A method and device for identifying the mental health status of a group of people

Abstract

The present invention discloses a method and device for identifying the mental health status of a group of people, which belongs to the field of artificial intelligence technology and is used to solve the technical problem that the traditional technical solution is not accurate enough in judging the mental health status of a group of people, and it is difficult to determine the mental health status of a group of people through specific indicators. The method includes: collecting facial image data and brain function image data of the user; differentially annotating the facial image data and brain function image data to obtain susceptibility indicators; by showing the preset scene interactive video to the user, collecting the user's eye movement gaze trajectory and head posture and performing negative bias analysis to obtain negative attention bias results; performing semantic and emotional analysis on the pre-collected user voice content to obtain language negative effect results; determining negative health high-risk results based on the negative attention bias results and the language negative effect results; determining the user's mental health status based on the susceptibility indicators and the negative health high-risk results.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence, and in particular to a method and device for identifying the mental health status of a group of people. Background Art

[0002] With the improvement of living standards, people are often prone to mental health problems in their daily lives and work. Among them, depression is a more serious mental health problem, which is divided into mild, moderate and severe depression according to the severity of depression. The treatment of mild depression generally adopts health monitoring, mental health education and exercise therapy; the treatment of moderate depression generally adopts health monitoring and cognitive behavioral therapy; the treatment of severe depression generally adopts health monitoring, cognitive behavioral therapy and drug therapy.

[0003] In cognitive behavioral therapy, since people with different mental health problems have different sensitivity types to non-reward signals, it is difficult for clinicians to determine the type of non-reward signals with personalized characteristics of people with mental health problems, and thus they cannot accurately determine the mental health status of people with mental health problems. Summary of the invention

[0004] The embodiments of the present application provide a method and device for identifying the mental health status of a group of people, which are used to solve the following technical problems: traditional technical solutions are not accurate enough in judging the mental health status of a group of people, and it is difficult to determine the mental health status of a group of people through specific indicators.

[0005] The present application embodiment adopts the following technical solutions:

[0006] On the one hand, an embodiment of the present application provides a method for identifying the mental health status of a group of people, the method comprising: collecting facial image data and brain function image data of the user by displaying preset physiological stimulation to the user; performing difference annotation on the facial image data and brain function image data to obtain a susceptibility index; collecting the user's eye gaze trajectory and head posture and performing negative bias analysis on the user by displaying a preset scene interaction video to the user to obtain a negative attention bias result; performing semantic analysis and sentiment analysis on the pre-collected user voice content to obtain a language negative effect result; and determining a negative health high-risk result based on the negative attention bias result and the language negative effect result; determining the user's mental health status based on the susceptibility index and the negative health high-risk result.

[0007] The embodiments of the present application can quickly and accurately judge the mental health status of the population by acquiring susceptibility indicators and negative health high-risk results, so as to determine the mental health status and mental state of the population, provide cognitive behavioral therapy and targeted treatment decisions for clinicians, and facilitate timely detection of people with abnormal mental health. The identified mental health status of the population can be used as an important reference indicator for clinicians to conduct auxiliary treatment.

[0008] In a feasible implementation, by displaying preset physiological stimuli to the user, facial image data and brain function image data of the user are collected, specifically including: using a facial collection device, collecting the user's expression, corresponding heart rate and heart rate variability to obtain the facial image data; wherein the facial collection device is obtained by connecting a facial collection camera with an image analysis device; using a brain function collection device, collecting brain function fMRI images and the time series of the brain function fMRI images to obtain the brain function image data; wherein the brain function collection device is obtained by connecting a brain function fMRI imaging device with the image analysis device; the preset physiological stimuli include any one or more of the following: visual stimulation, olfactory stimulation and taste stimulation.

[0009] The embodiment of the present application has an image analysis device with two interfaces, and the facial acquisition camera and the brain function fMRI imaging device are respectively connected to it, so as to better integrate the acquisition of facial image data and brain function image data.

