described below.

In this embodiment, it is assumed that a screening test has been conducted in advance on a plurality of users who are test subjects at the user terminal 100. Prior to step 302, the user terminal 100 displays a question for determining MCI on the display of the touch panel 106, receives an answer input from the user on the touch panel 106, and stores the input information in advance in the memory 104 as an answer result. It is stored as . Each subject was classified into a healthy group, an MCI group, and a dementia group.

In step 302, the server 200 (answer acquisition unit 212) acquires each test subject's answers to the test questions from the plurality of user terminals 100. One test includes multiple items, and each item has multiple questions. The answer results include a plurality of questions, the response time until the user answers the questions, and whether the answers to the questions are correct or incorrect for each item and each question.

Table 1 illustrates test items A to F of the screening test administered to the subjects. As shown in Table 1, the test items include, for example, item A in which the subject answers what he or she has memorized over time, item B in which the subject selects letters in a predetermined order, and item C in which the user terminal 100 is moved according to instructions. , item A for answering the time and place, item E for memorizing numbers, and item F for choosing a figure. The test items are not limited to those shown in the figure, and may include other items included in the generally known Mini-Mental State Examination (MMSE).

According to the present disclosure, it has been found that depending on the way the questions are presented, there are some questions that have a significant correlation with MCI and some questions that do not. Therefore, it is necessary to select the most appropriate way to ask questions from among multiple ways to ask questions. An appropriate way to ask questions is to use a combination of questions that can accurately distinguish between the MCI group and the dementia group with a small number of questions. Each item A to F shown in Table 1 includes multiple ways of asking the question.

In one example, in item A where you are asked to answer what you have memorized after a certain period of time, the question sentence "Please answer with the word you just memorized" is displayed in question A1, and the correct answer is selected from multiple words including the correct word. Let them choose a word. In problem A2, the question sentence ``Please select a word that is not included in the word you just memorized'' is displayed, and the student is asked to select a word that he or she does not remember from among multiple words that include the word that he or she memorized earlier and the word that he or she does not remember. . In question A3, the question sentence ``Please select the picture you just memorized'' is displayed, and the correct picture is selected from a plurality of pictures including the picture (here, symbol) that you memorized earlier. In question A4, the question text "Please answer with the picture you memorized earlier" is displayed, and the correct picture is selected from a plurality of pictures including the picture you memorized earlier (animal in this case).

In addition, in item B where you select letters in order, for example, problem B1 displays numbers that start from 1 and increases by 1, and you are asked to select these numbers in ascending order from 1, and problem B2 starts with a number other than 1 and increases by 1. In problem B3, hiragana characters starting with letters other than ``a'' are displayed and students are asked to select these hiragana in order. In problem B4, they have to select hiragana characters in order, starting with ``a''. Hiragana characters are displayed, and the user is asked to select these hiragana characters in order, starting with "a".

Also, in item C, which moves the device according to instructions, for example, in question C1 the question statement ``Please shake the device immediately when you see the blue circle'' is displayed, and in question C2 the question statement ``Please touch the screen immediately when it vibrates'' is displayed. (The user terminal 100 has a vibration function), and in problem C3, the problem sentence ``Please shake the device immediately when the sound sounds'' is displayed, and in problem C4, two instructions are displayed as the problem sentence ``If the sound sounds, shake the device immediately.'' "Please shake the device immediately. If you see a blue circle, please touch the screen immediately." is displayed first, then one of the instructions (sound or blue circle) is displayed, and the user moves the device according to the instructions given.

In addition, for item D that asks about time and place, for example, question D1 has the question ``What month and day is it today?'', question D2 has the question ``What day of the week is it today?'', and question D3 has the question a few days ago. In question D4, a question sentence asking about the day of the week, ``What day of the week was 2 days ago?'', and a question sentence asking about the day of several days ago, ``What month and day was 2 days ago?'' is displayed.

In addition, in item E where you have to answer memorized numbers, for example, in question E1, the message "Please press the memorized numbers in order" is displayed on the display, and the audio of the question is played, and in question E2, the question is displayed on the display. In question E3, only the audio of the question sentence is played, and in question E4, ``Please put the numbers you memorized in reverse order.''

In addition, in item F for selecting figures, for example, problem F1 asks you to select a piece to fill a holed image, problem F2 displays a diagram of multiple cubes stacked up, and asks you to select the number of cubes. included.

