A mental health evaluation method and system based on eye movement characteristics

Abstract

This invention discloses a method and system for mental health assessment based on eye movement features, belonging to the field of mental health assessment technology. It includes a data collection module, a calculation module, a threshold module, and a judgment module. When subjects read natural language passages containing elements triggering childhood trauma, the eye-tracking sampler in the data collection module collects eye movement data from a single area of ​​interest. The calculation module then substitutes the corresponding standard scores into the emotional arousal formula, emotional change formula, emotional matching formula, and cognitive load formula to calculate the subject's emotional arousal index, emotional change gradient, emotional matching score, and cognitive load index in that single area of ​​interest. The calculation module also substitutes the subject's average emotional arousal index, average emotional change gradient, average emotional matching score, and average cognitive load index into the trauma risk formula to calculate the trauma risk index.

Description

Technical Field

[0001] This invention relates to the field of mental health assessment technology, and in particular to a mental health assessment method and system based on eye movement characteristics. Background Technology

[0002] The core purpose of conducting a mental health assessment for childhood trauma is to understand how a person's current emotions, behaviors, and relationship patterns are deeply influenced by past traumatic experiences, thereby providing a key pathway to healing. This assessment scientifically identifies repressed or forgotten early traumas, connecting many seemingly inexplicable psychological distresses to their roots, and helping individuals break the vicious cycle of trauma.

[0003] The following problems often exist in existing mental health assessments: (1) Most of them focus on mechanical questioning and often neglect the collection of eye movement features. If the subjects are unwilling to answer mechanical questions, it will lead to incorrect diagnosis. As a subconscious eye movement, eye movement features generally cannot be forcibly changed. (2) Although there are corresponding methods for collecting and analyzing eye movement features in existing technologies, most of them are limited to a single eye movement feature and cannot comprehensively score multiple eye movement features, thereby reducing the randomness of a single eye movement feature. (3) When comprehensively scoring eye movement features, it is very likely that the subjects will be in a state of mental fatigue. However, existing analysis methods often only focus on the value of a single eye movement feature, and cannot quickly and effectively detect the mental fatigue of the subjects, which may cause the final score to deviate. Summary of the Invention

[0004] The purpose of this invention is to provide a mental health evaluation method and system based on eye movement features, aiming to solve the technical problems existing in the prior art, such as how to collect and quickly analyze eye movement features, how to comprehensively evaluate multiple eye movement features, and how to quickly and effectively detect mental fatigue of subjects during the comprehensive scoring process.

[0005] To address the aforementioned technical problems, the present invention adopts the following technical solution: a mental health assessment method based on eye movement features, comprising the following steps: Step S1: Present natural language segments containing elements that trigger childhood trauma to the subjects, and divide each natural language segment containing elements that trigger childhood trauma into five interest areas according to five elements: time, person, frequency, event, and feeling. Step S2: While the participants read natural language passages containing elements that trigger childhood trauma, the eye-tracking sampler in the data collection module collects eye-tracking data from a single area of ​​interest. Eye-tracking data for a single area of ​​interest Includes: average fixation duration Average number of fixations Average number of revisits Average number of eye twitches Average saccade amplitude Average pupil diameter Maximum pupil diameter Average blink rate and average blink time ; Step S3: Before step S1, in a calm state, the participants were asked to read natural language passages that did not contain triggers for childhood trauma. Subsequently, the eye-tracking sampler in the data collection module collected baseline eye-tracking data from the participants when they were reading the natural language passages that did not contain triggers for childhood trauma. ; Step S4: Subsequently, the calculation module will receive all eye-tracking data from the data collection module, and will also use a standardized formula to process eye-tracking data from individual interest areas. Standardization is performed to obtain eye-tracking data for a single area of ​​interest. The corresponding standard score ; Step S5: The calculation module then calculates the standard score. Substituting these values ​​into the emotional arousal formula, emotional change formula, emotional matching formula, and cognitive load formula, respectively, the emotional arousal index of the subjects in a single area of ​​interest was calculated. Emotional change gradient Emotional Matching Score and cognitive load index ; Step S6: The module then calculates the average emotional impact index of the participants. Average mood change gradient Average sentiment matching score and average cognitive load index Substituting these values ​​into the trauma risk formula, the trauma risk index is calculated. ; Step S7: The module then calculates the emotional engagement index of the subject in a single area of ​​interest. Emotional change gradient Emotional Matching Score Cognitive Load Index And the final trauma risk index The data is transmitted to the judgment module (4) to determine the mental health status of the test subject.

