WO2022102721A1

WO2022102721A1 - Depression-state-determining program

Info

Publication number: WO2022102721A1
Application number: PCT/JP2021/041602
Authority: WO
Inventors: 綾子澤田
Original assignee: Ａｓｓｅｓｔ株式会社
Priority date: 2020-11-11
Filing date: 2021-11-11
Publication date: 2022-05-19

Abstract

[Problem] To determine, to a high degree of accuracy, the depression state of a performer using a relatively simple method. [Solution] A depression-state-determining program for determining the depression state of a performer appearing in the media, the depression-state-determining program being characterized in causing a computer to execute: an information acquisition step for acquiring video image information comprising a video image of the performer; and a depression-state-determining step for using the degree of association, of three or more levels, between reference video image information in which an action of a person is imaged in advance as a video image and the level of a depression state, and determining the depression state of the performer on the basis of the degree of association, of three or more levels, between reference video image information corresponding to the video image information acquired in the information acquisition step and the level of a depression state.

Description

Depression state determination program

The present invention relates to a depression state determination program capable of determining the depression state of a performer such as a media suffering from stress or human relations and determining in advance that the performer falls into the depression state.

In modern society, mental stress is a problem. In particular, it has been reported that performers such as talents who appear in the media may fall into a depressed state due to stress or the like, and in particular, they may even commit suicide. In order to prevent such a performer from becoming depressed, it is desirable to monitor the performer's depression and confirm in advance whether or not there is a risk of becoming depressed. However, in the past, the current situation has not been proposed as a system capable of discriminating the depressed state of a performer such as a talent with high accuracy by a relatively simple method.

In recent years, the number of patients with depression is increasing, not limited to personalities. When depression occurs, it may appear as a sign to the outside even if the person is not aware of it. If the signs of possible depression can be detected in advance, various prescriptions can be given at that stage to avoid the risk of full-scale depression. However, in the past, no system has been proposed for automatically and highly accurately discriminating such signs of depression.

Therefore, the present invention has been devised in view of the above-mentioned problems, and an object thereof is to determine the depressed state of a performer appearing in the media with high accuracy by a relatively simple method. The purpose is to provide a possible depression state determination program. Another object of the present invention is to provide a depression sign discrimination program capable of automatically and accurately discriminating the signs of depression of a subject.

The depression state determination program according to the present invention is a depression state determination program for determining the depression state of a performer appearing in the media, and includes an information acquisition step for acquiring moving image information consisting of the moving image of the performer and a human operation. Using the degree of association between the reference moving image information previously captured as a moving image and the level of depression at three or more levels, the reference moving image information and depression according to the moving image information acquired in the above information acquisition step. It is characterized in that the computer executes the depression state determination step for determining the depression state of the performer based on the degree of association with the state level in three or more stages.

Even if you do not have any special skills or experience, it is possible to determine the depressed state of a performer with high accuracy using a relatively simple method.

It is a block diagram which shows the whole structure of the system to which this invention is applied. It is a figure which shows the specific configuration example of a search device. It is a figure for demonstrating the operation of this invention. It is a figure for demonstrating the operation of this invention. It is a figure for demonstrating the operation of this invention. It is a figure for demonstrating the operation of this invention. It is a figure for demonstrating the operation of this invention. It is a figure for demonstrating the operation of this invention. It is a figure for demonstrating the operation of this invention. It is a figure for demonstrating the operation of this invention. It is a figure for demonstrating the operation of this invention. It is a figure for demonstrating the operation of this invention. It is a figure for demonstrating the operation of this invention. It is a figure for demonstrating the operation of this invention. It is a figure for demonstrating the operation of this invention. It is a figure for demonstrating the operation of this invention.

Hereinafter, the depression state determination program to which the present invention is applied will be described in detail with reference to the drawings.

The first embodiment FIG. 1 is a block diagram showing an overall configuration of a depression state determination system 1 to which a depression state determination program to which the present invention is applied is implemented. The depression state determination system 1 includes an information acquisition unit 9, a determination device 2 connected to the information acquisition unit 9, and a database 3 connected to the determination device 2.

The information acquisition unit 9 is a device for a person using this system to input various commands and information, and specifically, is composed of a keyboard, buttons, a touch panel, a mouse, a switch, and the like. The information acquisition unit 9 is not limited to a device for inputting text information, and may be configured by a device such as a microphone that can detect voice and convert it into text information. Further, the information acquisition unit 9 may be configured as an image pickup device capable of taking an image of a camera or the like. The information acquisition unit 9 may be configured by a scanner having a function of recognizing a character string from a paper-based document. Further, the information acquisition unit 9 may be integrated with the estimation device 2 described later. The information acquisition unit 9 outputs the detected information to the estimation device 2. Further, the information acquisition unit 9 may be configured by means for specifying the position information by scanning the map information. Further, the information acquisition unit 9 may be composed of an illuminance sensor for measuring a temperature sensor, a humidity sensor, and a wind direction sensor. Further, the information acquisition unit 9 may be configured by a communication interface for acquiring data about the weather from the Japan Meteorological Agency or a private weather forecast company. Further, the information acquisition unit 9 may be composed of a body sensor that is attached to the body to detect body data, and the body sensor detects, for example, body temperature, heart rate, blood pressure, number of steps, walking speed, and acceleration. It may be composed of a sensor for the purpose. Further, the information acquisition unit 9 may be configured as a device for acquiring information such as drawings by scanning or reading from a database. In addition to these, the information acquisition unit 9 may be configured by an odor sensor that detects odors and scents.

Database 3 stores various information necessary for determining the depression state. Information necessary for determining the depression state includes reference moving image information in which human movements are captured in advance as moving images, reference audio information in which human voice is recorded, and reference text in which text data is categorized in advance. Information, reference schedule information that categorizes schedule data in advance, reference appearance frequency information that categorizes appearance frequency data in advance, a depressed state in which a judgment is actually made for these, and to improve the depressed state. A data set with improvement measures is stored.

That is, in addition to such reference moving image information, the database 3 contains one or more of reference audio information, reference text information, reference schedule information, and reference appearance frequency information, and is in a depressed state or improved. The measures are linked to each other and remembered. Further, in the database 3, in addition to the reference voice information, any one or more of the reference text information, the reference schedule information, and the reference appearance frequency information, and the depressed state or improvement measures are stored in association with each other. There is.

The discrimination device 2 is composed of, for example, an electronic device such as a personal computer (PC), but is embodied in any other electronic device such as a mobile phone, a smartphone, a tablet terminal, a wearable terminal, etc., in addition to the PC. It may be the one to be converted. The user can obtain a search solution by the discrimination device 2.

FIG. 2 shows a specific configuration example of the discrimination device 2. The discrimination device 2 performs wired communication or wireless communication with a control unit 24 for controlling the entire discrimination device 2 and an operation unit 25 for inputting various control commands via an operation button, a keyboard, or the like. A communication unit 26 for the purpose, an estimation unit 27 for making various judgments, and a storage unit 28 for storing a program for performing a search to be executed represented by a hard disk or the like are connected to the internal bus 21, respectively. .. Further, a display unit 23 as a monitor that actually displays information is connected to the internal bus 21.

The control unit 24 is a so-called central control unit for controlling each component mounted in the discrimination device 2 by transmitting a control signal via the internal bus 21. Further, the control unit 24 transmits various control commands via the internal bus 21 according to the operation via the operation unit 25.

The operation unit 25 is embodied by a keyboard or a touch panel, and an execution command for executing a program is input from the user. When the execution command is input by the user, the operation unit 25 notifies the control unit 24 of the execution command. Upon receiving this notification, the control unit 24, including the estimation unit 27, executes a desired processing operation in cooperation with each component. The operation unit 25 may be embodied as the information acquisition unit 9 described above.

The estimation unit 27 estimates the search solution. The estimation unit 27 reads out various information stored in the storage unit 28 and various information stored in the database 3 as necessary information when executing the estimation operation. The estimation unit 27 may be controlled by artificial intelligence. This artificial intelligence may be based on any well-known artificial intelligence technique.

The display unit 23 is configured by a graphic controller that creates a display image based on the control by the control unit 24. The display unit 23 is realized by, for example, a liquid crystal display (LCD) or the like.

When the storage unit 28 is composed of a hard disk, predetermined information is written to each address based on the control by the control unit 24, and is read out as needed. Further, the storage unit 28 stores a program for executing the present invention. This program will be read and executed by the control unit 24.

The operation in the depression state determination system 1 having the above-mentioned configuration will be described.

In the depression state determination system 1, for example, as shown in FIG. 3, it is premised that the degree of association between the reference moving image and the depression state is set and acquired in advance. The reference moving image is information obtained by capturing all movements and facial expressions of a person as a moving image. As the reference moving image information, the facial expression, movement, etc. of a person may be patterned by analyzing the moving image and analyzing the feature amount or the like by a deep learning technique or the like. Behaviors that show signs of depression include, for example, poor involvement with others, agitated attitude, looking at someone else's complexion, abnormal dietary attachment, and terrible restlessness. Behaviors are categorized and patterned in advance.