[0010] In a feasible implementation, the facial image data is differentially annotated, specifically comprising: before presenting the preset physiological stimulation process to the user, collecting the facial image data of the user, and analyzing the facial image data to obtain a pre-stimulation state time series; during the process of presenting the preset physiological stimulation to the user, re-collecting the facial image data of the user, and re-analyzing the facial image data to obtain a stimulation state time series; establishing a normal distribution curve based on the time series difference between the pre-stimulation state time series and the stimulation state time series; if the difference value on the normal distribution curve is less than 30% of the normal distribution probability, differentially annotating the corresponding stimulation in the preset physiological stimulation to obtain a facial stimulation index.

[0011] The embodiment of the present application collects the user's status before and after a preset physiological stimulation process, and establishes a normal distribution, which is conducive to better finding the corresponding stimulation at the difference and effectively obtaining the facial stimulation index.

[0012] In a feasible implementation, the facial image data and the brain function image data are differentially annotated to obtain a susceptibility index, specifically including: monitoring the activation time course of the lateral orbitofrontal cortex area in the brain function image data to obtain the activation time length of each stimulus in the preset physiological stimulation, and the activation degree of each stimulus in the lateral orbitofrontal cortex area; if there are differences between the activation time length and the activation degree of each stimulus in the preset physiological stimulation and the normal activation time course, the difference is differentially annotated to obtain a brain stimulation index; the facial stimulation index and the brain stimulation index after the difference annotation of the facial image data are aggregated and unified to obtain the susceptibility index, and the susceptibility index is displayed on the screen; wherein, the susceptibility index is used to measure the mental health state of the user after being subjected to the preset physiological stimulation.

[0013] The embodiment of the present application identifies the activation level of the user's lateral orbitofrontal cortex area, finds the corresponding stimulation in the preset physiological stimulation, and marks the difference, so as to better identify the difference points of the corresponding stimulation in the brain area.

[0014] In a feasible implementation manner, by displaying a preset scene interaction video to the user, collecting the user's eye gaze trajectory and head posture and performing negative bias analysis to obtain a negative attention bias result, specifically comprising: by displaying the preset scene interaction video to the user, collecting the user's eye gaze trajectory during the display process, and obtaining a trajectory distribution curve; wherein the content of the preset scene interaction video includes calm tone content and negative tone content; using a head camera to collect the user's head posture, and according to the orientation position of the head posture, establishing a head cube framework model; according to the head cube framework model, analyzing the user's head posture The method further comprises the steps of: performing orientation framework analysis on the trajectory distribution curve to obtain the head orientation of the user; if the distribution ratio of the negative tone content in the trajectory distribution curve is greater than the distribution ratio of the calm tone content, and the position of the head orientation is at the position where the negative tone content is located, then the user has a negative attention bias; if the distribution ratio of the negative tone content in the trajectory distribution curve is less than or equal to the distribution ratio of the calm tone content, or the position of the head orientation is not at the position where the negative tone content is located, then the user does not have a negative attention bias; wherein the negative attention bias result includes the existence of a negative attention bias and the non-existence of a negative attention bias.

[0015] In a feasible implementation, before performing semantic analysis and sentiment analysis on pre-collected user voice content to obtain language negative effect results, the method also includes: collecting the user's voice content and converting the voice content into text content; performing semantic analysis on the text content to obtain the abstract level, attribution method, and positive and negative valence of the text semantics; establishing a semantic model of the text content based on the abstract level and attribution method; and establishing a sentiment model of the text content based on the positive and negative valence.

[0016] In a feasible implementation manner, semantic analysis and sentiment analysis are performed on the pre-collected user voice content to obtain the language negative effect result, which specifically includes: performing semantic cause analysis on the user's text content according to the language key words and high-frequency vocabulary library in the semantic model to determine the cause analysis level; wherein the cause analysis level includes the interpersonal level and the abstract level; determining the attribution type of the text content according to the keyword path network of the attribution method in the semantic model; wherein the attribution type includes internal attribution and external attribution; performing similarity analysis on the emotion vocabulary set in the text content according to the emotion vocabulary library in the emotion model to obtain the emotion valence of the text content; wherein the emotion valence includes positive emotion and negative emotion; if the cause analysis level is the abstract level, and the attribution type is internal attribution, and the emotion valence is negative emotion, then the language negative effect result is that there is a language negative effect; if the cause analysis level is the interpersonal level, or the attribution type is external attribution, or the emotion valence is positive emotion, then the language negative effect result is that there is no language negative effect.