[0006] Preferably, a single region of interest contains one or more regions of interest; eye-tracking data for a single region of interest. Benchmark eye-tracking data Standard scores In Each of these indicates a data type number collected by the eye-tracking sampler. The value ranges from 1 to 9; a single area of ​​interest eye-tracking data It is obtained by averaging eye movement data from multiple regions of interest within a single region of interest; average fixation duration The average fixation time within a single area of ​​interest; the average number of fixations. The number of fixation events within a single area of ​​interest; the average number of regressions. The number of fixations from the later text to the earlier text within a single area of ​​interest; average number of saccades. The number of saccades within a single area of ​​interest; average saccade amplitude. The average angular span of a single saccade within a single area of ​​interest; average pupil diameter. The average pupil diameter within a single region of interest; maximum pupil diameter. The maximum pupil diameter within a single region of interest; average blink rate. The number of blink events within a single area of ​​interest; average blink duration. The average duration of a blink event within a single area of ​​interest; average fixation duration. Average number of fixations Average number of revisits Average number of eye twitches Average saccade amplitude Average pupil diameter Maximum pupil diameter Average blink rate and average blink time All data were obtained using an eye-tracking sampling device.

[0007] Preferably, interest zones are not set for natural language segments that do not contain elements that trigger childhood trauma; baseline eye-tracking data Compared with eye-tracking data from a single area of ​​interest It is a one-to-one correspondence; when the total number of subjects is less than 50, the baseline eye-tracking data... It is the average eye movement data of a single subject when faced with multiple natural language segments that do not contain elements that trigger childhood trauma; when the total number of subjects is greater than or equal to 50, the baseline eye movement data is... It consists of average eye movement data of subjects of the same age and gender when faced with a single natural language passage that does not contain elements that trigger childhood trauma.

[0008] Preferably, let As a benchmark eye-tracking data The corresponding standard deviation; the calculation module will use the baseline eye-tracking data. Eye-tracking data for a single area of ​​interest and standard deviation Substituting these values ​​into the standardized formula, we can obtain eye-tracking data for a single area of ​​interest. The corresponding standard score Then the standard score for: .

[0009] Preferably, the calculation module will calculate the average fixation time. Corresponding standard score Average number of fixations Corresponding standard score Average number of revisits Corresponding standard score Average number of eye twitches Corresponding standard score and average pupil diameter Corresponding standard score Substituting these values ​​into the emotional engagement formula, we can obtain the emotional engagement index. The formula for triggering emotions is as follows: ; In the formula: The pupil modulation coefficient ranges from 0.1 to 1.0.

[0010] Preferably, the calculation module calculates the average pupil diameter. Corresponding standard score and maximum pupil diameter Corresponding standard score Substituting these values ​​into the formula for emotion change, we can obtain the gradient of emotion change. The formula for mood changes is as follows: .

[0011] Preferably, the threshold module sets a separate type of eye-tracking data for interest areas based on previous training data and data recommended by psychological experts. The corresponding eye movement data threshold The calculation module then calculates the average fixation time. Corresponding standard score and the corresponding threshold Average number of fixations Corresponding standard score and the corresponding threshold Average number of revisits Corresponding standard score and the corresponding threshold Average number of eye twitches Corresponding standard score and the corresponding threshold Average saccade amplitude Corresponding standard score and the corresponding threshold Substitute these values ​​into the emotion matching formula to obtain the emotion matching score. The emotion matching formula is as follows: ; In the formula: This represents the eye movement feature vector of the subject. Represents the threshold expectation vector, This indicates finding the norm of a vector.

[0012] Preferably, the calculation module will calculate the average fixation time. Corresponding standard score Average number of fixations Corresponding standard score Average number of revisits Corresponding standard score Average number of eye twitches Corresponding standard score Average saccade amplitude Corresponding standard score Average blink rate Corresponding standard score and average blink time Corresponding standard score Substituting these values ​​into the cognitive load formula, we can obtain the cognitive load index. The formula for cognitive load is as follows: ; In the formula: This indicates the blink adjustment coefficient. The value range is between 0.05 and 0.5; when the cognitive load formula... When the term is less than or equal to 0, it indicates that the subject experiences relatively little psychological change within that area of ​​interest. In this case, if the cognitive load formula... A higher cognitive load index indicates that the subject is fatigued; therefore, a higher cognitive load index indicates higher fatigue. If the value is less than or equal to -5, the subject needs to adjust their state and the test needs to be repeated.

