Psychological health evaluation method and system based on image projection

Abstract

The invention provides a psychological health evaluation method and system based on image projection in the technical field of psychological health evaluation. The method comprises the steps that S1, a large number of psychological health scales are acquired and analyzed to obtain a plurality of psychological scale problems; s2, converting each psychological scale question into a corresponding question image through a text graph artificial intelligence technology; s3, classifying each problem image, and setting an interaction mode of each problem image and a score corresponding to the interaction mode; s4, performing mental health evaluation based on each question image, and obtaining interaction data and an accumulated score in the mental health evaluation process; and S5, generating a psychological health evaluation report based on the interaction data and the accumulated score, displaying the psychological health evaluation report, and pushing and storing the psychological health evaluation report. The method has the advantage that the accuracy and reliability of mental health evaluation are greatly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of mental health assessment, and in particular to a mental health assessment method and system based on image projection. Background Art

[0002] With the acceleration of social pace and the increase of life pressure, mental health problems are becoming more and more common. When people have mental health problems, they need to seek help from psychologists, who will conduct mental health assessments and then adopt corresponding psychological counseling or treatment plans based on the assessment results.

[0003] Traditionally, mental health assessment has been done using mental health scales, such as questionnaires and interviews. However, traditional methods usually rely on the subjective statements and completion of the test subjects, which may be affected by social desirability effects and memory biases, thereby reducing the accuracy and reliability of the assessment results.

[0004] Therefore, how to provide a mental health assessment method and system based on image projection to improve the accuracy and reliability of mental health assessment has become a technical problem that needs to be solved urgently. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide a method and system for psychological health assessment based on image projection, so as to improve the accuracy and reliability of psychological health assessment.

[0006] In a first aspect, the present invention provides a method for evaluating mental health based on image projection, comprising the following steps:

[0007] Step S1, creating a database, obtaining a large number of mental health scales, parsing each mental health scale to obtain a number of mental scale questions, and storing each mental scale question in a structured manner in the database;

[0008] Step S2, using the Wenshengtu artificial intelligence technology, converting each psychological scale question in the database into a corresponding question image, and storing each question image in a structured manner in the database;

[0009] Step S3, classifying each of the problem images, and setting an interaction mode and a score corresponding to each of the problem images;

[0010] Step S4: performing a mental health assessment based on each of the problem images, and obtaining interaction data and accumulated scores during the mental health assessment process;

[0011] Step S5: Generate a mental health assessment report based on the interaction data and the accumulated score, display the mental health assessment report, and push and store the mental health assessment report.

[0012] Furthermore, in step S1, the psychological scale question also carries the corresponding psychological scale answer and the score corresponding to each psychological scale answer.

[0013] Furthermore, the step S2 is specifically as follows:

[0014] Each psychological scale question in the database is input into a pre-trained text encoder to obtain the corresponding question semantic vector:

[0015] Z T =E text (T);

[0016] Among them, Z T Represents the question semantic vector; E text () indicates text encoder; T indicates psychological scale questions;

[0017] Each of the question semantic vectors is converted into a corresponding question image through a diffusion model; the conversion process of the diffusion model is as follows:

[0018] Get a real image, add noise to the real image step by step for forward diffusion, so as to obtain a completely random noise image:

[0019]

[0020] Among them, x t represents the noise image obtained by adding noise to the real image in t steps; x t-1 represents the noise image obtained by adding noise to the real image in t-1 steps; β t represents the noise parameter of the noise added in step t; I represents the identity matrix; q(x t x t-1 ) means generating x at step t t and generate x at step t-1 t-1 The probability distribution of ; N() represents the multivariate Gaussian distribution function; Indicates that from is the mean, and β t I is the covariance of the multivariate Gaussian distribution and x is obtained by sampling t ;

[0021] For the noise image x t Step by step, remove the noise and perform reverse diffusion to obtain the problem image x0:

[0022] p θ (xt-1 x t ,Z T )=N(x t-1 ;μ θ (x t ,Z T ),σ θ 2 I);

[0023] Among them, p θ (x t-1 x t ,Z T ) means that given x t and Z T , generating x t-1 The conditional probability distribution of θ (x t ,Z T ) represents the pixel mean of the noise image predicted by the question semantic vector; Indicates the noise size in the denoising process, which is a scalar parameter learned by the neural network; represents the covariance matrix of the normal distribution;

[0024] Each of the problem images x0 is structured and stored in the database.

