Immersive art treatment auxiliary system and application method thereof

Abstract

The invention discloses an immersive art treatment auxiliary system and an application method thereof, and relates to the technical field of psychotherapy, and the system comprises an information collection module, an environment construction module, a psychological analysis module, a scheme generation module, a model training module and a block chain recording module. The method has the advantages that a quantitative psychological assessment report is generated through multi-modal data analysis and a machine learning algorithm, the defect that traditional art therapy depends on subjective judgment is overcome, a patient can conduct art creation in an immersive virtual environment to obtain immersive experience, the sense of participation and the effect of therapy are enhanced, and the patient experience is improved. A personalized treatment scheme can be dynamically generated according to the psychological state and creation behavior of a patient, the treatment accuracy and effect are improved, a psychological analysis model is continuously optimized through the model training module, the accuracy and reliability of the model are improved, and the psychological analysis model can more accurately analyze the psychology of the patient in practical application.

Description

Technical Field

[0001] The present invention relates to the technical field of psychotherapy, and specifically to an immersive art therapy assistance system and its application method. Background Art

[0002] Mental health refers to an individual's good state in aspects such as cognition, emotion, and behavior, specifically manifested as a sound personality, strong adaptability, appropriate behavior, etc. Mental health is affected by both genetics and the environment. In particular, the parenting style of the original family in early childhood has a great impact on the development of mental health. With the increasing social pressure and the growing number of mental health problems, art therapy, as a non-drug treatment method, has gradually received wide attention. Traditional art therapy mainly relies on the subjective judgment of psychiatrists and lacks objective quantitative assessment tools, making it difficult to accurately measure the treatment effect;

[0003] Existing art therapy tools cannot provide an immersive creative environment, making it difficult for patients to obtain sufficient interaction and feedback during the creative process, affecting the sense of participation and effect of the treatment. It is impossible to conduct accurate psychological state analysis through the creative behaviors and works of patients, resulting in the difficulty of objectively evaluating the treatment effect. Although VR / AR technology has been applied in the medical field, its application in art therapy is still in its infancy, especially in terms of combining data analysis for psychological assessment, and there is no mature solution yet. For this reason, we propose an immersive art therapy assistance system and its application method. Summary of the Invention

[0004] The purpose of the present invention is to provide an immersive art therapy assistance system and its application method.

[0005] To achieve the above purpose, the present invention provides the following technical solutions: The treatment assistance system includes;

[0006] An information collection module that collects a large-scale dataset covering various art styles and mental health states;

[0007] An environment construction module that uses VR technology to create a virtual space to obtain a treatment space;

[0008] A psychological analysis module that real-time collects the patient's behavior data and biological signal data, and obtains a convolutional neural network CNN. Based on the convolutional neural network CNN, a psychological analysis model is constructed by combining the patient's behavior data and biological signal data, and the psychological state and the changing trend of the psychological state of the patient are analyzed using the psychological analysis model;

[0009] The solution generation module generates a treatment plan according to the patient's mental state and creative behavior, and generates a treatment plan according to the rate of change of the mental state collected in real time. At the same time, it deploys a molecular dynamics simulator based on a diffusion model, collects the patient's saliva, analyzes the patient's saliva metabolome data, and maps it to the virtual receptor protein target. Then, it predicts the neurotransmitter concentration field distribution through a three-dimensional graph convolutional network, and combines a reinforcement learning algorithm to generate a dynamic regulation strategy for dopamine and serotonin;

[0010] Neurotransmitters are chemical substances that transmit information between neurons. The neurotransmitter concentration field distribution refers to the distribution of neurotransmitter concentrations in a three-dimensional space within a specific spatial region of the nervous system (such as the brain). It describes the change in the content of neurotransmitters at different positions and reflects the inhomogeneous distribution characteristics of neurotransmitters in the nervous system;

[0011] Studying the neurotransmitter concentration field distribution usually requires the use of various techniques, such as immunohistochemistry, microdialysis, positron emission tomography (PET), etc. Immunohistochemistry can observe the distribution of neurotransmitters in tissue sections by labeling neurotransmitters. Microdialysis technology can monitor the concentration changes of neurotransmitters in specific brain regions in real time. PET can perform non-invasive neurotransmitter concentration imaging on live animals or humans to obtain three-dimensional distribution information of neurotransmitter concentrations throughout the brain;