[0013] Preferably, the calculation module calculates based on the emotional arousal index corresponding to the five interest areas. Emotional change gradient Emotional Matching Score Cognitive Load Index To obtain the average emotional impact index of the subjects. Average mood change gradient Average sentiment matching score and average cognitive load index Then, the average emotional impact index was calculated. Average mood change gradient Average sentiment matching score and average cognitive load index Substituting these values ​​into the trauma risk formula, the trauma risk index is calculated. The trauma risk formula is as follows: ; In the formula: This indicates the weight of emotional triggers. The value ranges from 0.5 to 1.5. Indicates the weight of emotional changes. The value ranges from 0.5 to 1.5. Indicates the weight of sentiment matching. The value ranges from 0.5 to 1.5. Indicates cognitive load weighting. The value range is between 0.5 and 1.5; , , , The specific value is determined based on the age group, gender, and family economic status of the subjects; When the judgment module detects the subject's trauma risk index When the value is greater than or equal to 0.55, the participant will be marked as a high-risk individual for mental health issues and will be connected to the manual assessment unit for in-depth mental health evaluation; when the judgment module detects the participant's trauma risk index... When the value is less than 0.55, the participants are marked as low-risk individuals for mental health issues. These individuals are then connected to a question-and-answer unit for automated AI-powered question-and-answer sessions to understand their childhood trauma. The assessment module will then filter out trauma risk indices. Participants with a trauma risk index between 0.45 and 0.65 simultaneously sent instructions to the host computer of the data collection module, causing the trauma risk index to... Data from subjects with scores between 0.45 and 0.65 were sent to the threshold module.

[0014] This invention also provides a mental health assessment system based on eye-tracking features, including a data collection module, a calculation module, a threshold module, and a judgment module. The data collection module is unidirectionally connected to the calculation module and the threshold module, respectively. The threshold module is unidirectionally connected to the calculation module, the calculation module is unidirectionally connected to the judgment module, and the judgment module is unidirectionally connected to the data collection module. The data collection module includes a display screen, a host computer, and an eye-tracking sampler. The display screen is used to present natural language segments stored on the host computer. The host computer is used to store natural language segments and all data collected by the eye-tracking sampler. The eye-tracking sampler is used to collect eye-tracking data. The host computer is unidirectionally connected to the calculation module and the threshold module, respectively. The judgment module includes a human assessment unit and a question-and-answer unit. The human assessment unit is used for human assessment of individuals at high risk of mental health issues. The question-and-answer unit is used for AI-based question-and-answer for individuals at low risk of mental health issues.

[0015] The beneficial effects of this invention compared with the prior art are: (1) The retrieval module outputs the statistical data to the statistics module, and then the network security logs of different servers are analyzed by the unique scoring unit in the statistics module. Perform asset vulnerability scoring External threat score and internal abnormal activity score The calculation then uses a unique scoring unit to determine the average asset vulnerability score for all servers. Average external threat score and average internal abnormal activity score The output is sent to the comprehensive scoring unit, which then calculates the comprehensive safety score based on the comprehensive evaluation formula. (2) The first adjustment unit within the adjustment module is based on the average asset vulnerability score. Calculate the vulnerability weight The second adjustment unit is based on the average external threat score. Calculate the threat weight The third adjustment unit is based on the average internal abnormal activity score. Calculate the outlier weights Then the adjustment module outputs the three weights to the comprehensive scoring unit, thereby realizing the dynamic adjustment of the weights. (3) In the asset vulnerability scoring formula This represents the vulnerability age decay coefficient, and is also part of the external threat scoring formula. This represents the threat age decay coefficient, when and As the value of increases, the asset vulnerability score will decrease due to past events. External threat score The impact of recent events on asset vulnerability scores. External threat score The impact of this, and the average deviation value in the internal abnormal activity scoring formula. It is calculated based on data within the current minute and is therefore highly timely. Attached Figure Description

[0016] Figure 1 This is a flowchart of the mental health assessment method based on eye movement features according to the present invention.

[0017] Figure 2 This is a schematic diagram of the architecture of the mental health assessment system based on eye movement features of this invention.

[0018] Figure 3 This is a schematic diagram of the structural principle of the mental health assessment system based on eye movement features of the present invention.