[0025] Furthermore, the step S3 is specifically as follows:

[0026] Classifying each question image based on the question type of each psychological scale question, and setting the interaction mode of each category of the question image and the score corresponding to the interaction mode respectively;

[0027] The interaction method is selection, clicking, dragging or voice input;

[0028] The score is the score of the psychological scale answer corresponding to the psychological scale question triggered on the question image based on each interaction mode.

[0029] Furthermore, the step S4 is specifically as follows:

[0030] Projecting the selected question images in sequence based on a preset display order, interacting through the question images to conduct a mental health assessment, and obtaining interaction data and accumulated scores during the mental health assessment;

[0031] The interaction data includes at least interaction operations, duration of each interaction operation, and selection preferences;

[0032] The cumulative score is obtained by adding up the scores obtained for each problem image.

[0033] In a second aspect, the present invention provides a mental health assessment system based on image projection, comprising the following modules:

[0034] A psychological scale question acquisition module is used to create a database, acquire a large number of psychological health scales, parse each of the psychological health scales to obtain a number of psychological scale questions, and store each of the psychological scale questions in a structured manner in the database;

[0035] A question image generation module, used to convert each psychological scale question in the database into a corresponding question image by using the Wensheng graph artificial intelligence technology, and store each question image in a structured manner in the database;

[0036] An interaction mode setting module, used to classify each of the problem images, and set the interaction mode of each of the problem images and the score corresponding to the interaction mode;

[0037] An image interaction module, used to perform a mental health assessment based on each of the problem images, and obtain interaction data and accumulated scores during the mental health assessment process;

[0038] The mental health assessment report generation module is used to generate a mental health assessment report based on the interaction data and the accumulated score, display the mental health assessment report, and push and store the mental health assessment report.

[0039] Furthermore, in the psychological scale question acquisition module, the psychological scale question also carries the corresponding psychological scale answer and the score corresponding to each psychological scale answer.

[0040] Furthermore, the problem image generation module is specifically used for:

[0041] Each psychological scale question in the database is input into a pre-trained text encoder to obtain the corresponding question semantic vector:

[0042] Z T =E text (T);

[0043] Among them, Z T Represents the question semantic vector; E text () indicates text encoder; T indicates psychological scale questions;

[0044] Each of the question semantic vectors is converted into a corresponding question image through a diffusion model; the conversion process of the diffusion model is:

[0045] Get a real image, add noise to the real image step by step for forward diffusion, so as to obtain a completely random noise image:

[0046]

[0047] Among them, x t represents the noise image obtained by adding noise to the real image in t steps; x t-1 represents the noise image obtained by adding noise to the real image in t-1 steps; β t represents the noise parameter of the noise added in step t; I represents the identity matrix; q(x t x t-1 ) means generating x at step t t and generate x at step t-1 t-1 The probability distribution of ; N() represents the multivariate Gaussian distribution function; Indicates that from is the mean, and β t I is the covariance of the multivariate Gaussian distribution and x is obtained by sampling t ;

[0048] For the noise image x t Step by step, remove the noise and perform reverse diffusion to obtain the problem image x0:

[0049] p θ (x t-1 x t ,Z T )=N(x t-1 ;μ θ (x t ,Z T ),σ θ 2 I);

[0050] Among them, p θ (x t-1 x t ,Z T ) means that given x t and Z T , generating x t-1 The conditional probability distribution of θ (x t ,Z T ) represents the pixel mean of the noise image predicted by the question semantic vector; Indicates the noise size in the denoising process, which is a scalar parameter learned by the neural network; represents the covariance matrix of the normal distribution;

[0051] Each of the problem images x0 is structured and stored in the database.