If the pattern is that the neck sways back and forth or the limbs are poorly shaken as a sign that the attitude is frightening, it is determined through the analysis of the moving image whether or not it applies to the pattern. In this discrimination process, existing image analysis techniques and deep learning techniques may be used. Then, when an action applied to this pattern is detected, it is applied to the reference moving image information classified according to each action pattern.

A trained model is created by similarly forming a degree of association between such reference moving image information and the level of depression. Then, when actually determining the depressed state of the performer, by performing image analysis in the same manner from the moving image of the performer, the image is applied to various types of pre-patterned behaviors, and this is applied to the moving image. It is information. The type of behavior pattern to be applied in this moving image information shall be in accordance with the above-mentioned reference moving image information.

Based on the behavior pattern information extracted from the image information in this way, it is possible to determine the actual possibility of abuse by referring to the degree of association formed between the reference behavior pattern information and the possibility of abuse. .. The process of determining the possibility of abuse is the same as the process of determining the possibility of abuse based on the image information shown in FIG. 3 described above.

The level of depression is an individual depression that is objectively evaluated. As an example of this depressed state, for example, it may be based on evaluation data, medical care data, medical examination results, etc., in which the depressed state is objectively evaluated by a doctor or a specialist having specialized knowledge. In addition, the evaluator of this depression is not limited to those who have specialized knowledge about mentality, and includes those who do not have that specialized knowledge. In other words, the evaluator of the depressed state may be evaluated by a third party other than the person himself / herself.

Examples of evaluation of depression include, for example, severe depression, mild depression, normal, slightly depressed, and distracted attention, but the present invention is not limited to these, and for example, depression is 0 points. When the normal score is 100 points, the score may be evaluated numerically in 100 steps between 0 and 100 points.

In the example of FIG. 3, it is assumed that the input data is, for example, reference moving images P01 to P03. The reference moving images P01 to P03 as such input data are connected to the output. In this output, the depressed state as the output solution is displayed.

The reference moving images are associated with each other through three or more levels of association with the level of depression as the output solution. The reference moving images are arranged on the left side through this degree of association, and each depression state is arranged on the right side through this degree of association. The degree of association indicates which depression state is more relevant to the reference moving images arranged on the left side. In other words, this degree of association is an indicator of what kind of depression each reference video is likely to be associated with, and is used to select the most probable depression from the reference video. It shows the accuracy. In the example of FIG. 3, w13 to w19 are shown as the degree of association. These w13 to w19 are shown in 10 stages as shown in Table 1 below, and the closer to 10 points, the higher the degree of relevance of each combination as an intermediate node to the depression state as an output. On the contrary, the closer to one point, the lower the degree of relevance of each combination as an intermediate node to the price as an output.

The discrimination device 2 acquires in advance the degree of association w13 to w19 of three or more stages shown in FIG. That is, the discriminating device 2 accumulates past data sets and analyzes which of the reference moving image and the depressed state in that case is adopted and evaluated in discriminating the actual search solution. , The degree of association shown in FIG. 3 is created by analysis.

For example, it is assumed that the reference moving image has a frightened facial expression, which is a sign of depression, detected four times in a row. As the level of depression for such a reference moving image, it is assumed that strong depression is often evaluated. By collecting and analyzing such a data set, the degree of association between the reference moving image (“frightened facial expression” four times in a row) and the depressed state (strong depressed state) becomes stronger.

This analysis may be performed by artificial intelligence. In such a case, for example, in the case of the reference moving image P01, analysis is performed from various data as a result of evaluating the past depression state. For example, the data set may be extracted by performing a text mining analysis from the electronic data of the diagnosis result or the evaluation result in the workplace. In the case of the reference moving image P01, if there are many cases of "strong depression", the degree of association that leads to the evaluation of this "strong depression" is set higher, and the case of "slightly depressed". If there are many, set a higher degree of association that leads to the evaluation of this depression. For example, in the example of the problem information P01 for reference, it is linked to "strong depression state" and "slightly depressed", but from the previous case, the degree of association of w13 that leads to "strong depression state" is 7 points. In addition, the degree of association of w14, which leads to "slightly depressed", is set to 2 points.

Further, the degree of association shown in FIG. 3 may be composed of the nodes of the neural network in artificial intelligence. That is, the weighting coefficient for the output of the node of this neural network corresponds to the above-mentioned degree of association. Further, the network is not limited to a neural network, and may be composed of all decision-making factors constituting artificial intelligence.

Such degree of association is what is called learned data in artificial intelligence. After creating such learned data, the above-mentioned learned data will be used to search for the depressed state in actually determining the depressed state for the performers. In such a case, the moving image is subjectively evaluated by the performer to be discriminated, or the moving image is newly acquired by asking through an interview or the like.

The newly acquired moving image is input by the above-mentioned information acquisition unit 9. When the performer is actually appearing on a TV program, the information acquisition unit 9 may obtain the information by recording the TV program, or the moving image posted on various video posting sites may be obtained. In some cases, it may be obtained from there.

Based on the newly acquired moving image in this way, the depressed state of the performer is determined. In such a case, the degree of association shown in FIG. 3 (Table 1) acquired in advance is referred to. For example, when the newly acquired moving image is the same as or similar to P02, "normal" is associated with w15 and "slightly depressed" is associated with the association degree w16 through the association degree. In such a case, "normal" with the highest degree of association is selected as the optimum solution. However, it is not essential to select the one with the highest degree of association as the optimum solution, and it is also possible to select "slightly depressed" as the optimum solution, although the degree of association is low but the association itself is recognized. In addition to this, it goes without saying that an output solution to which the arrows are not connected may be selected, and any other output solution may be selected in any other priority as long as it is based on the degree of association.

In this way, the most suitable depression state can be searched for and displayed to the user from the newly acquired moving image. By looking at this search result, the user can obtain an approach guideline for various improvement measures for caring for the performer, that is, for restoring mental health, based on the searched depression state. By the way, this depressive state is not limited to the case where it is merely a state evaluation, and it also mentions how to care for the performers, that is, to propose improvement measures for working styles for the depressive state. It may be the one that is.

In the above-described embodiment, the reference moving image information and the moving image information have been described by taking the case of a moving image in which a plurality of still images are continuously connected in chronological order as an example, but the description is limited to this. It's not something. Of course, it is also possible to extract features and various patterns from the reference moving image information and the individual still images constituting the moving image information, and to learn by associating these with the level of the depressed body.

In the example of FIG. 4, it is premised that a combination of the reference moving image information and the reference audio information is formed. The reference voice information includes all voice data recorded by the voice emitted by the performer.

This reference voice information may be composed of text data when the voice spoken by the subject who determines the sign of depression is converted into text data, and the text data is subjected to morphological analysis or syntactic analysis. It may be composed of data.

For example, as a result of performing morphological analysis, this reference voice information may be composed of information indicating how much pronouns are included in the text data. For example, the subject says, "I will go to Osaka with Fujimoto tomorrow by 13:00 on the Shinkansen," and "I will go to Osaka tomorrow. The meaning of the former is clearer, while the meaning of the latter is unclear. When a patient with depression has a high proportion of pronouns in the text data of his / her voice, he / she extracts this in units of text data to obtain voice information for reference.

Further, the reference voice information may be configured by measuring the interval time until the subject who receives the question from another person answers.

In addition to this, the reference voice information is information related to voice tones extracted from past subjects. The tone of this voice refers to, for example, the pitch of the sound (the number of vibrations per second of the sound wave, that is, the frequency), the sound itself, or the strength of the sound. The tone of the voice may be detected and analyzed by detecting the height and the strength of the voice through a general voice detector.

The reference voice information may be linked with the extracted text data. For example, in the phrase "I will go to Osaka with Mr. Fujimoto tomorrow by 13:00 on the Shinkansen", for each noun phrase (case component) such as "Tomorrow" and "with Mr. Fujimoto". , Each voice tone may be associated and used as reference voice information.

As input data, such reference audio information and reference video information are lined up. In the example of FIG. 4, it is assumed that the input data is, for example, reference moving images P01 to P03 and reference audio information P14 to 17. The intermediate node shown in FIG. 4 is a combination of reference audio information and reference video as such input data. Each intermediate node is further linked to the output. In this output, the depressed state as the output solution is displayed.

Each combination (intermediate node) of the reference moving image and the reference audio information is associated with each other through three or more levels of association with the depressed state as this output solution. The reference moving image and the reference audio information are arranged on the left side through this degree of association, and the depressed state is arranged on the right side through this degree of association. The degree of association indicates the degree of relevance to the depressed state with respect to the reference moving image and the reference audio information arranged on the left side. In other words, this degree of association is an index showing what kind of depression each reference moving image and reference audio information are likely to be associated with, and is the most from the reference moving image and reference audio information. It shows the accuracy in selecting a probable depression. Depression to be evaluated will differ depending on the actual state of attendance and departure, in addition to the degree of mental health that the performer subjectively inputs. Therefore, the optimum depression state is searched for by combining these reference moving images and reference audio information.