[0019] In the diagram: 1-Data collection module; 2-Calculation module; 3-Threshold module; 4-Judgment module; 11-Display screen; 12-Main unit; 13-Eye tracker; 41-Human evaluation unit; 42-Question and answer unit. Detailed Implementation

[0020] The technical solution of the present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0021] The accompanying drawings are for illustrative purposes only and are schematic diagrams, not actual pictures. They should not be construed as limiting the invention. To better illustrate the embodiments of the invention, some parts in the drawings may be omitted, enlarged, or reduced, and do not represent the actual product dimensions. It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0022] Figures 1 to 3 This is a preferred embodiment of the present invention.

[0023] Figure 1 A flowchart of the eye-tracking feature-based mental health assessment method of the present invention is provided, including the following steps: Step S1: Present natural language segments containing elements that trigger childhood trauma to the subjects, and divide each natural language segment containing elements that trigger childhood trauma into five interest areas according to five elements: time, person, frequency, event, and feeling. Step S2: While the subjects read natural language passages containing elements that trigger childhood trauma, the eye-tracking sampler 13 in the data collection module 1 collects eye-tracking data from a single area of ​​interest of the subjects. Eye-tracking data for a single area of ​​interest Includes: average fixation duration Average number of fixations Average number of revisits Average number of eye twitches Average saccade amplitude Average pupil diameter Maximum pupil diameter Average blink rate and average blink time ; Step S3: Before step S1, the participants, in a calm state, are asked to read natural language passages that do not contain triggers for childhood trauma. Subsequently, the eye-tracking sampler 13 in the data collection module 1 collects baseline eye-tracking data of the participants when they are faced with natural language passages that do not contain triggers for childhood trauma. ; Step S4: Subsequently, the calculation module 2 will receive all eye-tracking data from the data collection module 1, and will also use a standardized formula to process eye-tracking data from individual interest areas. Standardization is performed to obtain eye-tracking data for a single area of ​​interest. The corresponding standard score ; Step S5: Then, the calculation module 2 will calculate the standard score. Substituting these values ​​into the emotional arousal formula, emotional change formula, emotional matching formula, and cognitive load formula, respectively, the emotional arousal index of the subjects in a single area of ​​interest was calculated. Emotional change gradient Emotional Matching Score and cognitive load index ; Step S6: Then calculate the average emotional impact index of the subjects in module 2. Average mood change gradient Average sentiment matching score and average cognitive load index Substituting these values ​​into the trauma risk formula, the trauma risk index is calculated. ; Step S7: Then, module 2 calculates the emotional arousal index of the subject in a single area of ​​interest. Emotional change gradient Emotional Matching Score Cognitive Load Index And the final trauma risk index The data is transmitted to the judgment module 4 to determine the mental health status of the test subject.

[0024] The working principle of this invention is as follows: First, natural language passages containing elements triggering childhood trauma are presented to the subjects. These passages are then divided into five interest zones based on five elements: time, person, frequency, event, and feeling. Subsequently, while the subjects read these passages, the eye-tracking sampler 13 in the data collection module 1 collects eye-tracking data for each individual interest zone. Eye-tracking data for a single area of ​​interest Includes: average fixation duration Average number of fixations Average number of revisits Average number of eye twitches Average saccade amplitude Average pupil diameter Maximum pupil diameter Average blink rate and average blink time Simultaneously, when the participants were in a calm state, they were asked to read natural language passages that did not contain triggers for childhood trauma. Subsequently, the eye-tracking sampler 13 in the data collection module 1 collected baseline eye-tracking data of the participants when they were reading natural language passages that did not contain triggers for childhood trauma. Subsequently, the calculation module 2 receives all eye-tracking data from the data collection module 1, and also processes the eye-tracking data of a single area of ​​interest using a standardized formula. Standardization is performed to obtain eye-tracking data for a single area of ​​interest. The corresponding standard score Then, module 2 calculates the standard scores. Substituting these values ​​into the emotional arousal formula, emotional change formula, emotional matching formula, and cognitive load formula, respectively, the emotional arousal index of the subjects in a single area of ​​interest was calculated. Emotional change gradient Emotional Matching Score and cognitive load index ; then module 2 calculates the average emotional impact index of the subjects. Average mood change gradient Average sentiment matching score and average cognitive load index Substituting these values ​​into the trauma risk formula, the trauma risk index is calculated. ; then module 2 calculates the emotional engagement index of the subjects in a single area of ​​interest. Emotional change gradient Emotional Matching Score Cognitive Load Index And the final trauma risk index The data is transmitted to the judgment module 4 to determine the mental health status of the test subject.