[0052] Furthermore, the interaction mode setting module is specifically used for:

[0053] Classifying each question image based on the question type of each psychological scale question, and setting the interaction mode of each category of the question image and the score corresponding to the interaction mode respectively;

[0054] The interaction method is selection, clicking, dragging or voice input;

[0055] The score is the score of the psychological scale answer corresponding to the psychological scale question triggered on the question image based on each interaction mode.

[0056] Furthermore, the image interaction module is specifically used for:

[0057] Projecting the selected question images in sequence based on a preset display order, interacting through the question images to conduct a mental health assessment, and obtaining interaction data and accumulated scores during the mental health assessment;

[0058] The interaction data at least includes interaction operations, duration of each interaction operation, and selection preferences;

[0059] The cumulative score is obtained by adding up the scores obtained for each problem image.

[0060] The advantages of the present invention are:

[0061] By obtaining a large number of mental health scales, parsing each mental health scale to obtain a number of psychological scale questions, and storing each psychological scale question in a structured manner in the created database; then using the Wenshengtu artificial intelligence technology, each psychological scale question in the database is converted into a corresponding question image, and each question image is structured and stored in the database; then each question image is classified, and the interaction method of each question image and the score corresponding to the interaction method are set, a mental health assessment is conducted based on each question image, and the interaction data and cumulative scores in the mental health assessment process are obtained. Finally, a mental health assessment report is generated based on the interaction data and the cumulative scores, the mental health assessment report is displayed, and the mental health assessment report is pushed and stored; that is, the traditional mental health scale is converted into question images for mental health assessment, so as to improve the interactivity and fun of the mental health assessment, improve the participation of the test subjects, reduce fatigue and resistance in the assessment process, and minimize the impact of social expectation effect and memory bias, which ultimately greatly improves the accuracy and reliability of the mental health assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] The present invention will be further described below in conjunction with embodiments with reference to the accompanying drawings.

[0063] Figure 1 It is a flow chart of a mental health assessment method based on image projection of the present invention.

[0064] Figure 2 It is a structural schematic diagram of a mental health assessment system based on image projection of the present invention. DETAILED DESCRIPTION

[0065] The overall idea of ​​the technical solution in the embodiments of the present application is as follows: converting the traditional mental health scale into question images for mental health assessment, so as to improve the interactivity and fun of the mental health assessment, minimize the impact of social expectation effect and memory bias, and thereby improve the accuracy and reliability of the mental health assessment.

[0066] Please refer to Figure 1 to Figure 2 As shown, a preferred embodiment of a mental health assessment method based on image projection of the present invention comprises the following steps:

[0067] Step S1, creating a database, obtaining a large number of mental health scales, parsing each mental health scale to obtain a number of mental scale questions, and storing each mental scale question in a structured manner in the database;

[0068] Step S2, using the Wenshengtu artificial intelligence technology, converting each psychological scale question in the database into a corresponding question image, and storing each question image in a structured manner in the database; that is, reconstructing the psychological scale question and presenting it in the form of an image;

[0069] Step S3, classifying each of the problem images, and setting an interaction mode and a score corresponding to each of the problem images;

[0070] Step S4: performing a mental health assessment based on each of the problem images, and obtaining interaction data and accumulated scores during the mental health assessment process;

[0071] Step S5: Generate a mental health assessment report based on the interaction data and the accumulated score, display the mental health assessment report, and push and store the mental health assessment report.

[0072] In step S1, the psychological scale question also carries the corresponding psychological scale answer and the score corresponding to each psychological scale answer.