In the example of FIG. 4, w13 to w22 are shown as the degree of association. As shown in Table 1, these w13 to w22 are shown in 10 stages, and the closer to 10 points, the higher the degree of relevance of each combination as an intermediate node to the output, and conversely, 1 point. The closer they are, the less relevant each combination as an intermediate node is to the output.

The discrimination device 2 acquires in advance the degree of association w13 to w22 of three or more stages shown in FIG. That is, the discriminating device 2 accumulates past data as to which of the reference moving image, the reference audio information, and the depressed state in that case was suitable for discriminating the actual search solution. By analyzing and analyzing these, the degree of association shown in FIG. 4 is created.

For example, it is assumed that the reference moving image in the actual case in the past has 5 consecutive signs of frightening and normal signs every other day. Further, it is assumed that the reference voice information is a sign of tone down for three consecutive days. In such a case, if there are many "slightly depressed" depressed states, these are learned as a data set and defined in the form of the above-mentioned degree of association.

This analysis may be performed by artificial intelligence. In such a case, for example, in the case of the reference moving image P01 and the reference audio information P16, the depressive state is analyzed from the past data. When there are many cases of "normal" depression, the degree of association leading to this "normal" is set higher, and when there are many cases of "distracted attention" and few cases of "normal", the degree of association is set higher. Set the degree of association that leads to "distracted attention" high and the degree of association that leads to "normal" low. For example, in the example of the intermediate node 61a, it is linked to the output of "depressed state" and "normal", but from the previous case, the degree of association of w13 that leads to "depressed state" is set to 7 points, and w14 that leads to "normal". The degree of association is set to 2 points.

Further, the degree of association shown in FIG. 4 may be composed of the nodes of the neural network in artificial intelligence. That is, the weighting coefficient for the output of the node of this neural network corresponds to the above-mentioned degree of association. Further, the network is not limited to a neural network, and may be composed of all decision-making factors constituting artificial intelligence.

In the example of the degree of association shown in FIG. 4, the node 61b is a node of the combination of the reference audio information P14 with respect to the reference moving image P01, the degree of association of "slightly depressed" is w15, and "immediate hospitalization". The degree of association of "a level of severe depression that requires" is w16. The node 61c is a node of a combination of reference audio information P15 and P17 with respect to the reference moving image P02, and the degree of association of "normal" is w17 and the degree of association of "distracted attention" is w18. There is.

Such degree of association is what is called learned data in artificial intelligence. After creating such learned data, when actually determining the depressed state of the performer from now on, the above-mentioned learned data will be used. In such a case, the moving image and the audio information are actually acquired from the performer to be discriminated. The audio information corresponds to the reference audio information, and the performer records the audio generated in the moving image uploaded to the TV program, the radio program, or the video posting site.

Search for the optimal depression state based on the newly acquired moving image and audio information in this way. In such a case, the degree of association shown in FIG. 4 (Table 1) acquired in advance is referred to. For example, when the newly acquired moving image is the same as or similar to P02 and the audio information is P17, the node 61d is associated via the degree of association, and this node. In 61d, "slightly depressed" is associated with w19, and "attention distracted state" is associated with a degree of association w20. In such a case, the “slightly depressed” with the highest degree of association is selected as the optimum solution. However, it is not essential to select the one with the highest degree of association as the optimum solution, and the "distracted attention state" in which the degree of association itself is recognized although the degree of association is low may be selected as the optimum solution. In addition to this, it goes without saying that an output solution to which the arrows are not connected may be selected, and any other output solution may be selected in any other priority as long as it is based on the degree of association.

Table 2 below shows an example of the degree of association w1 to w12 extending from the input.

The intermediate node 61 may be selected based on the degree of association w1 to w12 extending from this input. That is, the larger the degree of association w1 to w12, the heavier the weighting in the selection of the intermediate node 61 may be. However, the degrees of association w1 to w12 may all have the same value, and the weights in the selection of the intermediate node 61 may all be the same.

FIG. 5 shows an example in which a combination of the above-mentioned reference moving image, reference text information, and a depression state with respect to the combination are set to three or more levels of association.

The reference text information is categorized in advance for text data. For example, words that correspond to false accusation or slander are categorized in advance. Then, it may indicate whether or not it corresponds to the categorized wording. Further, the degree of accusation may be indicated in a level, and the wording according to the level may be categorized in advance.

In the example of FIG. 5, it is assumed that the input data is, for example, reference moving images P01 to P03 and reference text information P18 to 21. The intermediate node shown in FIG. 5 is a combination of reference text information and reference video as such input data. Each intermediate node is further linked to the output. In this output, the depressed state as the output solution is displayed.

Each combination (intermediate node) of the reference moving image and the reference text information is associated with each other through three or more levels of association with the depressed state as this output solution. The reference moving image and the reference text information are arranged on the left side through this degree of association, and the depressed state is arranged on the right side through this degree of association. The degree of association indicates the degree of relevance to the depressed state with respect to the reference moving image and the reference text information arranged on the left side. In other words, this degree of association is an index showing what kind of depression each reference moving image and reference text information is likely to be associated with, and is the most from the reference moving image and reference text information. It shows the accuracy in selecting a probable depression. Depression to be evaluated will differ depending on what the actual income situation is, in addition to the degree of mental health that the performer subjectively inputs. Therefore, the optimum depression state is searched for by combining these reference moving images and reference text information.

In the example of FIG. 5, w13 to w22 are shown as the degree of association. As shown in Table 1, these w13 to w22 are shown in 10 stages, and the closer to 10 points, the higher the degree of relation between each combination as an intermediate node with the depression state as an output, and vice versa. The closer it is to one point, the lower the degree of association between each combination as an intermediate node and the depression state as an output.

The discrimination device 2 acquires in advance the degree of association w13 to w22 of three or more stages shown in FIG. That is, the discriminating device 2 accumulates past data as to which of the reference moving image, the reference text information, and the depression state in that case was suitable for discriminating the actual search solution. By analyzing and analyzing these, the degree of association shown in FIG. 5 is created.

This analysis may be performed by artificial intelligence. In such a case, for example, in the case of the reference moving image P01 and the reference text information P20, the depressive state is analyzed from the past data. When there are many cases of depression being "slightly depressed", the degree of association that leads to this "slightly depressed" is set higher, and there are many cases of "distracted attention", and cases of "slightly depressed". If the number is low, the degree of association that leads to "distracted attention" is set high, and the degree of association that leads to "slightly depressed" is set low. For example, in the example of the intermediate node 61a, it is linked to the output of "mild depression" and "normal", but from the previous case, the degree of association of w13 that leads to "mild depression" is set to 7 points, and "normal". The degree of association of w14 that leads to "" is set to 2 points.

Further, the degree of association shown in FIG. 5 may be composed of the nodes of the neural network in artificial intelligence. That is, the weighting coefficient for the output of the node of this neural network corresponds to the above-mentioned degree of association. Further, the network is not limited to a neural network, and may be composed of all decision-making factors constituting artificial intelligence.

In the example of the degree of association shown in FIG. 5, the node 61b is a node of the combination of the reference text information P18 with respect to the reference moving image P01, the degree of association of "slightly depressed" is w15, and "immediate hospitalization". The degree of association of "a level of severe depression that requires" is w16. The node 61c is a node in which the reference text information P19 and P21 are combined with respect to the reference moving image P02, and the degree of association of "normal" is w17 and the degree of association of "distracted attention" is w18. There is.

Such degree of association is what is called learned data in artificial intelligence. After creating such trained data, when actually searching for a depressed state from now on, the above-mentioned trained data will be used. In such a case, the moving image of the performer whose depression state is actually determined and the text information are acquired. The text information is composed of text data acquired from writing to various information sites and SNS on the Internet. The acquired text data is analyzed based on well-known morphological analysis techniques and syntactic analysis techniques, and which separation of the pre-categorized reference text information includes the words and words contained therein. Determine.

Search for the optimal depression state based on the newly acquired moving image and text information in this way. In such a case, the degree of association shown in FIG. 5 (Table 1) acquired in advance is referred to. For example, when the newly acquired moving image is the same as or similar to P02 and the text information is P21, the node 61d is associated via the degree of association, and this node. In 61d, "slightly depressed" is associated with w19, and "attention distracted state" is associated with a degree of association w20. In such a case, the “slightly depressed” with the highest degree of association is selected as the optimum solution. However, it is not essential to select the one with the highest degree of association as the optimum solution, and the "distracted attention state" in which the degree of association itself is recognized although the degree of association is low may be selected as the optimum solution. In addition to this, it goes without saying that an output solution to which the arrows are not connected may be selected, and any other output solution may be selected in any other priority as long as it is based on the degree of association.

FIG. 6 shows an example in which the combination of the above-mentioned reference moving image, the reference schedule information, and the depression state with respect to the combination are set to three or more levels of association.