[0017] In a feasible implementation, a negative health high-risk result is determined based on the negative attention bias result and the language negative effect result, specifically including: if the negative attention bias result is the existence of a negative attention bias, and the language negative effect result is the existence of a language negative effect, then the negative health high-risk result is a negative mental health high-risk period; if the negative attention bias result is the absence of a negative attention bias, or the language negative effect result is the absence of a language negative effect, then the negative health high-risk result is a negative mental health low-risk period.

[0018] In a feasible implementation, the user's mental health status is determined based on the susceptibility index and the negative health high-risk result, specifically including: if the index value of the susceptibility index is less than a first preset threshold, and the negative health high-risk result is a low-risk period of negative mental health, then the user's mental health status is mental health; if the index value of the susceptibility index is greater than or equal to the first preset threshold, and the negative health high-risk result is a low-risk period of negative mental health, then the user's mental health status is mild negative psychology; if the index value of the susceptibility index is less than the first preset threshold, and the negative health high-risk result is a high-risk period of negative mental health, then the user's mental health status is moderate negative psychology; if the index value of the susceptibility index is greater than or equal to the first preset threshold, and the negative health high-risk result is a high-risk period of negative mental health, then the user's mental health status is severe negative psychology.

[0019] The embodiment of the present application combines susceptibility indicators with negative health high-risk results to identify the mental health status of the detected user, which can make a based-on identification of their mental health, help determine people's mental health status, maintain their mental health, and improve their quality of life.

[0020] On the other hand, an embodiment of the present application also provides a device for identifying the mental health status of a group of people, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, so that the at least one processor can execute a method for identifying the mental health status of a group of people as described in any of the above embodiments.

[0021] The embodiments of the present application provide a method and device for identifying the mental health status of a group of people. By obtaining susceptibility indicators and negative health high-risk results, it is possible to quickly and accurately judge the mental health status of a group of people, so as to determine the mental health status and mental state of the group of people, and provide cognitive behavioral therapy and targeted treatment decisions for clinicians, which is conducive to timely discovering people with abnormal mental health, and using the identified mental health status of the group of people as an important reference indicator for clinicians to conduct auxiliary treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0023] Figure 1A flow chart of a method for identifying the mental health status of a group of people provided in an embodiment of the present application;

[0024] Figure 2 A schematic diagram of the structure of a device for identifying the mental health status of a group of people provided in an embodiment of the present application. DETAILED DESCRIPTION

[0025] In order to enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of them. Based on the embodiments of this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.

[0026] It should be noted that the mental health status in this application is mainly used to identify the user's psychological depression state, and can also be used to identify and determine the user's mental disorder.

[0027] The present application embodiment provides a method for identifying the mental health status of a group of people, such as Figure 1 As shown, the method for identifying the mental health status of a group of people specifically includes steps S101-S105:

[0028] S101. Collecting facial image data and brain function image data of a user by displaying preset physiological stimulation to the user.

[0029] Specifically, the facial image data is obtained by collecting the user's expression, corresponding heart rate and heart rate variability through a facial acquisition device. The facial acquisition device is obtained by connecting a facial acquisition camera with an image analysis device. The brain function acquisition device is used to collect brain function fMRI images and time series of brain function fMRI images to obtain brain function image data. The preset physiological stimulation obtained by connecting the brain function acquisition device with a brain function fMRI imaging device and an image analysis device includes any one or more of the following: visual stimulation, olfactory stimulation and taste stimulation.