[0073] The step S2 is specifically as follows:

[0074] Each psychological scale question in the database is input into a pre-trained text encoder to obtain the corresponding question semantic vector:

[0075] Z T =E text (T);

[0076] Among them, Z T Represents the question semantic vector; Etext () represents a text encoder; T represents a psychological scale question; the text encoder adopts CLIP (Contrast ive Language-Image Pretraining);

[0077] Each of the question semantic vectors is converted into a corresponding question image through a diffusion model; the conversion process of the diffusion model is as follows:

[0078] Get a real image, add noise to the real image step by step for forward diffusion, so as to obtain a completely random noise image:

[0079]

[0080] Among them, x t represents the noise image obtained by adding noise to the real image in t steps; x t-1 represents the noise image obtained by adding noise to the real image in t-1 steps; β t represents the noise parameter of the noise added in step t; I represents the identity matrix; q(x t x t-1 ) means generating x at step t t and generate x at step t-1 t-1 The probability distribution of ; N() represents the multivariate Gaussian distribution function; Indicates that from is the mean, and β t I is the covariance of the multivariate Gaussian distribution and x is obtained by sampling t ;

[0081] For the noise image x t Step by step, remove the noise and perform reverse diffusion to obtain the problem image x0:

[0082]

[0083] Among them, p θ (x t-1 x t ,Z T ) means that given x t and Z T , generating x t-1 The conditional probability distribution of θ (x t ,Z T ) represents the pixel mean of the noise image predicted by the question semantic vector; Indicates the noise size in the denoising process, which is a scalar parameter learned by the neural network; represents the covariance matrix of the normal distribution;

[0084] Each of the problem images x0 is structured and stored in the database.

[0085] The step S3 is specifically as follows:

[0086] Classifying each question image based on the question type of each psychological scale question, and setting the interaction mode of each category of the question image and the score corresponding to the interaction mode respectively;

[0087] The interaction method is selection, clicking, dragging or voice input;

[0088] The score is the score of the psychological scale answer corresponding to the psychological scale question triggered on the question image based on each interaction mode.

[0089] The step S4 is specifically as follows:

[0090] The selected question images are projected in sequence based on a preset display order, and the mental health assessment is performed through interaction through the question images, and the interaction data and accumulated scores in the mental health assessment process are obtained; in specific implementation, the corresponding question images can be selected for mental health assessment based on the user portrait generated based on age, gender, occupation, educational background, and marital status;

[0091] The interaction data at least includes the interaction operation, the duration of each interaction operation and the selection preference; in specific implementation, the image style of the question image can be dynamically adjusted based on the duration to ensure that the test subject is not likely to feel uncomfortable;

[0092] The cumulative score is obtained by adding up the scores obtained for each problem image.

[0093] The step S5 is specifically as follows:

[0094] Perform big data analysis of the interaction behavior on the interaction data, generate a mental health assessment report with the assessment time based on the big data analysis results and the cumulative score, display the mental health assessment report in real time on a display screen, encrypt the mental health assessment report into an encrypted report, push the encrypted report in real time to a pre-associated mobile terminal, and store and distribute the encrypted report.

[0095] The step of encrypting the mental health assessment report into an encrypted report is specifically as follows:

[0096] The mental health assessment report is encrypted into first-level encrypted data by the RC6 algorithm, each character of the first-level encrypted data is shifted right by 9 bits to obtain second-level encrypted data, the second-level encrypted data is encrypted by the 3DES algorithm to obtain third-level encrypted data, the third-level encrypted data is mapped based on a preset mapping rule to obtain fourth-level encrypted data, the number 8 and the letter B in the fourth-level encrypted data are swapped to obtain fifth-level encrypted data, and the fifth-level encrypted data is encrypted by the ChaCha20 algorithm to obtain an encrypted report.