In this way, even when a question is about one item, there are various ways to pose the question, and the difficulty level of the question varies greatly depending on the way the question is presented. If the difficulty level of the question is too high, the variation among individuals will be large, and if it is too low, the variation among each individual will be small, and no significant correlation will appear between the answers of each group. Therefore, in order to set an appropriate difficulty level for a question, questions that have a significant correlation with MCI are selected from the test subject's answers to multiple questions included in one item.

First, in step 302, answers to each question are obtained. The answer results include whether the answer to the question is correct or incorrect and the response time until the answer is accepted.

Next, in step 304, the feature calculation unit 214 calculates a score for each subject based on the answer results for each question. Table 1 exemplifies a method for calculating scores from the answer results for each item.

For example, the score is calculated by calculating the number of correct answers and the number of taps for each subject for the above-mentioned questions A1 to A4 among the items to be answered after a certain period of time in item A, and subtracting the number of taps from the number of correct answers (for example, It can be calculated by adding 1 point for each correct answer and subtracting 1 point for each tap. If you tap several times, you will eventually select the correct word, so the score will be higher if you select the correct word with fewer taps.

In another example, the score is calculated as 1 point for each subject who answers correctly for the above-mentioned questions D1 to D4 among the time and place answer items of item D, and 0 points when the answer is incorrect. be able to. In addition, for items E1 to E4 of the number memorization items, depending on the number memorized by each subject, for example, if the number memorized is 3, then 3 points, if 0, 0 points. Scores can be calculated.

Further, for example, the response time obtains the response time until input by the subject is obtained for questions D1 to D4 among the items for answering the time and place of item D. Furthermore, the feature amount calculation unit 214 normalizes the response times so that the average is 0 and the standard deviation is 1, and obtains the normalized response times of each subject.

Next, in step 306, the question selection unit 216 determines whether there is a significant correlation with the MCI based on the feature values calculated from the answer results (for example, the average score of each group and the average response time of each group). We will select issues that are acceptable. FIG. 4A shows an example of a histogram showing the feature amount (average score) for each group of questions, and FIG. 4B shows an example of a histogram showing the feature amount (average answer time) for each group of questions. . The score or response time should change depending on the progress of dementia. That is, the score and response time of the MCI group should be between those of the healthy group and the dementia group. Therefore, as shown in FIG. 4, problem H is a problem in which a significant correlation with MCI is recognized because the score of the MCI group is between the healthy group and the dementia group. On the other hand, Problem I has the highest score in the MCI group, so it is a problem for which no significant correlation with MCI is recognized. Furthermore, since the response time of the MCI group falls between those of the healthy group and the dementia group, Question J is a question that has a significant correlation with MCI. On the other hand, since problem K has the highest score in the MCI group, it is a problem for which no significant correlation with MCI is recognized. In this way, a question for which the score (response time) of the MCI group is between the scores (response time) of the healthy group and the dementia group is defined as a question that has a significant correlation with MCI.

According to the present disclosure, based on the results of a screening test for subjects (sample data), in item A, question A1 "Please answer with the word you memorized earlier" is displayed, and the correct answer is selected from multiple words including the correct word. Select a word, but in item B, question B3 displays hiragana that start with a letter other than ``a'' and asks you to select these hiragana in order, in item C, question C3 ``shake the device as soon as it sounds'', and in item D Question D4: ``What month and day was it two days ago?'', item E displays the question sentence from question E2, ``Please press the numbers you memorized in order'', and item F displays the perforated image from question F1. The problem that asked students to choose which piece to fill in was selected as the problem that had the most significant correlation with MCI.

A plurality of questions are included for each test item, and an example of a method for selecting a question that is recognized to have a high significant correlation with the MCI among the plurality of questions has been described above. However, the same procedure can be used for test items to select items that have a significant correlation with MCI. Table 1 lists the six items selected as having the highest significant correlation with MCI.

Next, in step 308, the model generation unit 218 analyzes the past answers of multiple subjects (sample data) and determines the new user's cognitive level (healthy person level, MCI level, dementia level). Create a model for A model for determining the user's cognitive level is created based on features obtained from past answers by the test subject to the selected questions. The feature quantity x includes the score obtained from the past answer results in step 304 and the normalized answer time. The feature quantity x may further include the subject's age and gender. If there are multiple questions (six questions in this disclosure), the feature values of score and response time are given to the model for each question. These are input into the model as feature quantities, along with the correct labels of the subjects classified into each group in advance, and the model is caused to perform machine learning. As an example, the correct label is a cognitive level "0" if the subject is healthy, a cognitive level "1" if the subject has MCI, and a cognitive level "2" if the subject has dementia. The test subject's correct label is compared with the output value from the model, the model is repeatedly modified, and the model is trained until the output value becomes close to the correct label to generate a trained model.