[0012] Reinforcement learning is a type of machine learning algorithm that learns optimal behavioral strategies based on environmental feedback;

[0013] The model training module extracts the information recorded in a large-scale dataset to obtain verification cases, and records them in a verification table. It analyzes the types of behavioral data in the verification cases, extracts the corresponding numerical values for different types, calculates the distance index of different verification cases through the Euclidean distance algorithm, numbers the verification cases, extracts two verification cases with similar distance indices from the verification table, deletes them from the verification table, establishes a comparison group, records the two verification cases in the comparison group, obtains X comparison groups, and X comparison groups constitute the verification dataset. At this time, the psychological analysis model is trained and verified for X - 1 rounds. In each round of training, one of the comparison groups is extracted as the verification set, and the remaining X - 1 comparison groups are used as the training set. The training set is used to train the psychological analysis model, the verification set is used to verify the psychological analysis model, and the verification result information is recorded until the X comparison groups are used as the verification set once.

[0014] As a further solution of the present invention: The treatment assistance system further includes;

[0015] The blockchain recording module uses the private chain technology based on Ethereum to record all hash values of the data processing process through smart contracts, and uses the proof-of-work consensus mechanism to ensure the immutability of the data.

[0016] As a further solution of the present invention: after the large-scale dataset in the information acquisition module is obtained, image segmentation technology and style classification technology are used to preprocess the data in the large-scale dataset, and at the same time, the preprocessed data is labeled using image segmentation technology and style classification technology.

[0017] As a further solution of the present invention: when the treatment space in the environment construction module is obtained, text descriptions of the patient's preferences for color, layout, and lighting are acquired to obtain ideal space information. Then, based on the ideal space information, the treatment space is constructed, and at the same time, the treatment space is stored to form a treatment set. The color characteristics and layout characteristics of the treatment space are analyzed and marked in the corresponding treatment space. After the color characteristics and layout characteristics in the environment construction module are marked in the treatment space, a retrieval establishment threshold is synchronously established. When the number of treatment spaces in the treatment set is higher than the retrieval establishment threshold, a retrieval unit is established. The retrieval unit is used to collect the patient's description of the ideal space to obtain description information, analyze the color characteristics and layout characteristics in the description information to obtain retrieval characteristics, and then match the treatment space according to the retrieval characteristics and display the matched treatment space to the patient. At the same time, the patient's modification opinions are collected, the modification opinions are matched through a preset rule library to adjust the space parameters, and the modified treatment space is stored in the treatment set.

[0018] As a further solution of the present invention: after the treatment plan in the plan generation module is obtained, the corresponding behavioral tasks, creative tasks, and psychological intervention measures of the treatment plan are analyzed and published to the user, and the user's feedback is collected in real time to adjust the treatment space according to the user's feedback information.

[0019] As a further solution of the present invention: when the psychological analysis model in the model training module is trained, X - 1 verification result information is obtained. At this time, the X - 1 verification result information constitutes a verification result table, and the obtained verification result table is displayed to Y professional personnel, and then the professional personnel conduct point-by-point evaluation, and then calculate the average value of the comments on different verification result information.

[0020] As a further solution of the present invention: when calculating the average value of the comments in the model training module, let the number of professional personnel be P, and let the different point-by-point evaluations be D F , let the average value of the comments be D J :

[0021]

[0022] The average value of the comments is calculated through the above formula.

[0023] As a further solution of the present invention: after the average evaluation in the model training module is calculated, a display threshold is established, where the display threshold is set by the staff, and the display threshold is set as Z Y , when D J ≥Z Y , the corresponding verification result of the average evaluation is displayed. When D J <Z Y , the corresponding verification result of the average evaluation will not be displayed.