[0097] The mental health assessment report is encrypted into first-level encrypted data through the RC6 algorithm, and each character of the first-level encrypted data is shifted right by 9 bits to obtain second-level encrypted data. The second-level encrypted data is encrypted through the 3DES algorithm to obtain third-level encrypted data. The third-level encrypted data is encrypted based on the mapping rule to obtain fourth-level encrypted data. The number 8 and the letter B in the fourth-level encrypted data are swapped to obtain fifth-level encrypted data. The fifth-level encrypted data is encrypted through the ChaCha20 algorithm to obtain an encrypted report. If the corresponding encryption algorithm or data transformation rule is not known, the encrypted report cannot be cracked. At least 6 security measures are taken before and after (RC6 algorithm, character shift right by 9 bits, 3DES algorithm, mapping rule, swapping the number 8 and the letter B, ChaCha20 algorithm), which greatly improves the security of the storage and push of the mental health assessment report.

[0098] The decryption process of the encrypted report is as follows: decrypt the encrypted report through the ChaCha20 algorithm to obtain level-five encrypted data, swap the number 8 and the letter B in the level-five encrypted data to obtain level-four encrypted data, map the level-four encrypted data based on a preset mapping rule to obtain level-three encrypted data, decrypt the level-three encrypted data through the 3DES algorithm to obtain level-two encrypted data, shift each character of the level-two encrypted data to the left by 9 bits to obtain level-one encrypted data, and decrypt the level-one encrypted data into a mental health assessment report through the RC6 algorithm.

[0099] A preferred embodiment of a mental health assessment system based on image projection of the present invention includes the following modules:

[0100] A psychological scale question acquisition module is used to create a database, acquire a large number of psychological health scales, parse each of the psychological health scales to obtain a number of psychological scale questions, and store each of the psychological scale questions in a structured manner in the database;

[0101] A question image generation module is used to convert each psychological scale question in the database into a corresponding question image through the Wensheng graph artificial intelligence technology, and store each of the question images in a structured manner in the database; that is, to reconstruct the psychological scale questions and present them in the form of images;

[0102] An interaction mode setting module, used to classify each of the problem images, and set the interaction mode of each of the problem images and the score corresponding to the interaction mode;

[0103] An image interaction module, used to perform a mental health assessment based on each of the problem images, and obtain interaction data and accumulated scores during the mental health assessment process;

[0104] The mental health assessment report generation module is used to generate a mental health assessment report based on the interaction data and the accumulated score, display the mental health assessment report, and push and store the mental health assessment report.

[0105] In the psychological scale question acquisition module, the psychological scale question also carries the corresponding psychological scale answer and the score corresponding to each psychological scale answer.

[0106] The problem image generation module is specifically used for:

[0107] Each psychological scale question in the database is input into a pre-trained text encoder to obtain the corresponding question semantic vector:

[0108] Z T =E text (T);

[0109] Among them, Z T Represents the question semantic vector; E text () represents a text encoder; T represents a psychological scale question; the text encoder adopts CLIP (Contrast ive Language-Image Pretraining);

[0110] Each of the question semantic vectors is converted into a corresponding question image through a diffusion model; the conversion process of the diffusion model is as follows:

[0111] Get a real image, add noise to the real image step by step for forward diffusion, so as to obtain a completely random noise image:

[0112]

[0113] Among them, x t represents the noise image obtained by adding noise to the real image in t steps; x t-1 represents the noise image obtained by adding noise to the real image in t-1 steps; β t represents the noise parameter of the noise added in step t; I represents the identity matrix; q(x t |x t-1 ) means generating x at step t tand generate x at step t-1 t-1 The probability distribution of ; N() represents the multivariate Gaussian distribution function; Indicates that from is the mean, and β t I is the covariance of the multivariate Gaussian distribution and x is obtained by sampling t ;

[0114] For the noise image x t Step by step, remove the noise and perform reverse diffusion to obtain the problem image x0:

[0115]

[0116] Among them, p θ (x t-1 x t ,Z T ) means that given x t and Z T , generating x t-1 The conditional probability distribution of θ (x t ,Z T ) represents the pixel mean of the noise image predicted by the question semantic vector; Indicates the noise size in the denoising process, which is a scalar parameter learned by the neural network; represents the covariance matrix of the normal distribution;

[0117] Each of the problem images x0 is structured and stored in the database.