[0025] Furthermore, a single region of interest contains one or more regions of interest; eye-tracking data for a single region of interest. Benchmark eye-tracking data Standard scores In Each of these indicates a data type number collected by eye-tracking sampler 13. The value ranges from 1 to 9; a single area of ​​interest eye-tracking data It is obtained by averaging eye movement data from multiple regions of interest within a single region of interest; average fixation duration The average fixation time within a single area of ​​interest; the average number of fixations. The number of fixation events within a single area of ​​interest; the average number of regressions. The number of fixations from the later text to the earlier text within a single area of ​​interest; average number of saccades. The number of saccades within a single area of ​​interest; average saccade amplitude. The average angular span of a single saccade within a single area of ​​interest; average pupil diameter. The average pupil diameter within a single region of interest; maximum pupil diameter. The maximum pupil diameter within a single region of interest; average blink rate. The number of blink events within a single area of ​​interest; average blink duration. The average duration of a blink event within a single area of ​​interest; average fixation duration. Average number of fixations Average number of revisits Average number of eye twitches Average saccade amplitude Average pupil diameter Maximum pupil diameter Average blink rate and average blink time All data were obtained using eye-tracking sampler 13.

[0026] Furthermore, no interest zones are set for natural language segments that do not contain elements that trigger childhood trauma; baseline eye-tracking data Compared with eye-tracking data from a single area of ​​interest It is a one-to-one correspondence; when the total number of subjects is less than 50, the baseline eye-tracking data... It is the average eye movement data of a single subject when faced with multiple natural language segments that do not contain elements that trigger childhood trauma; when the total number of subjects is greater than or equal to 50, the baseline eye movement data is... It consists of average eye movement data of subjects of the same age and gender when faced with a single natural language passage that does not contain elements that trigger childhood trauma.

[0027] Furthermore, set As a benchmark eye-tracking data The corresponding standard deviation; Calculation module 2 will use the baseline eye-tracking data Eye-tracking data for a single area of ​​interest and standard deviation Substituting these values ​​into the standardized formula, we can obtain eye-tracking data for a single area of ​​interest. The corresponding standard score Then the standard score for: .

[0028] Furthermore, the calculation module 2 will calculate the average fixation time. Corresponding standard score Average number of fixations Corresponding standard score Average number of revisits Corresponding standard score Average number of eye twitches Corresponding standard score and average pupil diameter Corresponding standard score Substituting these values ​​into the emotional engagement formula, we can obtain the emotional engagement index. The formula for triggering emotions is as follows: ; In the formula: The pupil modulation coefficient ranges from 0.1 to 1.0.

[0029] Furthermore, the calculation module 2 calculates the average pupil diameter. Corresponding standard score and maximum pupil diameter Corresponding standard score Substituting these values ​​into the formula for emotion change, we can obtain the gradient of emotion change. The formula for mood changes is as follows: .

[0030] Furthermore, threshold module 3 sets a separate type of eye-tracking data for interest areas based on previous training data and recommendations from psychological experts. The corresponding eye movement data threshold ;Then the calculation module 2 will calculate the average fixation time Corresponding standard score and the corresponding threshold Average number of fixations Corresponding standard score and the corresponding threshold Average number of revisits Corresponding standard score and the corresponding threshold Average number of eye twitches Corresponding standard score and the corresponding threshold Average saccade amplitude Corresponding standard score and the corresponding threshold Substitute these values ​​into the emotion matching formula to obtain the emotion matching score. The emotion matching formula is as follows: ; In the formula: This represents the eye movement feature vector of the subject. Represents the threshold expectation vector, This indicates finding the norm of a vector.

[0031] Furthermore, the calculation module 2 will calculate the average fixation time. Corresponding standard score Average number of fixations Corresponding standard score Average number of revisits Corresponding standard score Average number of eye twitches Corresponding standard score Average saccade amplitude Corresponding standard score Average blink rate Corresponding standard score and average blink time Corresponding standard score Substituting these values ​​into the cognitive load formula, we can obtain the cognitive load index. The formula for cognitive load is as follows: ; In the formula: This indicates the blink adjustment coefficient. The value range is between 0.05 and 0.5; when the cognitive load formula... When the term is less than or equal to 0, it indicates that the subject experiences relatively little psychological change within that area of ​​interest. In this case, if the cognitive load formula... A higher cognitive load index indicates that the subject is fatigued; therefore, a higher cognitive load index indicates higher fatigue. If the value is less than or equal to -5, the subject needs to adjust their state and the test needs to be repeated.