[0024] In addition, an application method of an immersive art therapy assistance system is also provided, including the following steps:

[0025] S100. Collect a large-scale dataset covering various art styles and mental health states, use VR technology to create a virtual space, and obtain a treatment space;

[0026] S200. Real-time collect the behavior data and biosignal data of the patient, build a psychological analysis model based on the convolutional neural network CNN combined with the patient's behavior data and biosignal data, and use the psychological analysis model to analyze the patient's psychological state and the changing trend of the psychological state;

[0027] S300. Generate a personalized treatment plan, obtain verification cases, generate dynamic regulation strategies for dopamine and serotonin, and record them in the verification form. Analyze the types of behavior data in the verification cases, extract the corresponding values of different types, and calculate the distance index of different verification cases through the Euclidean distance algorithm;

[0028] S400. Number the verification cases, extract two verification cases with similar distance indexes from the verification form, delete them from the verification form, establish a comparison group, record the two verification cases in the comparison group, and obtain X comparison groups, and the X comparison groups constitute a verification dataset;

[0029] S500. Conduct X-1 rounds of training and verification on the psychological analysis model. In each round of training, one of the comparison groups is extracted as the verification set, and the remaining X-1 comparison groups are used as the training set. Use the training set to train the psychological analysis model, use the verification set to verify the psychological analysis model, record the verification result information, and use the blockchain to record the entire data processing process.

[0030] Adopting the above technical solution, compared with the prior art, the beneficial effects of the present invention are as follows:

[0031] 1. The present invention generates a quantitative psychological assessment report through multi-modal data analysis and machine learning algorithms, making up for the deficiency of relying on subjective judgment in traditional art therapy. Patients can create art in an immersive virtual environment, obtaining an immersive experience, enhancing the sense of participation and effect of the treatment, being able to dynamically generate personalized treatment plans according to the patient's psychological state and creative behavior, improving the accuracy and effect of the treatment. The model training module helps to continuously optimize the psychological analysis model, improve the accuracy and reliability of the model, and enable it to more accurately analyze the patient's psychology in practical applications;

[0032] 2. The information collection module of the present invention helps to improve the quality and usability of data, making subsequent analysis and model training based on these data more accurate and efficient, facilitating subsequent targeted analysis and treatment plan formulation according to different styles. The environment construction module can fully meet the personalized needs of patients, create a treatment environment more in line with their psychological expectations for patients, improve the participation and satisfaction of patients, increase the generation speed of the treatment space, and improve the utilization rate of the previous treatment space;

[0033] 3. The model training module of the present invention makes the evaluation more comprehensive and scientific, can integrate the opinions of multiple professionals, more objectively reflect the advantages and disadvantages of the verification results, avoid the ambiguity and subjectivity in the calculation process, ensure the accuracy and reliability of the evaluation results, flexibly control the display content, improve work efficiency, and can also focus more on high-quality verification results to promote the optimization of the psychological analysis model. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a schematic diagram of the system flow in the embodiment of the present invention;

[0035] Figure 2 It is a schematic diagram of the method steps in the embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0036] The following further describes the specific embodiments of the present invention with reference to the drawings. It should be noted here that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation to the present invention.

[0037] In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0038] Please refer to the attached Figure 1 - attached Figure 2 , the immersive art therapy assistance system and its application method of the present invention, the therapy assistance system includes;

[0039] An information collection module, which collects a large-scale data set covering various art styles and mental health states;

[0040] An environmental construction module that uses VR technology to create a virtual space and obtain a treatment space;

[0041] A psychological analysis module that collects the patient's behavioral data and biosignal data in real time, obtains a convolutional neural network CNN, constructs a psychological analysis model based on the convolutional neural network CNN combined with the patient's behavioral data and biosignal data, and uses the psychological analysis model to analyze the patient's psychological state and the changing trend of the psychological state;

[0042] A treatment plan generation module that generates a treatment plan according to the patient's psychological state and creative behavior, and according to the rate of change of the psychological state collected in real time. At the same time, it deploys a molecular dynamics simulator based on a diffusion model, collects the patient's saliva, analyzes the patient's saliva metabolome data, maps it to virtual receptor protein targets, then predicts the distribution of neurotransmitter concentration fields through a three-dimensional graph convolutional network, and generates a dynamic regulation strategy for dopamine and serotonin in combination with a reinforcement learning algorithm;

[0043] The diffusion model is a generative model. It is based on modeling the data distribution and generates new samples by learning the diffusion process of the data. In molecular dynamics simulations, the diffusion model regards the movement of molecules as a diffusion process, and models the positions and states of molecules at different times to simulate the dynamic behavior of molecules. For example, the model first encodes the initial state of the molecular system, and then through a series of diffusion steps, gradually introduces noise to simulate the state changes of molecules under the influence of factors such as thermal motion. In the reverse process, the model learns to recover from the noisy state to the initial state, thus realizing the simulation of the molecular dynamics process;