[0118] The interaction mode setting module is specifically used for:

[0119] Classifying each question image based on the question type of each psychological scale question, and setting the interaction mode of each category of the question image and the score corresponding to the interaction mode respectively;

[0120] The interaction method is selection, clicking, dragging or voice input;

[0121] The score is the score of the psychological scale answer corresponding to the psychological scale question triggered on the question image based on each interaction mode.

[0122] The image interaction module is specifically used for:

[0123] The selected question images are projected in sequence based on a preset display order, and the mental health assessment is performed through interaction through the question images, and the interaction data and accumulated scores in the mental health assessment process are obtained; in specific implementation, the corresponding question images can be selected for mental health assessment based on the user portrait generated based on age, gender, occupation, educational background, and marital status;

[0124] The interaction data at least includes the interaction operation, the duration of each interaction operation and the selection preference; in specific implementation, the image style of the question image can be dynamically adjusted based on the duration to ensure that the test subject is not likely to feel uncomfortable;

[0125] The cumulative score is obtained by adding up the scores obtained for each problem image.

[0126] The mental health assessment report generation module is specifically used for:

[0127] Perform big data analysis of the interaction behavior on the interaction data, generate a mental health assessment report with the assessment time based on the big data analysis results and the cumulative score, display the mental health assessment report in real time on a display screen, encrypt the mental health assessment report into an encrypted report, push the encrypted report in real time to a pre-associated mobile terminal, and store and distribute the encrypted report.

[0128] The step of encrypting the mental health assessment report into an encrypted report is specifically as follows:

[0129] The mental health assessment report is encrypted into first-level encrypted data by the RC6 algorithm, each character of the first-level encrypted data is shifted right by 9 bits to obtain second-level encrypted data, the second-level encrypted data is encrypted by the 3DES algorithm to obtain third-level encrypted data, the third-level encrypted data is mapped based on a preset mapping rule to obtain fourth-level encrypted data, the number 8 and the letter B in the fourth-level encrypted data are swapped to obtain fifth-level encrypted data, and the fifth-level encrypted data is encrypted by the ChaCha20 algorithm to obtain an encrypted report.

[0130] The mental health assessment report is encrypted into first-level encrypted data through the RC6 algorithm, and each character of the first-level encrypted data is shifted right by 9 bits to obtain second-level encrypted data. The second-level encrypted data is encrypted through the 3DES algorithm to obtain third-level encrypted data. The third-level encrypted data is encrypted based on the mapping rule to obtain fourth-level encrypted data. The number 8 and the letter B in the fourth-level encrypted data are swapped to obtain fifth-level encrypted data. The fifth-level encrypted data is encrypted through the ChaCha20 algorithm to obtain an encrypted report. If the corresponding encryption algorithm or data transformation rule is not known, the encrypted report cannot be cracked. At least 6 security measures are taken before and after (RC6 algorithm, character shift right by 9 bits, 3DES algorithm, mapping rule, swapping the number 8 and the letter B, ChaCha20 algorithm), which greatly improves the security of the storage and push of the mental health assessment report.

[0131] The decryption process of the encrypted report is as follows: decrypt the encrypted report through the ChaCha20 algorithm to obtain level-five encrypted data, swap the number 8 and the letter B in the level-five encrypted data to obtain level-four encrypted data, map the level-four encrypted data based on a preset mapping rule to obtain level-three encrypted data, decrypt the level-three encrypted data through the 3DES algorithm to obtain level-two encrypted data, shift each character of the level-two encrypted data to the left by 9 bits to obtain level-one encrypted data, and decrypt the level-one encrypted data into a mental health assessment report through the RC6 algorithm.