[0032] Furthermore, the calculation module 2 calculates the emotional arousal index corresponding to the five interest areas. Emotional change gradient Emotional Matching Score Cognitive Load Index To obtain the average emotional impact index of the subjects. Average mood change gradient Average sentiment matching score and average cognitive load index Then, the average emotional impact index was calculated. Average mood change gradient Average sentiment matching score and average cognitive load index Substituting these values ​​into the trauma risk formula, the trauma risk index is calculated. The trauma risk formula is as follows: ; In the formula: This indicates the weight of emotional triggers. The value ranges from 0.5 to 1.5. Indicates the weight of emotional changes. The value ranges from 0.5 to 1.5. Indicates the weight of sentiment matching. The value ranges from 0.5 to 1.5. Indicates cognitive load weighting. The value range is between 0.5 and 1.5; , , , The specific values ​​are determined based on the participant's age, gender, and family economic situation. For example, participants aged 35 to 40, male, and from a middle-class family are generally experiencing a period of family harmony and are more mature and resilient. They are less sensitive to childhood trauma at this age. When childhood trauma is recalled, they may be suddenly moved, but the overall emotional change is not intense, and they will not experience severe depression afterward. Therefore... The specific value will tend to be larger. , , The specific value will be relatively small to prevent the emotional trigger index from being affected. Diluted, leading to a lower final trauma risk index Low The value is 1.4. The value is 0.8. The value is 1. The value is 0.6; When the judgment module 4 detects the subject's trauma risk index When the value is greater than or equal to 0.55, the participant will be marked as a high-risk individual for mental health issues and will be connected to the manual assessment unit 41 for in-depth mental health assessment; when the judgment module 4 detects the participant's trauma risk index... When the value is less than 0.55, the participants will be marked as low-risk individuals for mental health issues. These individuals will then be connected to question-and-answer unit 42 for automated question-and-answer sessions using artificial intelligence to understand their childhood trauma. Judgment module 4 will then filter out the trauma risk index. Subjects with a trauma risk index between 0.45 and 0.65 simultaneously sent instructions to the host 12 of the data collection module 1, causing the trauma risk index to... Data from subjects with values ​​between 0.45 and 0.65 are sent to threshold module 3.

[0033] For a mental health assessment system based on eye movement features, such as Figure 2 and Figure 3 As shown, the system includes a data collection module 1, a calculation module 2, a threshold module 3, and a judgment module 4. The data collection module 1 is unidirectionally connected to the calculation module 2 and the threshold module 3, respectively. The threshold module 3 is unidirectionally connected to the calculation module 2. The calculation module 2 is unidirectionally connected to the judgment module 4. The judgment module 4 is unidirectionally connected to the data collection module 1. The data collection module 1 includes a display screen 11, a host computer 12, and an eye-tracking sampler 13. The display screen 11 is used to display natural language segments stored in the host computer 12. The host computer 12 is used to store natural language segments and all data collected by the eye-tracking sampler 13. The eye-tracking sampler 13 is used to collect eye-tracking data. The host computer 12 is unidirectionally connected to the calculation module 2 and the threshold module 3, respectively. The judgment module 4 includes a human assessment unit 41 and a question-and-answer unit 42. The human assessment unit 41 is used for human assessment of individuals at high risk of mental health issues. The question-and-answer unit 42 is used for AI-based question-and-answer for individuals at low risk of mental health issues.

[0034] This invention is not limited to the specific embodiments described above. Any modifications made by those skilled in the art based on the above concept without creative effort are within the protection scope of this invention.