[0044] The patient's saliva metabolome data refers to a series of information obtained after qualitative and quantitative analysis of various metabolites in the patient's saliva. Various analysis techniques are used to detect and analyze the metabolites in the saliva. Commonly used techniques include nuclear magnetic resonance (NMR), gas chromatography-mass spectrometry (GC-MS), liquid chromatography-mass spectrometry (LC-MS), etc. These techniques can separate, identify, and quantify the metabolites in the saliva, thus obtaining the metabolome data;

[0045] Virtual receptor protein targets are analyzed using computer software for the three-dimensional structure of known proteins to determine the sites that may bind to ligands. These sites are usually pockets or clefts on the protein surface, with specific amino acid compositions and spatial structures, and can bind to ligands through hydrogen bonds, van der Waals forces, electrostatic interactions, etc. For example, when designing anti-cancer drugs, virtual targets can be constructed for the active sites of specific proteins (such as epidermal growth factor receptor) overexpressed in tumor cells;

[0046] The 3D graph convolutional network is a neural network architecture based on deep learning. It is designed and optimized for 3D data based on the traditional graph convolutional network. By defining convolutional operations on 3D graphs, it extends the concept of convolutional kernels in traditional convolutional neural networks to 3D graph structures. Utilizing the information of nodes and edges in the graph, the 3D graph convolutional network can automatically learn the spatial features and structural information of the data. Specifically, it aggregates the features of nodes and their neighbor nodes, and through multiple layers of non-linear transformations and convolutional operations, extracts higher-level feature representations, enabling effective modeling and analysis of 3D data.

[0047] The model training module extracts the information recorded in the large-scale dataset, obtains verification cases, and records them in the verification table. It analyzes the types of behavior data in the verification cases, extracts the corresponding numerical values for different types, calculates the distance indices of different verification cases through the Euclidean distance algorithm, numbers the verification cases, extracts two verification cases with similar distance indices from the verification table, deletes them from the verification table, establishes a comparison group, records the two verification cases in the comparison group, and obtains X comparison groups. The X comparison groups constitute the verification dataset. At this time, the psychological analysis model is trained and verified for X - 1 rounds. In each round of training, one of the comparison groups is extracted as the verification set, and the remaining X - 1 comparison groups are used as the training set. The training set is used to train the psychological analysis model, the verification set is used to verify the psychological analysis model, and the verification result information is recorded until each of the X comparison groups is used as the verification set once.

[0048] The treatment assistance system also includes;

[0049] The blockchain recording module adopts the private chain technology based on Ethereum, records all hash values of the data processing process through smart contracts, and uses the proof-of-work consensus mechanism to ensure the immutability of the data.

[0050] In an embodiment of the present invention: After the large-scale dataset in the information collection module is obtained, image segmentation technology and style classification technology are used to preprocess the data in the large-scale dataset, and at the same time, the preprocessed data is labeled using image segmentation technology and style classification technology.

[0051] In one embodiment of the present invention: when the treatment space in the environment construction module is obtained, it will acquire the patient's written descriptions of color, layout, and lighting to obtain the ideal space information, and then construct the treatment space according to the ideal space information. At the same time, the treatment space will be stored to form a treatment set, and the color characteristics and layout characteristics of the treatment space will be analyzed and marked in the corresponding treatment space. After the color characteristics and layout characteristics in the environment construction module are marked in the treatment space, a retrieval establishment threshold will be synchronously established. When the number of treatment spaces in the treatment set is higher than the retrieval establishment threshold, a retrieval unit will be established. The retrieval unit is used to collect the patient's descriptions of the ideal space to obtain description information, analyze the color characteristics and layout characteristics in the description information to obtain retrieval characteristics, and then match the treatment space according to the retrieval characteristics and display the matched treatment space to the patient. At the same time, the patient's modification opinions will be collected, the modification opinions will be matched through a preset rule library and the space parameters will be adjusted, and the modified treatment space will be stored in the treatment set.

[0052] In one embodiment of the present invention: after the treatment plan in the plan generation module is obtained, it will analyze the corresponding behavioral tasks, creative tasks, and psychological intervention measures of the treatment plan, publish them to the user, and collect the user's feedback in real time to adjust the treatment space according to the user's feedback information.