[0132] In summary, the advantages of the present invention are:

[0133] By obtaining a large number of mental health scales, parsing each mental health scale to obtain a number of psychological scale questions, and storing each psychological scale question in a structured manner in the created database; then using the Wenshengtu artificial intelligence technology, each psychological scale question in the database is converted into a corresponding question image, and each question image is structured and stored in the database; then each question image is classified, and the interaction method of each question image and the score corresponding to the interaction method are set, a mental health assessment is conducted based on each question image, and the interaction data and cumulative scores in the mental health assessment process are obtained. Finally, a mental health assessment report is generated based on the interaction data and the cumulative scores, the mental health assessment report is displayed, and the mental health assessment report is pushed and stored; that is, the traditional mental health scale is converted into question images for mental health assessment, so as to improve the interactivity and fun of the mental health assessment, improve the participation of the test subjects, reduce fatigue and resistance in the assessment process, and minimize the impact of social expectation effect and memory bias, which ultimately greatly improves the accuracy and reliability of the mental health assessment.

[0134] Although the specific implementation modes of the present invention are described above, those skilled in the art should understand that the specific implementation modes described are only illustrative and are not intended to limit the scope of the present invention. Equivalent modifications and changes made by those skilled in the art in accordance with the spirit of the present invention should be included in the scope of protection of the claims of the present invention.

Claims

1. A mental health assessment method based on image projection, characterized in that: The steps include: Step S1, creating a database, obtaining a large number of mental health scales, parsing each mental health scale to obtain a number of mental scale questions, and storing each mental scale question in a structured manner in the database; Step S2, using the Wenshengtu artificial intelligence technology, converting each psychological scale question in the database into a corresponding question image, and storing each question image in a structured manner in the database; Step S3, classifying each of the problem images, and setting an interaction mode and a score corresponding to each of the problem images; Step S4: performing a mental health assessment based on each of the problem images, and obtaining interaction data and accumulated scores during the mental health assessment process; Step S5: Generate a mental health assessment report based on the interaction data and the accumulated score, display the mental health assessment report, and push and store the mental health assessment report.

2. A mental health assessment method based on image projection as claimed in claim 1, characterized in that: In step S1, the psychological scale question also carries the corresponding psychological scale answer and the score corresponding to each psychological scale answer.

3. A mental health assessment method based on image projection as claimed in claim 1, characterized in that: The step S2 is specifically as follows: Each psychological scale question in the database is input into a pre-trained text encoder to obtain the corresponding question semantic vector: Z T =E text (T); Among them, Z T Represents the question semantic vector; E text () indicates text encoder; T indicates psychological scale questions; Each of the question semantic vectors is converted into a corresponding question image through a diffusion model; the conversion process of the diffusion model is as follows: Get a real image, add noise to the real image step by step for forward diffusion, so as to obtain a completely random noise image: Among them, x t represents the noise image obtained by adding noise to the real image in t steps; x t-1 represents the noise image obtained by adding noise to the real image in t-1 steps; β t represents the noise parameter of the noise added in step t; I represents the identity matrix; q(x t |x t-1 ) means generating x at step t t and generate x at step t-1 t-1 The probability distribution of ; N( ) represents the multivariate Gaussian distribution function; Indicates that from is the mean, and β t I is the covariance of the multivariate Gaussian distribution and x is obtained by sampling t ; For the noise image x t Step by step, remove the noise and perform reverse diffusion to obtain the problem image x0: Among them, p θ (x t-1 |x t ,Z T ) means that given x t and Z T , generating x t-1 The conditional probability distribution of θ (x t ,Z T ) represents the pixel mean of the noise image predicted by the question semantic vector; Indicates the noise size in the denoising process, which is a scalar parameter learned by the neural network; Represents the covariance matrix of the normal distribution; Each of the problem images x0 is structured and stored in the database.

4. The mental health assessment method based on image projection as claimed in claim 1, characterized in that: The step S3 is specifically as follows: Classifying each question image based on the question type of each psychological scale question, and setting the interaction mode of each category of the question image and the score corresponding to the interaction mode respectively; The interaction method is selection, clicking, dragging or voice input; The score is the score of the psychological scale answer corresponding to the psychological scale question triggered on the question image based on each interaction mode.