<Second phase>
Next, an MCI test is performed using the trained model generated for a new user to be tested. FIG. 5 is a block diagram showing the functional configuration of the user terminal 100 included in the MCI diagnosis support system 10 shown in FIG. User terminal 100 includes a processing section 510 and a storage section 530. The processing unit 510 further includes a question setting unit 512, an answer acquisition unit 514, an end determination unit 516, and a level determination unit 518. Storage unit 530 corresponds to memory 104 shown in FIG. The processing unit 510 and each unit 512 to 516 included in the processing unit 510 perform the MCI diagnosis processing according to the present disclosure, which is realized by the processor 102 shown in FIG. 1 reading out and executing the learned model program in the memory 104. represents the function realized by

Hereinafter, with reference to FIG. 6, an operation flow 600 of the user terminal 100 for diagnosing whether or not the user operating the user terminal 100 has MCI will be described. Prior to execution of the process of the operation flow 600 shown in FIG. is stored in advance. Further, it is assumed that an application for diagnosing MCI has been activated by the user.

First, in step 602, the question setting unit 512 displays the questions selected in the first phase on the display. For example, the question section 512 displays "Several hiragana are displayed on the screen. Please select a hiragana in the order of your choice..." as illustrated in the screen 800C of FIG. 8C.

Next, in step 604, the answer acquisition unit 514 receives various operations on the user terminal 100. The operations include, for example, a tap operation on a touch panel, a slide operation, and an operation to move the user terminal 100 (for example, a shake operation). For example, after detecting a tap operation on the start button 802 shown in FIG. 8C, a sliding motion of moving the fingertip while pressing it for a long time on the screen is detected. The answer acquisition unit 514 stores the results of the user's answers to the questions (correctness of the answer, response time from the start of the question to the completion of the user's input, etc.) in the storage unit 530.

Next, in step 606, the end determination unit 516 determines whether all the questions included in the MCI diagnostic test have been asked to the user, and if all the questions have been asked, the process proceeds to step 608. If all the questions have not been asked, the process returns to step 602.

Next, in step 608, the level determining unit 518 determines the cognitive level of the user, whether the user is healthy, has MCI, or has dementia, based on the user's answers stored in the memory.

FIG. 7 shows details of step 608 for level determination shown in FIG. Hereinafter, with reference to FIG. 7, a processing flow 700 for determining whether a user has MCI will be described.

In step 702, the level determining unit 518 determines whether the answer to each question is correct or incorrect based on the answer result for each question obtained in step 604 (FIG. 6), and further calculates a score from the correctness or incorrectness of the answer. The score calculation method is as described above.

Next, in step 704, the level determination unit 518 inputs the score calculated in step 702 and the normalized response time to the learned model. Furthermore, the user's age and gender may be input into the trained model.

Next, in step 706, the level determining unit 518 determines the user's cognitive level based on the output from the learned model. If the output value from the trained model is less than the first threshold, it is determined that the user is a healthy person (step 708). Furthermore, if the output from the trained model is greater than or equal to the first threshold and less than the second threshold, the user determines that the MCI is present (step 710). If the output from the trained model is greater than or equal to the second threshold, it is determined that the user has dementia (step 712). The first threshold value is preset so that the rate of correct answers for determining the healthy group and the MCI group is high, and the second threshold value is set in advance so that the rate of correct answers for determining the MCI group and the dementia group is high. .

8A to 8Q illustrate screens 800A to 800Q for a specific screening test. This test includes six questions, and the correctness of the answers and the time required to answer are obtained, and a score is calculated from the correctness of the answers. The content of this test is A. Reproduction of words, B. Character selection, C. Move the device according to instructions, D. Orientation to date and time, E. Memorizing numbers, F. Includes plane shape recognition. It can be determined whether the user is healthy, has MCI, or has dementia from the answers to the questions prepared in this test. First, a test is started after practice questions are asked to the user. The test questions may be asked in the order from A to F above, or in any other order.