[0053] In one embodiment of the present invention: when the psychological analysis model in the model training module is trained, X - 1 verification result information will be obtained. At this time, the X - 1 verification result information will form a verification result table, and the obtained verification result table will be displayed to Y professional personnel, and then the professional personnel will conduct point - by - point evaluation, and then calculate the average value of the point evaluations of different verification result information.

[0054] In one embodiment of the present invention: when calculating the average value of the point evaluations in the model training module, let the number of professional personnel be P, and let the different point - by - point evaluations be D F Let the average value of the point evaluations be D J :

[0055]

[0056] The average value of the point evaluations is calculated through the above formula.

[0057] In one embodiment of the present invention: after the average value of the point evaluations in the model training module is calculated, a display threshold will be established, where the display threshold is set by the staff independently. Let the display threshold be Z Y When D J ≥Z Y then the verification result corresponding to the average value of the point evaluations will be displayed. When D J <Z YWhen it is, the verification result corresponding to the comment average value will not be displayed.

[0058] Example 1. Please refer to the appendix Figure 1 - appendix Figure 2 Collect a large-scale dataset covering various artistic styles and mental health states, create a virtual space using VR technology to obtain a treatment space, collect the patient's behavioral data and biosignal data in real time, build a psychological analysis model based on the convolutional neural network CNN by combining the patient's behavioral data and biosignal data, analyze the patient's mental state and the changing trend of the mental state using the psychological analysis model, generate a personalized treatment plan, obtain verification cases, generate dynamic regulation strategies for dopamine and serotonin, record them in the verification table, analyze the types of behavioral data in the verification cases, extract the corresponding numerical values for different types, calculate the distance index of different verification cases through the Euclidean distance algorithm, number the verification cases, extract two verification cases with similar distance indexes from the verification table, delete them from the verification table, establish a comparison group, record the two verification cases in the comparison group, obtain X comparison groups, and the X comparison groups constitute the verification dataset. Conduct X - 1 rounds of training and verification on the psychological analysis model. In each round of training, one of the comparison groups is extracted as the verification set, and the remaining X - 1 comparison groups are the training set. Use the training set to train the psychological analysis model, use the verification set to conduct verification processing on the psychological analysis model, record the verification result information, and use the blockchain to record the entire data processing process.

[0059] Example 2. Please refer to the appendix Figure 1 - appendix Figure 2 Collect a large-scale dataset covering various artistic styles and mental health states, preprocess the data in the large-scale dataset using image segmentation technology and style classification technology, and at the same time label the preprocessed data using image segmentation technology and style classification technology. Create a virtual space using VR technology, obtain the text description of the ideal space by the patient, construct a treatment space according to the ideal space information, and at the same time store the treatment space to form a treatment set. Analyze the color characteristics and layout characteristics of the treatment space. When the number of treatment spaces in the treatment set is higher than the retrieval establishment threshold, use the retrieval unit to collect the description of the ideal space by the patient, analyze the color characteristics and layout characteristics in the description information, match the treatment space according to the retrieval characteristics, display the matched treatment space to the patient, collect the patient's modification opinions, automatically modify the treatment space according to the modification opinions, generate a personalized treatment plan, analyze the corresponding behavioral tasks, creative tasks, and psychological intervention measures of the treatment plan, publish them to the user, collect the user's feedback in real time, and adjust the treatment space according to the user's feedback information.

[0060] Example 3. Please refer to the appendix Figure 1 - appendixFigure 2 , extract the information recorded in the large-scale dataset to obtain verification cases, record them in the verification table, analyze the types of behavior data in the verification cases, extract the corresponding numerical values for different types, calculate the distance index of different verification cases through the Euclidean distance algorithm, number the verification cases, extract two verification cases with similar distance indices from the verification table, delete them from the verification table, establish a comparison group, record the two verification cases in the comparison group, obtain X comparison groups, and then use the X comparison groups to train the psychological analysis model to obtain X - 1 verification result information. At this time, the X - 1 verification result information will form a verification result table, and the obtained verification result table will be displayed to Y professional personnel, and then the professional personnel will conduct point-by-point evaluation, and then calculate the average value of the evaluations of different verification result information, and establish a display threshold, where the display threshold is set by the staff independently, and the display threshold is set to Z Y , when D J ≥Z Y , the verification result corresponding to the average evaluation value will be displayed. When D J <Z Y , the verification result corresponding to the average evaluation value will not be displayed.