Claims

1. A mental health assessment method based on eye-tracking features, characterized in that, Includes the following steps: Step S1: Present natural language segments containing elements that trigger childhood trauma to the subjects, and divide each natural language segment containing elements that trigger childhood trauma into five interest areas according to five elements: time, person, frequency, event, and feeling. Step S2: While the subjects read natural language passages containing elements that trigger childhood trauma, the eye-tracking sampler (13) in the data collection module (1) collects eye-tracking data of a single area of ​​interest from the subjects. Eye-tracking data for a single area of ​​interest Includes: average fixation duration Average number of fixations Average number of revisits Average number of eye twitches Average saccade amplitude Average pupil diameter Maximum pupil diameter Average blink rate and average blink time ; Step S3: Before step S1, the subjects, in a calm state, are asked to read natural language passages that do not contain elements that trigger childhood trauma. Subsequently, the eye-tracking sampler (13) in the data collection module (1) will collect baseline eye-tracking data of the subjects when they are faced with natural language passages that do not contain elements that trigger childhood trauma. ; Step S4: Subsequently, the calculation module (2) will receive all eye-tracking data from the data collection module (1), and will also process the eye-tracking data of a single area of ​​interest using a standardized formula. Standardization is performed to obtain eye-tracking data for a single area of ​​interest. The corresponding standard score ; Step S5: Then the calculation module (2) will calculate the standard score. Substituting these values ​​into the emotional arousal formula, emotional change formula, emotional matching formula, and cognitive load formula, respectively, the emotional arousal index of the subjects in a single area of ​​interest was calculated. Emotional change gradient Emotional Matching Score and cognitive load index ; Step S6: Then calculate the average emotional impact index of the subjects in module (2). Average mood change gradient Average sentiment matching score and average cognitive load index Substituting these values ​​into the trauma risk formula, the trauma risk index is calculated. ; Step S7: Then calculate module (2) the emotional arousal index of the subject in a single area of ​​interest. Emotional change gradient Emotional Matching Score Cognitive Load Index And the final trauma risk index The data is transmitted to the judgment module (4) to determine the mental health status of the test subject.

2. The mental health assessment method based on eye movement features as described in claim 1, characterized in that: A single area of ​​interest may contain one or more areas of interest; eye-tracking data for a single area of ​​interest. Benchmark eye-tracking data Standard score In All of these represent the type number of the data collected by the eye-tracking sampler (13). The value ranges from 1 to 9; a single area of ​​interest eye-tracking data It is obtained by averaging eye movement data from multiple regions of interest within a single region of interest; average fixation duration The average fixation time within a single area of ​​interest; the average number of fixations. The number of gaze events within a single area of ​​interest; Average number of replays The number of fixations from the later text to the earlier text within a single area of ​​interest; average number of saccades. The number of saccades within a single area of ​​interest; average saccade amplitude. The average angular span of a single saccade within a single area of ​​interest; average pupil diameter. The average pupil diameter within a single region of interest; maximum pupil diameter. The maximum pupil diameter within a single region of interest; average blink rate. The number of blink events within a single area of ​​interest; average blink duration. The average duration of a blink event within a single area of ​​interest; average fixation duration. Average number of fixations Average number of revisits Average number of eye twitches Average saccade amplitude Average pupil diameter Maximum pupil diameter Average blink rate and average blink time All were obtained by eye-tracking sampling device (13).

3. The mental health assessment method based on eye movement features as described in claim 2, characterized in that: Natural language segments that do not contain elements that trigger childhood trauma are not assigned zones of interest; baseline eye-tracking data Compared with eye-tracking data from a single area of ​​interest It is a one-to-one correspondence; when the total number of subjects is less than 50, the baseline eye-tracking data... It is the average eye movement data of a single subject when faced with multiple natural language segments that do not contain elements that trigger childhood trauma; when the total number of subjects is greater than or equal to 50, the baseline eye movement data is... It consists of average eye movement data of subjects of the same age and gender when faced with a single natural language passage that does not contain elements that trigger childhood trauma.

4. The mental health assessment method based on eye movement features as described in claim 3, characterized in that: set up As a benchmark eye-tracking data The corresponding standard deviation; the calculation module (2) calculates the baseline eye-tracking data. Eye-tracking data for a single area of ​​interest and standard deviation Substituting these values ​​into the standardized formula, we can obtain eye-tracking data for a single area of ​​interest. The corresponding standard score Then the standard score for: 。 5. The mental health assessment method based on eye movement features as described in claim 4, characterized in that: The calculation module (2) will calculate the average fixation time. Corresponding standard score Average number of fixations Corresponding standard score Average number of revisits Corresponding standard score Average number of eye twitches Corresponding standard score and average pupil diameter Corresponding standard score Substituting these values ​​into the emotional engagement formula, we can obtain the emotional engagement index. The formula for triggering emotions is as follows: ； In the formula: The pupil modulation coefficient ranges from 0.1 to 1.

0.