The reference schedule information includes not only the program appearance schedule of the performers but also information on all work schedules. In other words, the reference schedule information shows how the autograph session, lecture, other interview time, working time, travel time, practice time, training time, etc. are organized in chronological order. Since this reference schedule information is only reference information, it may be categorized by replacing it with the schedule of not only people who are engaged in talent activities such as performers but also ordinary people. In such a case, for example, in a one-week schedule, various schedules such as working hours, break times, traveling times, meeting hours, and free hours are acquired as learning data.

In the example of FIG. 6, it is assumed that the input data is, for example, reference moving images P01 to P03 and reference schedule information P22 to 25. The intermediate node shown in FIG. 6 is a combination of the reference moving image and the reference schedule information as such input data. Each intermediate node is further linked to the output. In this output, the depressed state as the output solution is displayed.

Each combination (intermediate node) of the reference moving image and the reference schedule information is associated with each other through three or more levels of association with the depressed state as this output solution. The reference moving image and the reference schedule information are arranged on the left side through this degree of association, and the depressed state is arranged on the right side through this degree of association. The degree of association indicates the degree of relevance to the depressed state with respect to the reference moving image and the reference schedule information arranged on the left side. In other words, this degree of association is an index showing what kind of depression each reference moving image and reference schedule information are likely to be associated with, and is the most from the reference moving image and reference schedule information. It shows the accuracy in selecting a probable depression.

In the example of FIG. 6, w13 to w22 are shown as the degree of association. As shown in Table 1, these w13 to w22 are shown in 10 stages, and the closer to 10 points, the higher the degree of relation between each combination as an intermediate node with the depression state as an output, and vice versa. The closer it is to one point, the lower the degree of association between each combination as an intermediate node and the depression state as an output.

The discrimination device 2 acquires in advance the degree of association w13 to w22 of three or more stages shown in FIG. That is, the discriminating device 2 accumulates past data as to which of the reference moving image, the reference schedule information, and the depression state in that case was suitable for discriminating the actual search solution. By analyzing and analyzing these, the degree of association shown in FIG. 6 is created.

This analysis may be performed by artificial intelligence. In such a case, for example, in the case of the reference moving image P01 and the reference schedule information P24, the depression state is analyzed from the past data. When there are many cases of depression being "slightly depressed", the degree of association that leads to this "slightly depressed" is set higher, and there are many cases of "distracted attention", and cases of "slightly depressed". If the number is low, the degree of association that leads to "distracted attention" is set high, and the degree of association that leads to "slightly depressed" is set low. For example, in the example of the intermediate node 61a, it is linked to the output of "mild depression" and "normal", but from the previous case, the degree of association of w13 that leads to "depression" is set to 7 points and becomes "normal". The degree of association of the connected w14 is set to 2 points.

Further, the degree of association shown in FIG. 6 may be composed of the nodes of the neural network in artificial intelligence. That is, the weighting coefficient for the output of the node of this neural network corresponds to the above-mentioned degree of association. Further, the network is not limited to a neural network, and may be composed of all decision-making factors constituting artificial intelligence.

In the example of the degree of association shown in FIG. 6, the node 61b is a node in which the reference schedule information P22 is combined with the reference moving image P01, the degree of association of "slightly depressed" is w15, and "immediate hospitalization". The degree of association of "a level of severe depression that requires" is w16. The node 61c is a node in which the reference schedule information P23 and P25 are combined with respect to the reference moving image P02, and the "normal" association degree is w17 and the "attention distracted state" association degree is w18. There is.

Such degree of association is what is called learned data in artificial intelligence. After creating such trained data, when actually searching for a depressed state from now on, the above-mentioned trained data will be used. In such a case, the moving image of the performer whose depression state is actually determined and the schedule information are acquired. The schedule information may be manually input or directly acquired from the attribute data of the performers managed by the company.

Search for the optimal depression state based on the newly acquired moving image and schedule information in this way. In such a case, the degree of association shown in FIG. 6 (Table 1) acquired in advance is referred to. For example, when the newly acquired moving image is the same as or similar to P02 and the schedule information is P25, the node 61d is associated via the degree of association, and this node. In 61d, "slightly depressed" is associated with w19, and "attention distracted state" is associated with a degree of association w20. In such a case, the “slightly depressed” with the highest degree of association is selected as the optimum solution. However, it is not essential to select the one with the highest degree of association as the optimum solution, and the "distracted attention state" in which the degree of association itself is recognized although the degree of association is low may be selected as the optimum solution. In addition to this, it goes without saying that an output solution to which the arrows are not connected may be selected, and any other output solution may be selected in any other priority as long as it is based on the degree of association.

The reference schedule information shown in FIG. 6 may be replaced with reference appearance frequency information. The reference appearance frequency information referred to here is information on appearance frequency in TV programs and radio programs, and for example, appearance time, appearance frequency, etc. per unit time (for example, one week, one month, one year, etc.). It can be shown by frequency data. The appearance frequency information for reference may be categorized as appearance frequency data, for example, less than once a week, less than once or twice a week, twice a week to less than three times a week, and the like. May be. In particular, talents can tell whether or not they have a lot of work depending on the frequency of appearances, and depression may be controlled accordingly. Therefore, this reference appearance frequency information is also added to the explanatory variables.

In such a case, create in advance three or more levels of association between the reference moving image information, the combination having the reference appearance frequency information categorized in advance, and the level of depression. Then, the appearance frequency information regarding the appearance frequency of the performers in various media is newly provided. The type of the appearance frequency information as data is the same as the above-mentioned reference appearance frequency information. Then, based on the combination of the reference moving image information according to the acquired moving image information and the reference appearance frequency information according to the appearance frequency information, and the degree of association between the depression level and the above-mentioned appearance. Determine the depressed state of a person.

The present invention is not limited to the above-described embodiment, and as shown in FIG. 7, for example, instead of the reference moving image information, there are three or more stages of reference audio information and a level of depression. You may try to use the degree of association of. In such a case, the depressed state of the performer is determined based on the degree of association between the reference voice information according to the newly acquired voice information and the level of the depressed state at three or more levels.

Although not shown, in addition to the reference moving image information, not only one of the reference audio information, the reference text information, the reference schedule information, and the reference appearance frequency information, but also any two or more, and depression. The state or improvement measures may be stored in association with each other. In addition to the reference voice information, not only one of the reference text information, the reference schedule information, and the reference appearance frequency information, but also any two or more, the depressed state, or the improvement measures are linked to each other. It may be remembered. In addition to the reference moving image information, in addition to any one or more of the reference audio information, the reference text information, the reference schedule information, and the reference appearance frequency information, the depression state or improvement can be achieved by combining with other data. The measures may be linked and memorized. In addition to the reference voice information, in addition to any one or more of the reference text information, the reference schedule information, and the reference appearance frequency information, in combination with other data, the depressed state or improvement measures are linked to each other. It may be remembered.

Similarly, in these cases, when the information corresponding to the reference information used as the learning data is input, the solution search is performed based on the above-mentioned method.

In the above-mentioned degree of association, the degree of association is expressed by a 10-step evaluation, but it is not limited to this, and it may be expressed by a degree of association of 3 or more levels, and conversely, it may be expressed by 3 or more levels. For example, 100 steps or 1000 steps may be used. On the other hand, this degree of association does not include those expressed in two stages, that is, whether or not they are related to each other, either 1 or 0.

According to the present invention having the above-mentioned configuration, anyone can easily search for depression and improvement measures without any special skill or experience. Further, according to the present invention, it is possible to make a judgment of this search solution with higher accuracy than that made by a human being. Further, by configuring the above-mentioned degree of association with artificial intelligence (neural network or the like), it is possible to further improve the discrimination accuracy by learning this.

Note that the above-mentioned input data and output data may not be completely the same in the process of learning, so that the input data and the output data may be classified by type. That is, the information P01, P02, ... P15, 16, ... That constitute the input data are classified according to the criteria classified in advance on the system side or the user side according to the content of the information, and the classified inputs. A dataset may be created between the data and the output data and trained.

The example of FIG. 8 is an example of using three or more levels of association between the reference moving image information and the level of depression. Focusing only on this degree of association, it is the same as in FIG. 3, but in this example, other reference information different from the reference moving image information is further associated with the level of depression as this search solution.

The level of depression required through the degree of association may be further modified based on reference information, or the weighting may be changed.

The reference information referred to here includes the above-mentioned reference voice information, reference text information, reference schedule information, reference appearance frequency information, and the like.

For example, as one of the reference information, it is assumed that there were many writings in the reference text information that slandered the performer. In such cases, the level of depression is often high. At this time, the "strong depression state" searched from the property information via the degree of association is subjected to a process of increasing the weighting, in other words, a process of leading to a search solution having a high level of depression. Is set in advance.

For example, it is assumed that the reference information G is an analysis result suggesting a higher level of depression, and the reference information F is an analysis result suggesting a lower level of depression. After the setting with the reference information in this way, if the actually acquired information is the same as or similar to the reference information G, a process of increasing the weighting at a high level of depression is performed. On the other hand, when the actually acquired information is the same as or similar to the reference information F, a process of increasing the weighting at a low level of depression is performed. That is, the degree of association itself leading to the level of depression may be controlled based on the reference information F to H. Alternatively, after the level of depression is determined only by the above-mentioned degree of association, the obtained search solution may be modified based on the reference information F to H. In the latter case, how to modify the level of depression as a search solution based on the reference information F to H will reflect what was designed on the system side each time.