Screen 800A (FIG. 8A) and screen 800B (FIG. 8B) are screen displays for testing whether the user in item A above can reproduce the memorized words at a later time (word reproduction). . The question setting section 512 (FIG. 5) displays a screen 800A illustrated in FIG. 8A on the display 106, and the speaker 114 displays a plurality of predetermined words, for example, "telescope," "tangerine," and "wallet." Read one out loud. FIG. 8B illustrates the screen when the second word is being read aloud. In this example, the words read aloud are first memorized, and the answers to the memorized questions are accepted later.

Screen 800C (FIG. 8C) and screen 800D (FIG. 8D) are screen displays for testing whether the order of the letters in item B can be selected correctly. The question section 512 displays on the display 106 the question sentence ``Several hiragana are displayed on the screen. Please press the letters for ``Ai Ue O order'''' as shown in FIG. Instruct the user to trace the following in order (screen 800C). When the user selects the start button 802, the screen changes to a screen 800D shown in FIG. 8D. On the screen 800D, a plurality of consecutive characters starting from characters other than "a" are arranged randomly and randomly. In the illustrated example, hiragana from "ku" to "to" are displayed. The elapsed time 804 from the start of display is displayed on the screen 800D. The answer acquisition unit 514 receives a touch input from the user on the touch panel, and acquires the correctness of the answer and the time from the start of displaying the screen 800D until the user finishes selecting all hiragana as the answer time. The score is calculated by subtracting points from the full score according to the number of mistaps. For example, if you can select everything from "ku" to "to" in order, you will get a full score of 13 points, and if you tap incorrectly once, you will lose 1 point.

Screen 800E (FIG. 8E) and screen 800F (FIG. 8F) are screen displays for testing whether the user of item C can move the terminal according to the instructions. The question setting unit 512 displays the question sentence “Please shake your smartphone as soon as you hear the alarm sound” illustrated at 800F in FIG. 8E on the display 106. When the user selects the start button 802, the screen changes to a screen 800F shown in FIG. 8F, and the alarm sound is output from the speaker 114 multiple times, three times in this example. Ask the user to shake the device each time an alarm sounds. The answer acquisition unit 514 acquires the output of the acceleration sensor and detects the number of times the terminal has been shaken. The response acquisition unit 514 acquires the output from the acceleration sensor after the alarm sound is sounded and the time from when the alarm sound is output until the terminal is shaken as the response time. The score is calculated as 1 point for a correct answer if the terminal is shaken after the alarm tone sounds, and a full score if the terminal is shaken the same number of times as the alarm tone sounds.

Screen 800G (FIG. 8G) and screen 800H (FIG. 8H) are screen displays for accepting responses from the user to the word reproduction test of item A above. The question section 512 presents the question sentence ``We will ask you the word you memorized in the first question. Please select and answer from the options'' (screen 800G), and when the user selects the start button 802, the screen shown in FIG. 8H appears. The screen changes to the screen 800H shown in FIG. On the screen 800H, there are several correct words (in this example, "telescope," "tangerine," and "wallet") and several incorrect words (in this example, "sheep," "light bulb," and "gyoza." ”, “tuna”, and “battery”). The user performs a touch input on the touch panel 106 to select a memorized word. The answer acquisition unit 514 receives touch input from the user on the touch panel 106. The answer acquisition unit 514 acquires the correctness of the answer and the time required to receive input from the user for each word option as the answer time. The score is calculated by subtracting the number of taps from the number of correct words.

Screen 800I (FIG. 8I) and screen 800J (FIG. 8J) are screen displays for testing the user's date and time orientation for item D above. The question asking section 512 presents the question sentence ``I will ask you a question regarding dates starting from ``today.'' Please select from the options and answer.'' as exemplified in the screen 800I of FIG. 8I. For example, if the test implementation date is April 8th, the question setting section 512 displays multiple dates before and after the test implementation date as options on the display. In this example, 10 days before the test implementation date are displayed as options, and the user is asked to select a date, for example, a date 3 days before. The answer acquisition unit 514 receives touch input from the user on the touch panel 106 for date options. Further, when the "Next" button on the screen 800J is selected, the answer acquisition unit 514 acquires the correctness of the answer and the answer time from the start of presentation of date options until the date selection input is accepted. The score is a binary value of correct and incorrect answers, and is calculated by setting a correct answer as 1 point and an incorrect answer as 0 points, for example.