[0030] In one embodiment, a part of the acquisition main device includes an image analysis device with two interfaces, the two interfaces are the interface of the image analysis device connected to the facial acquisition camera, and the interface of the image analysis device connected to the brain function fMRI imaging device, and the other part also includes a display screen with automatic display function and recommendation display function, a visual stimulus library storage, an olfactory stimulus generator, and a taste stimulus generator. The display screen with automatic display function and recommendation function mainly plays the following list of visual stimulation, olfactory stimulation, and taste stimulation.

[0031] Among them, visual stimulation mainly includes: young opposite sex, beautiful opposite sex, young opposite sex smiling at users, scenery of blue sky and white clouds, videos of opposite sex courtship, marriage videos, videos of couples living together, parent-child activity videos, videos of obtaining great wealth, videos of obtaining great power, videos of people serving the society, videos of people helping their relatives, videos of self-control of one's own behavior (difficult sports, gymnastics, etc.), mildly dangerous activities (extreme sports videos), physical exercise videos, videos of the process of solving complex intellectual problems, videos of collecting food and other necessities of life, videos of finding and purchasing housing, etc.

[0032] Olfactory stimulation mainly includes: odor generators simulating the production of pheromones, etc.

[0033] Taste stimulation mainly includes: simulating salty, sweet, spicy, bitter, etc.

[0034] S102: Differentially annotate the facial image data and the brain function image data to obtain a susceptibility index.

[0035] Specifically, before presenting the preset physiological stimulation process to the user, the facial image data of the user is collected, and the facial image data is analyzed to obtain a pre-stimulation state time series. During the process of presenting the preset physiological stimulation to the user, the facial image data of the user is recollected, and the facial image data is reanalyzed to obtain a stimulation state time series. According to the time series difference between the pre-stimulation state time series and the stimulation state time series, a normal distribution curve is established. If the difference value on the normal distribution curve is less than 30% of the normal distribution probability, the corresponding stimulation in the preset physiological stimulation is marked with a difference to obtain a facial stimulation index.

[0036] In one embodiment, the facial image features of the user are collected and analyzed by a facial camera, and the expression, heart rate, and heart rate variability are analyzed, which are defined as: FE, HR, HRV. Among them, FE is further divided into four types, which are defined as PA, NA, NA-, and PA-. When the above-mentioned stimulation video and the stimulation of taste and smell are displayed, the user's state before the stimulation is recorded and analyzed, which is defined as the pre-stimulation state T1 (FE, HR, HRV). After the stimulation begins, the state of the user in the process of watching and tasting the stimulation is recorded, which is defined as the stimulation state T2 (FE, HR, HRV). The difference in the data time series of T1 and T2 is recorded, and a normal distribution is established. When the difference value is in the normal distribution probability less than 30%, the corresponding stimulus is marked and annotated, which can be expressed by the formula P (Δ (T2-T1) | visual stimulation, olfactory stimulation, taste stimulation), where P is the annotation of the corresponding stimulus on the facial image data.

[0037] In addition, the activation time course of the lateral orbitofrontal cortex region in the brain function image data is monitored to obtain the activation time length of each stimulus in the preset physiological stimulation and the activation degree of each stimulus in the lateral orbitofrontal cortex region. If there are differences between the activation time length and activation degree of each stimulus in the preset physiological stimulation and the normal activation time course, the difference is marked to obtain the brain stimulation index.

[0038] In one embodiment, in an image analysis device connected to a brain function fMRI imaging device, the time series data of fMRI is mainly analyzed, and the activation time course of the lateral orbitofrontal cortex (vmOFC) region in the fMRI image data is monitored. When the activation time course is significantly different from the activation time length (t) and activation degree (a) of other stimuli, a mark is made on the corresponding stimulus, which can be expressed by the formula P(Σ(a, t) vmOFC |Visual stimulation, olfactory stimulation, taste stimulation).

[0039] Furthermore, the facial stimulation index and the brain stimulation index after the difference annotation of the facial image data are aggregated and unified to obtain the susceptibility index, and the susceptibility index is displayed on the screen. Among them, the susceptibility index is used to measure the user's mental health state after being subjected to a preset physiological stimulation.