Further, the reference information is not limited to the case where it is composed of any one type, and the solution search may be performed based on two or more types of reference information. Similarly, in such a case, the higher the level of depression suggested by the reference information is, the higher the level of depression as the search solution obtained through the degree of association is corrected, and the reference information is corrected. The lower the level of depression suggested by, the lower the level of depression as a search solution obtained through the degree of association.

Further, according to the present invention, there is a feature that the optimum solution search is performed through the degree of association set in three or more stages. The degree of association can be described by, for example, a numerical value from 0 to 100% in addition to the above-mentioned 10 steps, but is not limited to this, and any step can be described as long as it can be described by a numerical value of 3 or more steps. It may be configured.

By discriminating the most probable depression state and improvement measures based on the degree of association expressed by such numerical values of three or more stages, in a situation where there are multiple possible candidates for search solutions, the degree of association can be determined. It is also possible to search and display in descending order. If the user can be displayed in descending order of the degree of association in this way, it is possible to preferentially display more probable search solutions.

In addition to this, according to the present invention, it is possible to judge without overlooking the discrimination result of the extremely low output such as 1% of the degree of association. It warns the user that even a judgment result with an extremely low degree of association is connected as a slight sign, and may be useful as the judgment result once every tens or hundreds of times. be able to.

Further, according to the present invention, there is a merit that the search policy can be determined by the method of setting the threshold value by performing the search based on the degree of association of three or more stages. If the threshold value is lowered, even if the above-mentioned degree of association is 1%, it can be picked up without omission, but it is unlikely that a more appropriate discrimination result can be detected favorably, and a lot of noise may be picked up. be. On the other hand, if the threshold value is raised, there is a high possibility that the optimum search solution can be detected with high probability, but the degree of association is usually low and it is passed through, but it is suitable to appear once in tens or hundreds of times. Sometimes the solution is overlooked. It is possible to decide which one should be emphasized based on the ideas of the user side and the system side, but it is possible to increase the degree of freedom in selecting the points to be emphasized.

Further, in the present invention, the above-mentioned degree of association may be updated. This update may reflect information provided, for example, via a public communication network such as the Internet. In addition, when moving images for reference, audio information for reference, text information for reference, schedule information for reference, etc. are acquired, and when knowledge, information, and data regarding depression and improvement measures for these are acquired, the degree of association is determined accordingly. Raise or lower.

In other words, this update is equivalent to learning in terms of artificial intelligence. It can be said that it is a learning act because it acquires new data and reflects it in the learned data.

In addition, this update of the degree of association is done by the system side or the user side based on the contents of research data, papers, conference presentations, newspaper articles, books, etc. by experts, except when it is based on information that can be obtained from the public communication network. It may be updated artificially or automatically. Artificial intelligence may be utilized in these update processes.

In addition, the process of first creating a trained model and the above-mentioned update may use not only supervised learning but also unsupervised learning, deep learning, reinforcement learning, and the like. In the case of unsupervised learning, instead of reading and training the data set of input data and output data, information corresponding to the input data is read and trained, and the degree of association related to the output data is self-formed from there. You may let it.

2nd Embodiment In the 2nd embodiment, it is premised that a combination of the reference pronoun frequency information and the reference tone information is formed, for example, as shown in FIG.

The reference pronoun frequency information is information indicating how much the pronoun is included in the text data when the voice spoken by the subject who determines the sign of depression is converted into text data. For example, the subject says, "I will go to Osaka with Fujimoto tomorrow by 13:00 on the Shinkansen," and "I will go to Osaka tomorrow. The meaning of the former is clearer, while the meaning of the latter is unclear. Depressed patients have a higher proportion of pronouns in their spoken textual data. By extracting this in units of text data, it becomes reference pronoun frequency information.

In actually quantifying the frequency of pronouns in text data, the volume of the entire text data should be detected by counting the number of phrases, words, case components, noun phrases, and characters in the text data. You may do it. Then, the volume of the pronouns included in the volume of the entire text data is similarly counted via the number of clauses, the number of words, the number of case components, the number of noun phrases, the number of characters, and the like. Then, the ratio of the volume of the pronoun to the volume of the entire text data is detected as the frequency described above. It is premised that the units (number of phrases, number of words, number of case components, number of noun phrases, number of characters, etc.) for measuring the entire text data and the volume of pronouns are shared with each other.

In the present invention, in detecting the reference pronoun frequency information, the input of the voice of the past subject is accepted. This input may be accepted via a microphone or the like. Then, the voice of the subject is converted into text data and morphologically analyzed to extract the pronoun included in the text data. For pronouns, words such as "that", "it", and "kore" are extracted by morphological analysis. Similarly, morphological analysis is used when extracting phrase structures (case components, noun phrases, etc.).

Reference tone information is information related to voice tones extracted from past subjects. The tone of this voice refers to, for example, the pitch of the sound (the number of vibrations per second of the sound wave, that is, the frequency), the sound itself, or the strength of the sound. The tone of the voice may be detected and analyzed by detecting the height and the strength of the voice through a general voice detector.

The reference tone information may be linked with the text data in the reference pronoun frequency information. For example, in the phrase "I'm going to Osaka tomorrow, that, that, that," the voice tone is associated with each noun phrase (case component) such as "tomorrow" and "that". It may be used as reference tone information.

As input data, such reference pronoun frequency information and reference tone information are lined up. The intermediate node 61 shown in FIG. 9 is a combination of the reference pronoun frequency information and the reference tone information as such input data. Each intermediate node 61 is further connected to an output. In this output, the discriminant types A to E of the signs of depression are displayed. This discriminant type of sign of depression indicates all types of signs of depression, respectively. The types of discrimination of signs of depression are, for example, A is "no abnormality", B is "severe depression", and C is "a preliminary group that is not depressed but has signs of depression". In addition to this, the discriminant type of the sign of depression may indicate the magnitude and degree of the symptom of depression.

Each combination (intermediate node) of the reference pronoun frequency information and the reference tone information is associated with each other through three or more levels of association with the discrimination type of the sign of depression as this output solution. ..

The search device 2 acquires in advance the degree of association w13 to w22 of three or more stages shown in FIG. That is, the search device 2 accumulates past data as to which of the reference pronoun frequency information, the reference tone information, and the discrimination type of the sign of depression in that case was suitable for discriminating the actual search solution. By analyzing and analyzing these, the degree of association shown in FIG. 9 is created.

In the process of accumulating these data, past subjects who actually suffered from depression, past subjects who did not suffer from depression, and even those who did not have depression, were placed in the preliminary group. Reference synonym frequency information and reference tone information are detected from the past subjects who have become. At the same time, the subject actually applied the result of discrimination of the signs of depression by a specialist or a doctor to a predetermined discrimination type, digitized this, and the above-mentioned reference pronoun frequency. You may want to train a dataset of information and reference tone information.

In the process of constructing this learning data, it is not limited to the case of actually extracting data from the subject, but assuming a fictitious subject, a certain reference pronoun frequency information and a reference tone. If it is information, it may be determined what kind of depressive sign discrimination type is actually applied to it, and it may be converted into data and learned.

This analysis may be performed by artificial intelligence. Further, the degree of association shown in FIG. 9 may be composed of the nodes of the neural network in artificial intelligence. That is, the weighting coefficient for the output of the node of this neural network corresponds to the above-mentioned degree of association. Further, the network is not limited to a neural network, and may be composed of all decision-making factors constituting artificial intelligence.

After creating such learned data, the above-mentioned learned data will be used in the actual search for proposing the type of discrimination of signs of depression. In such a case, tone information is acquired in addition to pronoun frequency information from a new subject who newly proposes a discrimination type of signs of depression. In addition to such pronoun frequency information, the method of acquiring tone information is the same as the above-mentioned reference pronoun frequency information and reference tone information.

Based on the newly acquired pronoun frequency information and tone information in this way, the optimum type of discrimination for signs of depression is searched for. In such a case, the degree of association shown in FIG. 9 (Table 1) acquired in advance is referred to. For example, when the newly acquired pronoun frequency information is the same as or similar to P02 and the tone information is the same as or similar to P17, the node 61d is associated via the degree of association. The node 61d is associated with a depression sign discrimination type C at w19 and a depression sign discrimination type D at a degree of association w20. In such a case, the discriminant type C of the sign of depression with the highest degree of association is selected as the optimum solution. However, it is not essential to select the one with the highest degree of association as the optimal solution, and the discriminant type D of the signs of depression in which the degree of association itself is recognized although the degree of association is low may be selected as the optimum solution. .. In addition to this, it goes without saying that an output solution to which the arrows are not connected may be selected, and any other output solution may be selected in any other priority as long as it is based on the degree of association.

It should be noted that the combination of the reference pronoun frequency information and the reference facial expression image information and the discrimination type of the sign of depression for the combination may be set to three or more levels of association.