Screen 800K (FIG. 8K), screen 800L (FIG. 8L), and screen 800M (FIG. 8M) are screen displays for testing the user's numerical memory ability for item E. The question section 512 presents the question sentence ``Numbers will be displayed on the screen. Remember the numbers circled and answer in the order in which they are displayed. (Repeat this question until you get it wrong.)'' as exemplified on screen 800K in FIG. 8K. do. Next, the question setting section 512 sequentially displays on the display a plurality of random integers from 0 to 9 that are displayed every few seconds (screen 800L). Next, the question setting unit 512 sequentially receives touch inputs of numbers memorized by the user on the answer screen (FIG. 8M). If the number selected by the user is correct, the question setting unit 512 adds a score according to the memorized number and obtains the answer time for each tap.

Screen 800N (FIG. 8N) and screen 800O (FIG. 8O) are screen displays for testing the user's visual recognition of figures for item F. After presenting the problem statement ``An image with a part missing'' is displayed on the screen.Please choose an appropriate picture to fill in the missing part,'' illustrated on the screen 800N in FIG. The next screen displays an image with a portion missing (screen 800O). In this example, an image is displayed in which rectangles of three colors are randomly arranged next to each other to form one large rectangle, of which a portion is filled in black. Below the large rectangle, four shapes including the missing image are displayed. The answer acquisition unit 514 receives a touch input from the user on one of the four images on the screen 800O. The answer acquisition unit 514 acquires whether or not the correct image was selected (correctness of the answer) and the response time from the start of presenting the image options until the selection input is accepted.

The test ends when all questions have been asked. After the test is completed, the test results of the screening test according to the present disclosure are displayed on the display. Screen 800P (FIG. 8P) and screen 800Q (FIG. 8Q) are screen displays for displaying the user's test results.

The screen 800P in FIG. 8P includes a question that asks you to reproduce the words in item A (memorize the words), a question that asks you to select the letters in order in item B (tracing them in order), and a question that asks whether you can follow the instructions in item C (follow instructions). ), item D, which asks about date and time orientation (answering the date), item E, which tests number memorization ability (remembering numbers), and item F, which asks visual recognition of shapes (selecting images). For each question, the test results and the state of the user's cognitive function are displayed. The state of cognitive function is visually represented by the number of marks 806 (star marks in this example), and in this example, the larger the number of marks 806, the higher the cognitive function. The state with the highest cognitive function is represented by three stars, the state with slightly lower cognitive function is represented by two stars, and the state with lower cognitive function is represented by one star. In item A, which asks you to reproduce words (memorize words), ○ is displayed if you remember all the words, △ if you got some of them wrong, and × if you got all of them wrong. For items B, in which letters are selected in sequence, the time from the start of question display until all letters are selected is displayed. In item C, which asks whether or not the instructions can be followed, the number of times the instructions were correctly followed out of the three instructions given is displayed (two times in the illustrated example). In item D, which asks about date and time orientation, a circle is displayed if the answer is correct, and an x is displayed if the answer is incorrect. For item E, which tests the number memorization ability, the number of digits of the number that could be memorized is displayed. In item F, a question asking about visual recognition of figures (choosing an image), a circle is displayed if the correct image is selected, and an x is displayed if the correct image is not selected.

The screen 800Q in FIG. 8Q visually displays a graph of changes in values regarding the test results of the users who took the test for each question. This allows the user to recognize changes in his or her brain's cognitive function during the screening test period.

Since the screening test of the present disclosure does not require the presence of a doctor, the time burden and mental burden on the user taking the test are less compared to screening tests performed in the presence of a doctor. Additionally, by installing the application on their own devices, users can easily take the test at any time, increasing the possibility of early detection of MCI patients. As a result, it can lead to early treatment of MCI patients.

Although the embodiments of the present invention have been described above, the embodiments of the invention described above are for facilitating understanding of the present invention, and are not intended to limit the present invention. The present invention may be modified and improved without departing from its spirit, and it goes without saying that the present invention includes equivalents thereof. Further, any combination of the embodiments and modifications is possible as long as at least part of the above-mentioned problems can be solved or at least part of the effects can be achieved, and the embodiments and modifications can be combined in any way as long as they are not described in the claims and the specification. Any combination or omission of each component is possible.