[0040] As a feasible implementation method, the facial stimulation indicators and brain stimulation indicators marked during the test are summarized and displayed on a display screen to assist psychological counselors or clinicians in making treatment target decisions for cognitive behavioral therapy for mental health conditions.

[0041] S103: By showing the preset scene interactive video to the user, the user's eye gaze trajectory and head posture are collected and negative bias analysis is performed to obtain a negative attention bias result.

[0042] Specifically, by showing the preset scene interaction video to the user, the user's eye movement gaze trajectory during the display process is collected, and a trajectory distribution curve is obtained. The content of the preset scene interaction video includes calm tone content and negative tone content. The user's head posture is collected through the head camera, and a head cube frame model is established according to the orientation position of the head posture. According to the head cube frame model, the user's head posture is subjected to orientation frame analysis to obtain the user's head orientation.

[0043] If the distribution ratio of negative tone content in the trajectory distribution curve is greater than the distribution ratio of calm tone content, and the position of the head direction is at the location of the negative tone content, then the user has negative attention bias. If the distribution ratio of negative tone content in the trajectory distribution curve is less than or equal to the distribution ratio of calm tone content, or the position of the head direction is not at the location of the negative tone content, then the user does not have negative attention bias. Among them, the negative attention bias results include the presence of negative attention bias and the absence of negative attention bias.

[0044] In one embodiment, a set of social and humanistic scene interaction videos pre-set in a storage device are displayed on a display screen, specifically: two people with no facial expression introduce themselves in a calm tone and a depressed tone respectively; a person with a blank face and a person with a "sad fake smile" facial expression both introduce themselves in a calm tone. At the same time, an eye tracking calculation and analysis device is used to capture the user's eye gaze trajectory, which is defined as TES; a camera is used to capture the user's head posture, and a cubic frame model is established to analyze the user's head orientation, which is defined as HP. While playing the stimulus video, the user's TES and HP are calculated, and it is judged that the distribution ratio of TES in the "depressed tone" and "sad fake smile" directions is greater than that in other stimulus videos, that is, it is judged whether the user has a negative attention bias. At this time, HP meets the conditions of "towards the same stimulus" or "towards the middle position", that is, the position of the head orientation is at the position where the negative tone content is located, and the display result "there is a negative attention bias" is displayed; if other conditions occur, the display result "there is no negative attention bias" is displayed.

[0045] S104: Perform semantic analysis and sentiment analysis on the pre-collected user voice content to obtain the language negative effect results, and determine the negative health high-risk results based on the negative attention bias results and the language negative effect results.

[0046] Specifically, the user's voice content is collected and converted into text content. The text content is semantically analyzed to obtain the abstract level, attribution method, and positive and negative valence of the text semantics. According to the abstract level and attribution method, a semantic model of the text content is established. According to the positive and negative valence, an emotional model of the text content is established.

[0047] In one embodiment, a virtual person is displayed on a display screen, and the virtual person asks the user a question, which requires the user to describe his or her recent experience and analyze the cause of this experience. During the user's answer, the user's answer is collected through a microphone, and the voice is automatically converted into text. Semantic modeling and sentiment modeling are performed on the text, and the abstract level and attribution method of semantic analysis in semantic modeling are defined, and the positive and negative valence of language analysis in sentiment modeling are defined.

[0048] Furthermore, according to the key language words and high-frequency vocabulary in the semantic model, the semantic cause analysis is performed on the text content of the user to determine the cause analysis level; wherein the cause analysis level includes the interpersonal level and the abstract level. According to the keyword path network of the attribution method in the semantic model, the attribution type of the text content is determined. Among them, the attribution type includes internal attribution and external attribution. According to the emotional vocabulary library in the emotional model, the emotional vocabulary set in the text content is similarity analyzed to obtain the emotional valence of the text content. Among them, the emotional valence includes positive emotion and negative emotion. If the cause analysis level is the abstract level, and the attribution type is internal attribution, and the emotional valence is negative emotion, then the language negative effect result is that there is a language negative effect. If the cause analysis level is the interpersonal level, or the attribution type is external attribution, or the emotional valence is positive emotion, then the language negative effect result is that there is no language negative effect.