5. The method for evaluating mental health based on image projection according to claim 1, characterized in that: The step S4 is specifically as follows: Projecting the selected question images in sequence based on a preset display order, interacting through the question images to conduct a mental health assessment, and obtaining interaction data and accumulated scores during the mental health assessment; The interaction data at least includes interaction operations, duration of each interaction operation, and selection preferences; The cumulative score is obtained by adding up the scores obtained for each problem image.

6. A mental health assessment system based on image projection, characterized in that: Includes the following modules: A psychological scale question acquisition module is used to create a database, acquire a large number of psychological health scales, parse each of the psychological health scales to obtain a number of psychological scale questions, and store each of the psychological scale questions in a structured manner in the database; A question image generation module, used to convert each psychological scale question in the database into a corresponding question image by using the Wensheng graph artificial intelligence technology, and store each question image in a structured manner in the database; An interaction mode setting module, used to classify each of the problem images, and set the interaction mode of each of the problem images and the score corresponding to the interaction mode; An image interaction module, used to perform a mental health assessment based on each of the problem images, and obtain interaction data and accumulated scores during the mental health assessment process; The mental health assessment report generation module is used to generate a mental health assessment report based on the interaction data and the accumulated score, display the mental health assessment report, and push and store the mental health assessment report.

7. A mental health assessment system based on image projection as claimed in claim 6, characterized in that: In the psychological scale question acquisition module, the psychological scale question also carries the corresponding psychological scale answer and the score corresponding to each psychological scale answer.

8. The mental health assessment system based on image projection as claimed in claim 6, characterized in that: The problem image generation module is specifically used for: Each psychological scale question in the database is input into a pre-trained text encoder to obtain the corresponding question semantic vector: Z T =E text (T); Among them, Z T Represents the question semantic vector; E text ( ) indicates text encoder; T indicates psychological scale questions; Each of the question semantic vectors is converted into a corresponding question image through a diffusion model; the conversion process of the diffusion model is as follows: Get a real image, add noise to the real image step by step for forward diffusion, so as to obtain a completely random noise image: Among them, x t represents the noise image obtained by adding noise to the real image in t steps; x t-1 represents the noise image obtained by adding noise to the real image in t-1 steps; β t represents the noise parameter of the noise added in step t; I represents the identity matrix; q(x t |x t-1 ) means generating x at step t t and generate x at step t-1 t-1 The probability distribution of ; N() represents the multivariate Gaussian distribution function; Indicates that from is the mean, and β t I is the covariance of the multivariate Gaussian distribution and x is obtained by sampling t ; For the noise image x t Step by step, remove the noise and perform reverse diffusion to obtain the problem image x0: Among them, p θ (x t-1 |x t ,Z T ) means that given x t and Z T , generating x t-1 The conditional probability distribution of θ (x t ,Z T ) represents the pixel mean of the noise image predicted by the question semantic vector; Indicates the noise size in the denoising process, which is a scalar parameter learned by the neural network; Represents the covariance matrix of the normal distribution; Each of the problem images x0 is structured and stored in the database.

9. The mental health assessment system based on image projection as claimed in claim 6, characterized in that: The interaction mode setting module is specifically used for: Classifying each question image based on the question type of each psychological scale question, and setting the interaction mode of each category of the question image and the score corresponding to the interaction mode respectively; The interaction method is selection, clicking, dragging or voice input; The score is the score of the psychological scale answer corresponding to the psychological scale question triggered on the question image based on each interaction mode.

10. The mental health assessment system based on image projection as claimed in claim 6, characterized in that: The image interaction module is specifically used for: Projecting the selected question images in sequence based on a preset display order, interacting through the question images to conduct a mental health assessment, and obtaining interaction data and accumulated scores during the mental health assessment; The interaction data at least includes interaction operations, duration of each interaction operation, and selection preferences; The cumulative score is obtained by adding up the scores obtained for each problem image.

Images