The reference facial expression image information is information related to the facial expression image of the subject who acquires the reference pronoun frequency information. The reference facial expression image information may be obtained by analyzing the image data obtained by capturing the facial expression of the subject with a camera to extract a characteristic portion for detecting the depression. .. Assuming that a depressed patient may show a peculiar smile in his facial expression, the presence or absence of the peculiar smile may be detected by analyzing a facial image. Further, the facial expression image information for reference may be composed of not only a still image but also a moving image. In the case of a moving image, it may be linked with the text data in the reference pronoun frequency information. For example, in the phrase "I will go to Osaka tomorrow, that, that,", the content of the video is in chronological order for each noun phrase (case component) such as "tomorrow" and "that". It may be associated with the target and used as reference facial expression image information. In addition to image analysis, the method of capturing such reference facial expression image information and facial expression image information may be, if necessary, using deep learning technology, automatically discriminating based on the feature amount of the analyzed image, and converting it into data. ..

The discrimination type is searched based on the newly acquired pronoun frequency information and facial expression image information in this way. Since the actual search method is the same as described above, the description below will be omitted.

As an alternative to the reference pronoun frequency information, the reference intent information described below may be used. This reference intent information is information managed in each processing operation unit included in the text data, and defines an action name.

An intent usually defines an action that specifies a business process (processing operation). For example, all actions such as "throw in the trash", "eat rice", "watch TV", "go shopping", "ride the train", "listen to music" are defined as intents.

The text data is morphologically analyzed and intents are assigned to each of them. This intent allocation refers to the intent table created and saved in advance.

The intent table defines which intent contains the wording of the morphological analysis. For example, in the case of the intent "go shopping", the morphologically analyzed wording included in this is "go shopping", "go shopping", "go shopping", and "procure" in addition to "go shopping". Various things such as "come" are included. Similarly, in the case of the intent "get on the train", there are various morphologically analyzed words included in this, such as "go on the Yamanote line", "get on the Chuo line", "use the train", and "use the train". Things are included.

In the intent table, various words that have been morphologically analyzed are associated and recorded for each such intent, and by reading this, it is possible to assign intents to each of the words that have been morphologically analyzed. Will be.

The combination of the reference intent information consisting of such intents and the reference facial expression image information, and the discrimination type of the sign of depression as output data are associated with each other through the degree of association of the intermediate node 61. Let me learn.

Then, intent information and facial expression image information are newly extracted from the subject, and the signs of depression as a search solution are analyzed via the corresponding reference intent information.

It should be noted that the combination of the reference intent information and the reference tone information and the discrimination type of the sign of depression as the output data are related to each other through the degree of association of the intermediate node 61 and learned. , When the intent information and the tone information are newly input from the subject, it is possible to search the search solution in the same manner.

As an alternative to the reference pronoun frequency information, three or more levels of association between the combination with the reference EEG information described below and the discrimination type of the sign of depression for the combination may be set. ..

The reference brain wave information is information related to the brain wave of the subject. The electroencephalogram of the subject can be measured from a commercially available electroencephalograph. Since it may be possible to grasp the signs of depression by making a judgment by combining such reference EEG information, this is added as an explanatory variable. This reference electroencephalogram information may be composed of information that captures changes over time. In such a case, it may be linked with the text data in the reference pronoun frequency information. For example, in the phrase "I will go to Osaka tomorrow, that, that,", the changes in brain waves over time for each noun phrase (case component) such as "tomorrow" and "that". May be associated with and used as reference pronoun frequency information.

Similarly, when such a degree of association is set, the sign of depression is discriminated based on the pronoun frequency information newly acquired from the subject to be examined and the electroencephalogram information.

As an alternative to the reference pronoun frequency information, an example is shown in which a combination with the reference attribute information described below and a three-level or higher degree of association with the discrimination type of the sign of depression for the combination are set. ..

Reference attribute information is information indicating the attributes of the subject. The attributes of the subject are the subject's age, gender, occupation, current social activities, information on behaviors and behaviors related to depression from the past to the present, various diseases other than depression, etc. Information about is also included.

Similarly, even when such a degree of association is set, it is determined based on the pronoun frequency information newly acquired from the subject and the attribute information.

The present invention is not limited to the case of searching for a discrimination type as a search solution, and as shown in FIG. 10, the prescription can be used as a search solution by learning a prescription according to the discrimination type in advance. Can be output. After verifying in advance what kind of prescription is effective for what kind of discrimination type, an effective prescription is associated with each discrimination type. Then, the discriminant type may be searched in the same manner as described above, and the prescription associated with the searched discriminant type may be output together with the discriminant type or as a substitute for the discriminant type. Further, as an alternative to the discrimination type, the prescription itself may be trained as a data set with the above-mentioned reference information. As a result, when the input data acquired from the subject is input, a more effective prescription will be output straight.

FIG. 10 shows an example in which the degree of association is formed by combining the reference synonym frequency information and the reference tone information, but the present invention is not limited to this, and all the above-mentioned reference information can be used. As for the degree of association as input data, it is of course possible to train the prescription itself as a data set with the above-mentioned reference information.

Third Embodiment In the third embodiment, the discrimination type of the sign of depression is discriminated only from the reference pronoun frequency information. For example, as shown in FIG. 11, the degree of association between the reference pronoun frequency information acquired in the past and the discrimination type of the sign of depression actually discriminated in the past is used.

Such degree of association is what is called learned data in artificial intelligence. After creating such trained data, the above-mentioned trained data will be used when actually determining a new sign of depression from now on. In such a case, the pronoun frequency information is newly acquired. The acquisition method is the same as that of the first and second embodiments.

Based on the pronoun frequency information newly acquired in this way, the discrimination type of the sign of depression is discriminated. In such a case, the degree of association shown in FIG. 11 acquired in advance is referred to. Since the specific method for estimating the type of discrimination of signs of depression is the same as that of the first and second embodiments, the description below will be omitted.

Further, in the third embodiment, the discrimination type of the sign of depression may be discriminated only from the reference intent information. In such a case, for example, as shown in FIG. 12, the degree of association between the reference intent information acquired in the past and the discrimination type of the sign of depression actually discriminated in the past is used.

Such degree of association is what is called learned data in artificial intelligence. After creating such trained data, the above-mentioned trained data will be used when actually determining a new sign of depression from now on. In such a case, intent information is newly acquired. The acquisition method is the same as that of the first and second embodiments.

Based on the newly acquired intent information in this way, the type of discrimination of signs of depression is discriminated. In such a case, the degree of association shown in FIG. 12 acquired in advance is referred to. Since the specific method for estimating the type of discrimination of signs of depression is the same as that of the first and second embodiments, the description below will be omitted.

Further, in the third embodiment, the discrimination type of the sign of depression may be discriminated only from the reference tone information. In such a case, for example, as shown in FIG. 13, the degree of association between the reference tone information acquired in the past and the discrimination type of the sign of depression actually discriminated in the past is used.

Such degree of association is what is called learned data in artificial intelligence. After creating such trained data, the above-mentioned trained data will be used when actually determining a new sign of depression from now on. In such a case, tone information is newly acquired. The acquisition method is the same as that of the first and second embodiments.

Based on the tone information newly acquired in this way, the discrimination type of the sign of depression is discriminated. In such a case, the degree of association shown in FIG. 13 acquired in advance is referred to. Since the specific method for estimating the type of discrimination of signs of depression is the same as that of the first and second embodiments, the description below will be omitted.

Similarly, in the third embodiment, the search solution is not limited to the case of searching for the discrimination type, and the prescription is output as the search solution by learning the prescription according to the discrimination type in advance. Of course, you may do so.

Further, both the first embodiment to the third embodiment are not limited to the above-described embodiments, and as shown in FIG. 13, for example, reference information as a keynote and a type for discriminating signs of depression. You may use the degree of association of 3 or more levels. In such a case, the solution search will be performed based on the degree of association with the discrimination type of the sign of depression according to the newly acquired information in three or more stages. The underlying reference information is, for example, reference pronoun frequency information, etc., but is not limited thereto, and any reference information (reference pronoun frequency information, reference) in the first to third embodiments. Tone information for reference, facial expression image information, intent information for reference, brain wave information for reference, attribute information for reference, etc.) can also be applied.

The search solution obtained through the degree of association may be further modified based on other reference information, or the weighting may be changed, as shown in FIG.

The other reference information referred to here corresponds to any reference information other than the reference information which is the keynote when any of the above-mentioned reference information is used as the keynote reference information.

For example, as one of the other reference information, it is assumed that in a certain reference tone information P14, B was previously discriminated as a discriminant type of a sign of depression. When the tone information corresponding to the reference tone information P14 is newly acquired, the search solution B as the discrimination type of the sign of depression is processed to increase the weight, in other words, the sign of depression. It is set in advance to perform a process that leads to the search solution B as the discrimination type of.