[0049] In one embodiment, the abstract level refers to that when the user analyzes his own experience and uses artificial intelligence natural language technology to analyze its semantic abstract level, two levels of abstract levels are pre-defined in the semantic model, namely, the basic level and the abstract level. The basic level pre-sets language focus words and high-frequency vocabulary library, and the vocabulary library specifies relationship demonstrative pronouns such as "you", "I", and "he", indicating that the cause analysis level is at the interpersonal level; the abstract level pre-sets language focus words and high-frequency vocabulary library, and the vocabulary library specifies abstract entity demonstrative words such as "society", "culture", and "environment", indicating that the cause analysis level is at the abstract level. By detecting the conformity of the text focus words and high-frequency words of the user's analysis of his own experience with the two levels, it is determined whether the abstract level of the user's analysis is at the interpersonal level or the abstract level.

[0050] In one embodiment, the attribution method refers to that when a user analyzes his own experience and uses artificial intelligence natural language technology to analyze his semantic attribution method, two types of attribution methods are pre-defined in the semantic model, external attribution and internal attribution. A keyword path network is pre-set in the external attribution type, and the path from "causal entity recognition" to "non-self-referential entity vocabulary" is defined as the external attribution type; a keyword path network is pre-set in the internal attribution type, and the path from "causal entity recognition" to "self-referential entity vocabulary" is defined as the internal attribution type.

[0051] In one embodiment, positive and negative valence refers to the user analyzing his own experience and using artificial intelligence natural language technology to analyze its semantic emotional valence, pre-defining a "positive vocabulary library" and a "negative vocabulary library" in the analysis language of the emotional model, identifying the similarity between the vocabulary set narrated by the user and the above two vocabulary libraries, and judging the emotional valence of the user's narration.

[0052] Furthermore, if the negative attention bias result is that there is a negative attention bias, and the language negative effect result is that there is a language negative effect, then the negative health high-risk result is a negative mental health high-risk period. If the negative attention bias result is that there is no negative attention bias, or the language negative effect result is that there is no language negative effect, then the negative health high-risk result is a negative mental health low-risk period.

[0053] S105. Determine the user's mental health status based on the susceptibility index and the negative health high-risk results.

[0054] Specifically, if the index value of the susceptibility index is less than the first preset threshold value, and the negative health high-risk result is a negative mental health low-risk period, the user's mental health state is mental health. If the index value of the susceptibility index is greater than or equal to the first preset threshold value, and the negative health high-risk result is a negative mental health low-risk period, the user's mental health state is mild negative psychology. If the index value of the susceptibility index is less than the first preset threshold value, and the negative health high-risk result is a negative mental health high-risk period, the user's mental health state is moderate negative psychology. If the index value of the susceptibility index is greater than or equal to the first preset threshold value, and the negative health high-risk result is a negative mental health high-risk period, the user's mental health state is severe negative psychology.

[0055] As a feasible implementation method, when a user is identified to have mild negative psychology, a clinical physician or psychological counselor may be prompted to conduct normal intervention treatment. When a user is identified to have moderate negative psychology, a clinical physician or psychological counselor may be prompted to conduct urgent intervention treatment. When a user is identified to have severe negative psychology, a clinical physician or psychological counselor may be prompted to conduct emergency intervention treatment. This is conducive to clinical physicians or psychological counselors making auxiliary treatment decisions based on the accurately identified user's mental health status, helping users with abnormal mental health to conduct targeted treatment more quickly, improving the user's mental health status, and helping them return to normal life.