[0061] Specifically, when calculating the distance index of different verification cases using the Euclidean distance formula, the numerical values of the same information type in different verification cases will be extracted. Let the different numerical values of the first verification case be B S , let the different numerical values of the second verification case be E S , let the number of different numerical values be M, and let the distance index be J Z :

[0062]

[0063] The distance index is calculated through the above formula, and at the same time, the information type distribution order of the first verification case and the second verification case is the same.

[0064] Specifically, when determining the display threshold, the highest point-by-point evaluation and the lowest point-by-point evaluation can also be extracted, and then the average evaluation of the highest point-by-point evaluation and the lowest point-by-point evaluation is calculated to obtain the range mean. At this time, the range mean can be used as the display threshold.

[0065] Specifically, a convolutional neural network is a deep learning model designed specifically for processing data with a grid structure. In the mental health census of a large population, a convolutional neural network can quickly analyze various psychological assessment data, such as questionnaire survey results, behavioral test data, etc., generate mental health assessment reports for each person, and screen out people who may have psychological problems. Biological signal data refers to the data contained in various signals that can reflect the physiological and pathological states of organisms, and these signals can be collected and analyzed through different technical means.

[0066] Working principle:

[0067] First, collect a large-scale dataset covering various art styles and mental health states, preprocess the data in the large-scale dataset using image segmentation technology and style classification technology, and at the same time label the preprocessed data using image segmentation technology and style classification technology. Use VR technology to create a virtual space, obtain the text description of the ideal space by the patient, construct a treatment space according to the ideal space information, and store the treatment space at the same time to form a treatment set. Analyze the color characteristics and layout characteristics of the treatment space. When the number of treatment spaces in the treatment set is higher than the retrieval establishment threshold, a retrieval unit will be established. Use the retrieval unit to collect the description of the ideal space by the patient, analyze the color characteristics and layout characteristics in the description information, match the treatment space according to the retrieval characteristics, display the matched treatment space to the patient, collect the modification opinions of the patient, automatically modify the treatment space according to the modification opinions, generate a personalized treatment plan, generate a dynamic regulation strategy for dopamine and serotonin, analyze the corresponding behavioral tasks, creative tasks and psychological intervention measures of the treatment plan, publish them to the user, collect the feedback of the user in real time, adjust the treatment space according to the feedback information of the user, obtain verification cases, and record them in the verification form. Analyze the types of behavioral data in the verification cases, extract the corresponding numerical values of different types, calculate the distance index of different verification cases through the Euclidean distance algorithm, number the verification cases, extract two verification cases with similar distance indexes from the verification form, and delete them from the verification form. Establish a comparison group, record the two verification cases in the comparison group, obtain X comparison groups, and X comparison groups constitute the verification dataset. Conduct X-1 rounds of training and verification on the psychological analysis model. In each round of training, one of the comparison groups will be extracted as the verification set, and the remaining X-1 comparison groups will be the training set. Use the training set to train the psychological analysis model, use the verification set to verify the psychological analysis model, and record the verification result information. Use the blockchain to record the entire data processing process, and obtain Verification result information, then X-1 verification result information will form a verification result table, and the obtained verification result table will be displayed to Y professionals, who will then score and evaluate it, and then calculate the average value of the comments on different verification result information, and establish a display threshold, where the display threshold is set by the staff independently, and the display threshold is set to Z Y , when D J ≥Z Y When D J <Z Y When the verification result of the review average is displayed, the blockchain is used to record the entire data processing process, and the entire workflow is completed.

[0068] Although the present invention is disclosed as above in terms of preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art may make possible changes and modifications without departing from the spirit and scope of the present invention. Therefore, any modification, equivalent change and modification made to the above embodiments according to the technical essence of the present invention without departing from the content of the technical solution of the present invention shall fall within the scope of protection defined by the claims of the present invention.