For example, the other reference information G is an analysis result suggesting a search solution C as a discriminant type of a sign of more depression, and the reference information F is a search as a discriminant type of a sign of more depression. It is assumed that the analysis result suggests the solution D. When the information actually acquired is the same as or similar to the reference information G after the setting with the reference information in this way, the weighting of the search solution C as a discrimination type of the sign of depression is increased. Perform processing. On the other hand, when the actually acquired information is the same as or similar to the reference information F, the process of increasing the weighting of the search solution D as the discrimination type of the sign of depression is performed. That is, the degree of association itself leading to the discrimination type of the sign of depression may be controlled based on the reference information F to H. Alternatively, after determining the discrimination type of the sign of depression only by the above-mentioned degree of association, the obtained search solution may be modified based on the reference information F to H. In the latter case, how to modify the discrimination type of the sign of depression as a search solution based on the reference information F to H should reflect the one designed on the system side each time. Become.

Further, the reference information is not limited to the case where it is composed of any one type, and the solution search may be performed based on two or more types of reference information. Similarly, in such a case, the more the case leads to the discrimination type with the sign of depression suggested by the reference information, the higher the modification of the discrimination type as the search solution obtained through the degree of association may be made. ..

Similarly, as shown in FIG. 15, it also becomes a keynote when forming a degree of association with a discriminant type of a sign of depression for a combination having a keynote reference information and another reference information. The reference information is any reference information (reference pronoun frequency information, reference tone information, facial expression image information, reference intent information, reference brain wave information, reference attribute information) in the first to third embodiments. Etc.) are also applicable. Other reference information includes any reference information in the first to third embodiments other than the underlying reference information.

At this time, if the reference information as the keynote is the reference pronoun frequency information, the other reference information includes any reference information in the other 1st to 3rd embodiments.

In such a case, it is possible to estimate the discrimination type of signs of depression by performing a solution search in the same manner. At this time, as shown in FIG. 14 described above, the signs of depression are discriminated from the search solution obtained through the degree of association through further reference information (reference information F, G, H, etc.). You may try to modify the type.

Also in the third embodiment, the degree of association may be learned by combining not only one but also two or more other reference information.

As the above-mentioned search solution, a prescription for depression may be searched as a search solution as an alternative to the type of discrimination of signs of depression. In such a case, the search solution can be similarly obtained by preparing data associated with the prescription of depression for the combination of the above-mentioned basic reference information and other reference information through three or more levels of association. Can be explored.

Fourth Embodiment As shown in FIG. 16, in the fourth embodiment, various references are used when the degree of association between the reference information as the keynote and the discrimination type of the sign of depression is used in three or more stages. An example of various variation development will be explained.

As an example of the reference information that serves as the keynote, there is reference magnetoencephalogram information obtained by performing a magnetoencephalogram examination on a past subject. This reference magnetoencephalogram information may be composed of image data obtained by MEG (magnetoencephalography examination), or may be composed of MRI data scanned for the brain. The reference magnetoencephalogram information may be further composed of data obtained by extracting characteristic portions by using a deep learning technique known for these image data. Since the degree of depression and these magnetoencephalographic information are considered to be correlated, the above-mentioned degree of association is constructed in advance by learning the data sets in which they are associated with each other.

Such degree of association is what is called learned data in artificial intelligence. After creating such trained data, the above-mentioned trained data will be used when actually determining a new sign of depression from now on. In such a case, the magnetoencephalogram information is newly acquired.

The newly acquired magnetoencephalogram information is acquired from a subject who newly discriminates a sign of depression, and the type and acquisition method of the information are the same as the above-mentioned reference magnetoencephalogram information. Since the specific solution search method is the same as that of each of the above-described embodiments, the description below will be omitted.

As an example of the reference information that serves as the keynote, there is reference meal content information regarding the meal content of the subject in the past. This reference meal content information is the content of the meal actually ingested by the subject in the past (meal, etc.), the amount of the meal, the time of the meal, the time interval for taking three meals, and the number of nutrients and calories ingested. This reference meal content may be obtained by dividing it into periods such as one day, one week, and one month.

Since the degree of depression and the information on these dietary contents are considered to be correlated, the above-mentioned degree of association is constructed in advance by learning the data sets in which they are associated with each other.

Such degree of association is what is called learned data in artificial intelligence. After creating such trained data, the above-mentioned trained data will be used when actually determining a new sign of depression from now on. In such a case, new meal content information is acquired.

The newly acquired meal content information is acquired from the subject who newly determines the sign of depression, and the type and acquisition method of the information is the same as the above-mentioned reference meal content information. Since the specific solution search method is the same as that of each of the above-described embodiments, the description below will be omitted.

As an example of the reference information that serves as the keynote, there is reference sleep information regarding the sleep state of the subject in the past. This reference sleep information is all information about the sleep actually performed by the subject in the past, for example, sleep time, sleep depth, number of times of waking up during sleep and time zone, body movement during sleep. Contains all the data about. Since patients with depression have data such as sleep division and abundant awakenings, the above-mentioned degree of association is constructed in advance by learning a data set in which these are associated with each other.

Such degree of association is what is called learned data in artificial intelligence. After creating such trained data, the above-mentioned trained data will be used when actually determining a new sign of depression from now on. In such a case, sleep information is newly acquired.

The newly acquired sleep information is acquired from a subject who newly determines a sign of depression, and the type and acquisition method of the information are the same as the above-mentioned reference meal content information. Since the specific solution search method is the same as that of each of the above-described embodiments, the description below will be omitted.

As an example of the reference information that serves as the keynote, there is reference exercise information regarding the exercise state of the subject in the past. This reference exercise information is all information related to the exercise actually performed by the subject in the past, and is composed of, for example, the number of steps per unit time, the time of going out, the number of times of going out, the distance walked, the distance traveled, and the like. May be good. Further, the reference exercise information may be composed of information obtained by measuring the swing of the arm, the movement of the foot, the walking speed, etc. through the acceleration sensor actually attached to the subject. Further, the reference exercise information may indicate the heart rate detected through a heart rate monitor worn on the body, and may indicate the actual activity range via the position information detected through GPS or the like.

Since patients with depression are considered to have a correlation with the amount of exercise, the above-mentioned degree of association is constructed in advance by learning a data set in which these are associated with each other.

Such degree of association is what is called learned data in artificial intelligence. After creating such trained data, the above-mentioned trained data will be used when actually determining a new sign of depression from now on. In such a case, exercise information is newly acquired.

The newly acquired exercise information is acquired from a subject who newly determines the signs of depression, and the type and acquisition method of the information is the same as the above-mentioned reference exercise information. Since the specific solution search method is the same as that of each of the above-described embodiments, the description below will be omitted.

As an example of the reference information that serves as the keynote, there is reference life time information regarding the life time zone of the subject in the past. This reference life time information is any information about the life time actually performed by the subject in the past, and for example, the time zone of the toilet may be acquired through the electricity usage time of the toilet. The time actually awake and the time of sleep may be detected through the time of electricity usage in the room. Further, it may be possible to detect which room and how many hours have been stayed through the door open / close sensor. The electricity usage time may be detected via an electricity usage meter set in a house, or may be detected via an electricity meter installed in a so-called smart house or the like.

Since patients with depression are considered to have a correlation with their life time, the above-mentioned degree of association is constructed in advance by learning a data set in which they are associated with each other.

Such degree of association is what is called learned data in artificial intelligence. After creating such trained data, the above-mentioned trained data will be used when actually determining a new sign of depression from now on. In such a case, new life time information is acquired.

The newly acquired life time information is acquired from the subject who newly determines the sign of depression, and the type and acquisition method of the information are the same as the above-mentioned reference life time information. Since the specific solution search method is the same as that of each of the above-described embodiments, the description below will be omitted.

As an example of the reference information that serves as the keynote, there is reference operation information related to various operations in the subject's house in the past. This reference operation information includes electrical appliances (TVs, air conditioners, PCs, videos, vacuum cleaners, refrigerators, washing machines), lighting switches, bath gas switches, and system kitchens that past subjects actually have in their homes. Anything that any resident of a house can operate, such as the operation buttons and levers of the house, the operation of windows that open and close with a push button, is targeted. These reference operation information may be detected through IoT sensors installed on push buttons such as electric appliances, lighting switches in homes, gas switches for baths, operation buttons and levers in system kitchens, and the like. It may be detected via an electric meter installed in a so-called smart house or the like.

Depressed patients are considered to have a correlation with the frequency of pressing mistakes, pressing time, pressing strength, etc. when operating buttons, etc., so by learning the data sets associated with each other, the above-mentioned Build the degree of association in advance.

Such degree of association is what is called learned data in artificial intelligence. After creating such trained data, the above-mentioned trained data will be used when actually determining a new sign of depression from now on. In such a case, the operation information is newly acquired.

The newly acquired operation information is acquired from a subject who newly determines a sign of depression, and the type and acquisition method of the information are the same as the above-mentioned reference operation information. Since the specific solution search method is the same as that of each of the above-described embodiments, the description below will be omitted.