[0056] In addition, the present application also provides a device for identifying the mental health status of a group of people, such as Figure 2 As shown, the device 200 for identifying the mental health status of a group of people specifically includes:

[0057] At least one processor 201, and a memory 202 in communication with the at least one processor 201; wherein the memory 202 stores instructions executable by the at least one processor 201, so that the at least one processor 201 can execute:

[0058] By showing preset physiological stimulation to the user, facial image data and brain function image data of the user are collected;

[0059] Differentially annotate facial image data and brain function image data to obtain susceptibility indicators;

[0060] By showing the user a preset scene interaction video, the user's eye gaze trajectory and head posture are collected and negative bias analysis is performed to obtain negative attention bias results;

[0061] Perform semantic analysis and sentiment analysis on the pre-collected user voice content to obtain the results of negative language effects; and determine negative health high-risk results based on the negative attention bias results and the negative language effect results;

[0062] Determine the user's mental health status based on susceptibility indicators and negative health risk outcomes.

[0063] The embodiments of the present application provide a method and device for identifying the mental health status of a group of people. By obtaining susceptibility indicators and negative health high-risk results, it is possible to quickly and accurately judge the mental health status of a group of people, so as to determine the mental health status and mental state of the group of people, and provide cognitive behavioral therapy and targeted treatment decisions for clinicians, which is conducive to timely discovering people with abnormal mental health, and using the identified mental health status of the group of people as an important reference indicator for clinicians to conduct auxiliary treatment.

[0064] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0065] The above describes specific embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0066] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the embodiments of the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the embodiments of the present application should be included in the scope of the claims of the present application.

Claims

1. A device for identifying the mental health status of a group of people, characterized in that: The device comprises: at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the device for identifying the mental health status of a group of people to perform: By showing a preset physiological stimulus to a user, facial image data and brain function image data of the user are collected; Performing difference annotation on the facial image data and the brain function image data to obtain a susceptibility index; By showing the preset scene interaction video to the user, collecting the user's eye gaze trajectory and head posture and performing negative bias analysis, a negative attention bias result is obtained; Performing semantic analysis and sentiment analysis on the pre-collected user voice content to obtain language negative effect results; and determining negative health high-risk results based on the negative attention bias results and the language negative effect results; Determining the user's mental health status according to the susceptibility index and the negative health high-risk result; Among them, by showing the preset scene interaction video to the user, collecting the user's eye movement gaze trajectory and head posture and performing negative bias analysis to obtain negative attention bias results, the device can specifically perform: By showing the preset scene interaction video to the user, collecting the user's eye movement gaze trajectory during the display process, and obtaining a trajectory distribution curve; wherein the content of the preset scene interaction video includes calm tone content and negative tone content; The head posture of the user is collected by a head camera, and a head cube frame model is established according to the orientation position of the head posture; According to the head cube framework model, performing orientation framework analysis on the head posture of the user to obtain the head orientation of the user; If the distribution ratio of the negative tone content in the trajectory distribution curve is greater than the distribution ratio of the calm tone content, and the position of the head direction is at the position where the negative tone content is located, then the user has a negative attention bias; If the distribution ratio of the negative tone content in the trajectory distribution curve is less than or equal to the distribution ratio of the calm tone content, or the position of the head direction is not at the position where the negative tone content is located, then the user does not have a negative attention bias; The negative attention bias result includes the existence of negative attention bias and the absence of negative attention bias.

2. The device according to claim 1, characterized in that By displaying preset physiological stimulation to the user and collecting the user's facial image data and brain function image data, the following can be performed: The facial image data is obtained by collecting the facial expression, the corresponding heart rate and the heart rate variability of the user through a facial collection device; wherein the facial collection device is obtained by connecting a facial collection camera and an image analysis device; The brain function fMRI image and the time series of the brain function fMRI image are collected by a brain function collection device to obtain the brain function image data; wherein the brain function collection device is obtained by connecting a brain function fMRI imaging device and the image analysis device; The preset physiological stimulation includes any one or more of the following: visual stimulation, olfactory stimulation and taste stimulation.