10...Diagnosis support system 12...Communication network 100...User terminal 102...Processor 104...Memory 106...Touch panel 106...Display 108...Acceleration sensor 112...Communication interface 114...Speaker 200...Server 202...Processor 204...Memory 206...Communication interface 210 ...Processing section 212...Answer acquisition section 214...Feature amount calculation section 216...Question selection section 218...Model generation section 230...Storage section 510...Processing section 512...Question question section 514...Answer acquisition section 516...Completion judgment section 518...Level judgment Section 530...Storage section

Claims (7)

An information processing method executed by a processor that executes instructions stored in a storage unit, the method comprising:
A step of displaying the text of a test question determined to have a statistically significant difference in order to determine the user's cognitive level, the cognitive level being between a healthy person level and a mild cognitive impairment (MCI) level. , a step including a dementia level;
a step of accepting the user's answer to the question statement;
inputting feature quantities calculated based on the user's answer results into a trained model to determine whether the user is MCI;
The determined test questions are divided into multiple items, and the multiple questions included in each item are questions in which one question is asked in a different way. From the answers of the test subjects, the scores of the healthy group, the MCI group, and the dementia group are obtained, and questions that have a correlation between the obtained scores and the cognitive level are determined based on the statistically significant difference. Steps to determine that there is a problem
If the score of the MCI group is between the score of the healthy group and the score of the dementia group, it is determined that the problem has a correlation with the cognitive level.
Information processing method. The information processing method according to claim 1,
Each item consists of selecting a memorized word after a certain time, selecting a letter, moving the user's terminal when a sound sounds , answering the date and time, answering a memorized number, and recognizing a plane figure. An information processing method, wherein one of the plurality of problems included in the item is the problem determined to be statistically significant. The information processing method according to claim 2, further comprising:
From the results of the test subjects' answers to the question sentences of the plurality of questions, the response times of the healthy group, the MCI group, and the dementia group were obtained, and the relationship between the obtained response times and the cognitive level was calculated. The step of determining a question with a high correlation as a question with the statistically significant difference, wherein the response time of the MCI group is between the response time of the healthy group and the response time of the dementia group. An information processing method in which it is determined that the question has a high correlation with the cognitive level. 4. The information processing method according to claim 1, wherein the feature amount is a score calculated based on whether the answer is correct or incorrect obtained from the user's answer to the problem statement; and a response time until receiving a response from the user to the question statement. An information processing device comprising a storage unit that stores instructions and a processor, the information processing device comprising:
The processor executes instructions stored in the storage unit, thereby
In order to determine the user's cognitive level, the text of the test question determined to have a statistically significant difference is displayed, and the cognitive level includes a healthy level, an MCI level, and a dementia level. ,
Accepting the user's answer to the problem statement,
Inputting feature amounts calculated based on the user's answer results into a trained model to determine whether the user is MCI,
The determined test questions are divided into multiple items, and the multiple questions included in each item are questions in which one question is asked in a different way. From the answers of the test subjects, the scores of the healthy group, the MCI group, and the dementia group are obtained, and questions that have a correlation between the obtained scores and the cognitive level are determined based on the statistically significant difference. If the score of the MCI group is between the score of the healthy group and the score of the dementia group, it is determined that the problem has a correlation with the cognitive level. ,
Information processing device. A program executed by a processor,
When the processor executes the program,
In order to determine the user's cognitive level, the text of the test question determined to have a statistically significant difference is displayed, and the cognitive level includes a healthy level, an MCI level, and a dementia level. ,
Accepting the user's answer to the problem statement,
Inputting feature amounts calculated based on the user's answer results into a trained model to determine whether the user is MCI,
The determined test questions are divided into multiple items, and the multiple questions included in each item are questions in which one question is asked in a different way. From the answers of the test subjects, the scores of the healthy group, the MCI group, and the dementia group are obtained, and questions that have a correlation between the obtained scores and the cognitive level are determined based on the statistically significant difference. If the score of the MCI group is between the score of the healthy group and the score of the dementia group, it is determined that the problem has a correlation with the cognitive level. ,
program. A recording medium storing a program executed by a processor,
By the processor executing the program recorded on the recording medium ,
In order to determine the user's cognitive level, the text of the test question determined to have a statistically significant difference is displayed, and the cognitive level includes a healthy level, an MCI level, and a dementia level. ,
Accepting the user's answer to the problem statement,
Inputting feature amounts calculated based on the user's answer results into a trained model to determine whether the user is MCI,
The determined test questions are divided into multiple items, and the multiple questions included in each item are questions in which one question is asked in a different way. From the answers of the test subjects, the scores of the healthy group, the MCI group, and the dementia group are obtained, and questions that have a correlation between the obtained scores and the cognitive level are determined based on the statistically significant difference. Determining that the problem is a certain problem, and determining that the problem has a correlation with the cognitive level if the score of the MCI group is between the score of the healthy group and the score of the dementia group.
recoding media.

Images