As an example of the reference information that serves as the keynote, there is reference line-of-sight image information regarding the movement of the line-of-sight of the subject in the past. This reference line-of-sight image information is an image of the line-of-sight of the subject in the past for the question given. In performing the image analysis of the line of sight, a deep learning technique may be used to automatically determine the image based on the feature amount of the analyzed image and convert it into data. Further, an infrared camera may be used to detect the movement of the subject's line of sight to a high degree.

Questions to be asked include, for example, displaying figures of various shapes at the same time, counting how many of the same things exist, analyzing the line of sight when moving the mark displayed on the screen, or It may be a problem that tests spatial cognitive ability, judgment, etc., such as a problem that causes a so-called mistaken search.

Depressed patients are thought to have characteristic gaze movements for such problems, so by training the datasets in which they are associated with each other, the above-mentioned degree of association is pre-constructed. ..

Such degree of association is what is called learned data in artificial intelligence. After creating such trained data, the above-mentioned trained data will be used when actually determining a new sign of depression from now on. In such a case, the line-of-sight image information is newly acquired.

The newly acquired operation information is an image of the movement of the line of sight with respect to the problem given to the subject who newly determines the sign of depression, and the type and acquisition method of the information are the above-mentioned reference operations. Similar to information. Since the specific solution search method is the same as that of each of the above-described embodiments, the description below will be omitted.

As an example of the reference information that serves as the keynote, there is reference conversation amount information regarding the conversation amount per unit time. The unit time may be any time unit such as 1 minute, 10 minutes, 60 minutes, 3 hours, and one day. The conversation volume may be simply composed of the time during which the voice is emitted, that is, the time during which the voice exceeding a predetermined volume is detected divided by the unit time. In addition, the amount of conversation is the amount of each part of speech (verb, adjective, object, pronoun, adverb, etc.) in the text data when the subject's voice is converted into text data. It may be configured with. In addition, the volume of the entire text data may be detected by converting the voice into text data and counting the number of phrases, the number of words, the number of case components, the number of noun phrases, and the number of characters. Then, the amount of the volume of the entire text data per unit time may be used as the conversation amount. It is premised that the units (number of phrases, number of words, number of case components, number of noun phrases, number of characters, etc.) for measuring the entire text data and the volume of pronouns are shared with each other.

Since the degree of symptoms of a depressed patient may be related to such conversational volume, the above-mentioned degree of association is constructed in advance by learning a data set in which these are associated with each other.

Such degree of association is what is called learned data in artificial intelligence. After creating such trained data, the above-mentioned trained data will be used when actually determining a new sign of depression from now on. In such a case, the conversation volume information is newly acquired.

For the newly acquired conversation amount information, the voice of the subject who newly determines the sign of depression is converted into text data, and the conversation amount per unit time is obtained in the same manner as described above. Since the specific solution search method is the same as that of each of the above-described embodiments, the description below will be omitted.

Further, both the first to fourth embodiments are not limited to the above-described embodiments, and as shown in FIG. 13, for example, reference information as a keynote and a type for discriminating signs of depression. You may use the degree of association of 3 or more levels. In such a case, the solution search will be performed based on the degree of association with the discrimination type of the sign of depression according to the newly acquired information in three or more stages. The underlying reference information is, for example, reference pronoun frequency information, etc., but is not limited thereto, and any reference information (reference pronoun frequency information, reference) in the first to fourth embodiments. Tone information for reference, facial image information, intent information for reference, brain wave information for reference, attribute information for reference, encephalogram information for reference, meal content information for reference, sleep information for reference, driving information for reference, operation information for reference. , Reference line-of-sight image information, reference conversation volume information, reference life time information, etc.) are also applicable.

The search solution obtained through the degree of association may be further modified based on other reference information, or the weighting may be changed.

For example, as one of the other reference information, it is assumed that B was previously discriminated as a discriminant type of a sign of depression in a certain reference life time information P14. When the tone information corresponding to the reference life time information P14 is newly acquired, the search solution B as the discrimination type of the sign of depression is processed to increase the weight, in other words, the depression. It is set in advance to perform a process that leads to the search solution B as a symptom discrimination type.

Similarly, as shown in FIG. 15, it also becomes a keynote when forming a degree of association with a discriminant type of a sign of depression for a combination having a keynote reference information and another reference information. The reference information is any reference information (reference pronoun frequency information, reference tone information, facial image image information, reference intent information, reference brain wave information, reference attribute information) in the first to fourth embodiments. , Reference electroencephalogram information, reference meal content information, reference sleep information, reference driving information, reference operation information, reference line-of-sight image information, reference conversation volume information, reference life time information, etc.) are also applicable. Is. Other reference information includes any reference information in the first to fourth embodiments other than the underlying reference information.

At this time, if the reference information as the keynote is the reference magnetoencephalogram information, the other reference information includes any reference information in the other 1st to 4th embodiments.

Also in the fourth embodiment, the degree of association may be learned by combining not only one but also two or more other reference information.

1 Depression state discrimination system 2 Discrimination device 21 Internal bus 23 Display unit 24 Control unit 25 Operation unit 26 Communication unit 27 Estimating unit 28 Storage unit 61 Node

Claims

In the depression state determination program that determines the depression state of performers appearing in the media
The information acquisition step to acquire the moving image information consisting of the moving images of the above performers,
The reference moving image according to the moving image information acquired in the above information acquisition step by using the three or more levels of association between the reference moving image information captured in advance as a moving image of a person's movement and the level of depression. A depression state determination program characterized by having a computer execute the depression state determination step for determining the depression state of the performer based on the degree of association between the information and the level of the depression state.
In the above information acquisition step, the voice information of the above performers is acquired, and the voice information is acquired.
In the depression state determination step, the information acquisition is performed by using the combination of the reference moving image information and the reference voice information recorded with human voice, and the degree of association between the level of the depression state and the level of the depression state. Based on the degree of association between the combination of the reference moving image information according to the moving image information acquired in the step and the reference audio information according to the audio information and the level of the depressed state, the above performers The depression state determination program according to claim 1, wherein the depression state is determined.
In the depression state determination program that determines the depression state of performers appearing in the media
The information acquisition step to acquire the voice information of the above performers,
Using the three or more levels of association between the reference voice information recorded from human voice and the level of depression, the reference voice information and the level of depression according to the voice information acquired in the above information acquisition step. A depression state determination program characterized by causing a computer to execute the depression state determination step for determining the depression state of the performer based on the degree of association of three or more stages.
In the above information acquisition step, the text information written about the above performers is acquired on various information sites, and the text information is acquired.
In the depression state determination step, the information acquisition is performed by using the combination of the reference video information, the reference text information in which the text data is categorized in advance, and the degree of association between the depression level and the depression state. Based on the degree of association between the combination of the reference video information according to the video information acquired in the step and the reference text information according to the text information and the level of depression, the above performers The depression state determination program according to claim 1, wherein the depression state is determined.
In a depression symptom determination program that determines the signs of depression in a subject
A reception step that accepts the voice input of the subject,
An information acquisition step of acquiring intent information related to an intent that defines an action that specifies a processing operation included in the text data by converting the voice received by the reception step into text data and performing morphological analysis. When,
Refer to the three or more levels of association between the reference intent information regarding the intent that defines the action that specifies the processing operation and the discrimination type of the sign of depression, and use the intent information acquired by the above information acquisition step. Based on the above, a depression sign determination program characterized by causing a computer to perform a determination step for determining a sign of depression in a subject.
In a depression symptom determination program that determines the signs of depression in a subject
A reception step that accepts the voice input of the subject,
An information acquisition step for acquiring tone information regarding the tone of the voice received by the above reception step, and
Based on the tone information acquired by the information acquisition step, the degree of association between the reference tone information regarding the tone of the voice and the discrimination type of the sign of depression is referred to, and the depression of the subject is described. A depression symptom discrimination program characterized by having a computer perform a symptom discrimination step.
In a depression symptom determination program that determines the signs of depression in a subject
The information acquisition step to acquire the line-of-sight image information that captures the line-of-sight of the subject for the question that was asked,
The line-of-sight acquired by the above information acquisition step with reference to the degree of association between the reference line-of-sight image information obtained by capturing the line-of-sight of the subject in the past for the question asked and the discrimination type of the sign of depression in three or more stages. A depression sign discrimination program characterized by causing a computer to perform a discrimination step for discriminating a sign of depression of a subject based on image information.
In a depression symptom determination program that determines the signs of depression in a subject
A reception step that accepts the voice input of the subject,
The information acquisition step of converting the voice received by the above reception step into text data and acquiring the conversation amount information obtained by calculating the conversation amount per unit time, and the information acquisition step.
Based on the conversation volume information acquired by the information acquisition step, the subject is examined by referring to the degree of association between the reference conversation volume information regarding the conversation volume per unit time and the discrimination type of the sign of depression at three or more levels. A depression sign discrimination program characterized by having a computer perform a discrimination step for discriminating a person's signs of depression.
The depression sign discrimination program according to any one of claims 1 to 8, wherein in the discrimination step, the degree of association corresponding to the weighting coefficient of each output of the node of the neural network in artificial intelligence is used. ..