3. The device according to claim 1, characterized in that The facial image data is differentially annotated, and specifically, the following steps can be performed: Before showing the preset physiological stimulation process to the user, collecting facial image data of the user, and analyzing the facial image data to obtain a pre-stimulation state time series; In the process of showing the preset physiological stimulation to the user, re-collecting the facial image data of the user, re-analyzing the facial image data, and obtaining a state time series during the stimulation; Establishing a normal distribution curve according to the time series difference between the pre-stimulation state time series and the during-stimulation state time series; If the difference value on the normal distribution curve is less than 30% of the normal distribution probability, the corresponding stimulus in the preset physiological stimulus is marked with a difference to obtain a facial stimulation index.

4. The device according to claim 3, characterized in that The facial image data and the brain function image data are differentially annotated to obtain a susceptibility index, which can be specifically performed as follows: Monitoring the activation time course of the lateral orbitofrontal cortex region in the brain function image data to obtain the activation time length of each stimulus in the preset physiological stimulus and the activation degree of each stimulus in the lateral orbitofrontal cortex region; If the activation time length and the activation degree of each stimulus in the preset physiological stimulation are different from the normal activation time course, the difference is marked to obtain a brain stimulation index; The facial stimulation index after the difference of the facial image data is annotated and the brain stimulation index are aggregated and unified to obtain the susceptibility index, and the susceptibility index is displayed on the screen; wherein the susceptibility index is used to measure the mental health state of the user after being subjected to the preset physiological stimulation.

5. The device according to claim 1, characterized in that Before performing semantic analysis and sentiment analysis on the pre-collected user voice content to obtain the language negative effect result, the device can also perform: Collecting the user's voice content and converting the voice content into text content; Performing semantic analysis on the text content to obtain the abstract level, attribution method, and positive and negative valence of the text semantics; Establishing a semantic model of the text content according to the abstraction level and the attribution method; A sentiment model of the text content is established according to the positive and negative valences.

6. The device according to claim 5, characterized in that Perform semantic analysis and sentiment analysis on the pre-collected user voice content to obtain the results of the negative effects of language. Specifically, the following can be performed: According to the language key words and high-frequency vocabulary library in the semantic model, a semantic cause analysis is performed on the text content of the user to determine the cause analysis level; wherein the cause analysis level includes an interpersonal level and an abstract level; Determining the attribution type of the text content according to the keyword path network of the attribution method in the semantic model; wherein the attribution type includes internal attribution and external attribution; According to the emotional vocabulary library in the emotional model, similarity analysis is performed on the emotional vocabulary set in the text content to obtain the emotional valence of the text content; wherein the emotional valence includes positive emotion and negative emotion; If the cause analysis level is the abstract level, the attribution type is internal attribution, and the emotional valence is negative emotion, then the negative language effect result is that there is a negative language effect; If the causal analysis level is the interpersonal level, or the attribution type is external attribution, or the emotional valence is positive emotion, then the negative language effect result is that there is no negative language effect.

7. The device according to claim 1, characterized in that According to the negative attention bias results and the negative language effect results, negative health high-risk results are determined, and specifically the following can be performed: If the negative attention bias result is the existence of negative attention bias, and the language negative effect result is the existence of language negative effect, then the negative health high-risk result is the negative mental health high-risk period; If the negative attention bias result is that there is no negative attention bias, or the negative language effect result is that there is no negative language effect, then the negative health high-risk result is a negative mental health low-risk period.

8. The device according to claim 1, characterized in that Determining the user's mental health status according to the susceptibility index and the negative health high-risk result may specifically be performed: If the index value of the susceptibility index is less than the first preset threshold value, and the negative health high-risk result is a negative mental health low-risk period, then the user's mental health state is mental health; If the index value of the susceptibility index is greater than or equal to the first preset threshold, and the negative health high-risk result is a negative mental health low-risk period, then the user's mental health state is mildly negative; If the index value of the susceptibility index is less than the first preset threshold value, and the negative health high-risk result is a negative mental health high-risk period, then the user's mental health state is moderately negative; If the index value of the susceptibility index is greater than or equal to the first preset threshold, and the negative health high-risk result is a negative mental health high-risk period, then the user's mental health state is severe negative psychology.

Images