6. The mental health assessment method based on eye movement features as described in claim 5, characterized in that: The calculation module (2) calculates the average pupil diameter. Corresponding standard score and maximum pupil diameter Corresponding standard score Substituting these values ​​into the formula for emotion change, we can obtain the gradient of emotion change. The formula for mood changes is as follows: 。 7. The mental health assessment method based on eye movement features as described in claim 6, characterized in that: The threshold module (3) sets a separate eye-tracking data set for interest areas based on previous training data and data recommended by psychological experts. The corresponding eye movement data threshold The calculation module (2) then calculates the average fixation time. Corresponding standard score and the corresponding threshold Average number of fixations Corresponding standard score and the corresponding threshold Average number of revisits Corresponding standard score and the corresponding threshold Average number of eye twitches Corresponding standard score and the corresponding threshold Average saccade amplitude Corresponding standard score and the corresponding threshold Substitute these values ​​into the emotion matching formula to obtain the emotion matching score. The emotion matching formula is as follows: ； In the formula: This represents the eye movement feature vector of the subject. Represents the threshold expectation vector, This indicates finding the norm of a vector.

8. A mental health assessment method based on eye movement features as described in claim 7, characterized in that: The calculation module (2) will calculate the average fixation time. Corresponding standard score Average number of fixations Corresponding standard score Average number of revisits Corresponding standard score Average number of eye twitches Corresponding standard score Average saccade amplitude Corresponding standard score Average blink rate Corresponding standard score and average blink time Corresponding standard score Substituting these values ​​into the cognitive load formula, we can obtain the cognitive load index. The formula for cognitive load is as follows: ； In the formula: This indicates the blink adjustment coefficient. The value range is between 0.05 and 0.5; when the cognitive load formula... When the term is less than or equal to 0, it indicates that the subject experiences relatively little psychological change within that area of ​​interest. In this case, if the cognitive load formula... A larger value indicates that the subject is in a state of fatigue; Therefore, when the cognitive load index If the value is less than or equal to -5, the subject needs to adjust their state and the test needs to be repeated.

9. A mental health assessment method based on eye movement features as described in claim 8, characterized in that: The calculation module (2) calculates the emotional arousal index corresponding to the five interest areas. Emotional change gradient Emotional Matching Score Cognitive Load Index To obtain the average emotional impact index of the subjects. Average mood change gradient Average sentiment matching score and average cognitive load index Then, the average emotional impact index was calculated. Average mood change gradient Average sentiment matching score and average cognitive load index Substituting these values ​​into the trauma risk formula, the trauma risk index is calculated. The trauma risk formula is as follows: ； In the formula: This indicates the weight of emotional triggers. The value ranges from 0.5 to 1.

5. Indicates the weight of emotional changes. The value ranges from 0.5 to 1.

5. Indicates the weight of sentiment matching. The value ranges from 0.5 to 1.

5. Indicates cognitive load weighting. The value range is between 0.5 and 1.5; , , , The specific value is determined based on the age group, gender, and family economic status of the subjects; When the judgment module (4) detects the subject's trauma risk index When the value is greater than or equal to 0.55, the subject will be marked as a high-risk individual for mental health, and will be connected to the manual assessment unit (41) for in-depth mental health assessment; when the judgment module (4) detects the subject's trauma risk index When the value is less than 0.55, the participants will be marked as low-risk individuals for mental health, and they will be connected to the question-and-answer unit (42) for automatic question-and-answer by artificial intelligence to understand their childhood trauma; the judgment module (4) will filter out the trauma risk index. Subjects with values ​​between 0.45 and 0.65 simultaneously sent instructions to the host (12) of the data collection module (1) to adjust the trauma risk index. Data from subjects with values ​​between 0.45 and 0.65 are sent to the threshold module (3).

10. A system based on the eye-tracking feature-based mental health assessment method according to claim 9, characterized in that: The system includes a data collection module (1), a calculation module (2), a threshold module (3), and a judgment module (4). The data collection module (1) is unidirectionally connected to the calculation module (2) and the threshold module (3), respectively. The threshold module (3) is unidirectionally connected to the calculation module (2). The calculation module (2) is unidirectionally connected to the judgment module (4). The judgment module (4) is unidirectionally connected to the data collection module (1). The data collection module (1) includes a display screen (11), a host computer (12), and an eye-tracking sampler (13). The display screen (11) is used to display the host computer. (12) Stored natural language segments; host (12) is used to store natural language segments and all data collected by eye tracker (13); eye tracker (13) is used to collect eye track data; host (12) is unidirectionally connected to calculation module (2) and threshold module (3) respectively; judgment module (4) includes human assessment unit (41) and question answering unit (42); human assessment unit (41) is used for human assessment of people with high mental health risk; question answering unit (42) is used for artificial intelligence question answering of people with low mental health risk.

Images