Claims

1. Immersive art therapy assistance system, characterized by: The treatment assistance system includes; An information collection module that collects a large-scale dataset covering various art styles and mental health states; An environment construction module that uses VR technology to create a virtual space to obtain a treatment space; A psychological analysis module that collects the patient's behavior data and biometric signal data in real time, and obtains a convolutional neural network CNN. Based on the convolutional neural network CNN, a psychological analysis model is constructed by combining the patient's behavior data and biometric signal data, and the psychological analysis model is used to analyze the patient's psychological state and the changing trend of the psychological state; A treatment plan generation module that generates a treatment plan according to the patient's psychological state and creative behavior, and according to the rate of change of the psychological state collected in real time. At the same time, a molecular dynamics simulator based on a diffusion model is deployed to collect the patient's saliva, analyze the patient's saliva metabolome data, and map it to a virtual receptor protein target. Then, the three-dimensional graph convolutional network is used to predict the distribution of neurotransmitter concentration fields, and a dynamic regulation strategy for dopamine and serotonin is generated by combining a reinforcement learning algorithm; A model training module that extracts the information recorded in the large-scale dataset to obtain verification cases, records them in a verification table, analyzes the types of behavior data in the verification cases, extracts the corresponding numerical values for different types, calculates the distance index of different verification cases through the Euclidean distance algorithm, numbers the verification cases, extracts two verification cases with similar distance indices from the verification table, deletes them from the verification table, establishes a comparison group, records the two verification cases in the comparison group, and obtains X comparison groups. The X comparison groups constitute a verification dataset. At this time, the psychological analysis model is trained and verified for X-1 rounds. In each round of training, one of the comparison groups is extracted as the verification set, and the remaining X-1 comparison groups are used as the training set. The training set is used to train the psychological analysis model, the verification set is used to verify the psychological analysis model, and the verification result information is recorded until the X comparison groups are used as the verification set once.

2. The immersive art therapy assistance system according to claim 1, wherein: The treatment assistance system further includes; A blockchain recording module that uses the private chain technology based on Ethereum to record all hash values of the data processing process through a smart contract, and uses the proof-of-work consensus mechanism to ensure the immutability of the data.

3. The immersive art therapy assistance system according to claim 1, wherein: After the large-scale dataset in the information collection module is obtained, image segmentation technology and style classification technology are used to preprocess the data in the large-scale dataset, and at the same time, the preprocessed data is labeled using image segmentation technology and style classification technology.

4. The immersive art therapy assistance system according to claim 1, wherein: When the treatment space in the environment construction module is obtained, it will acquire the patient's text descriptions of color, layout, and lighting, obtain the ideal space information, then construct the treatment space according to the ideal space information, and at the same time store the treatment space to form a treatment set. It will also analyze the color characteristics and layout characteristics of the treatment space, and mark the color characteristics and layout characteristics in the corresponding treatment space. After the color characteristics and layout characteristics in the environment construction module are marked in the treatment space, a retrieval establishment threshold will be synchronously established. When the number of treatment spaces in the treatment set is higher than the retrieval establishment threshold, a retrieval unit will be established. The retrieval unit is used to collect the patient's description of the ideal space, obtain the description information, analyze the color characteristics and layout characteristics in the description information, obtain the retrieval characteristics, then match the treatment space according to the retrieval characteristics, and display the matched treatment space to the patient. At the same time, it will collect the patient's modification opinions, match the modification opinions through a preset rule base and adjust the space parameters, and store the modified treatment space into the treatment set.

5. The immersive art therapy assistance system according to claim 1, wherein: After the treatment plan in the plan generation module is obtained, it will analyze the corresponding behavioral tasks, creative tasks, and psychological intervention measures of the treatment plan, publish them to the user, and collect the user's feedback in real time to adjust the treatment space according to the user's feedback information.

6. The immersive art therapy assistance system according to claim 1, characterized in that: When the psychological analysis model in the model training module is trained, it will obtain X - 1 verification result information. At this time, the X - 1 verification result information will form a verification result table, and the obtained verification result table will be displayed to Y professional personnel, and then the professional personnel will conduct point - by - point evaluation, and then calculate the average value of the comments on different verification result information.

7. The immersive art therapy assistance system according to claim 6, wherein: When calculating the average comment value in the model training module, let the number of professionals be P, and let different point score evaluations be D F , and let the average comment value be D J : Calculate the average value of the comments through the above formula.

8. The immersive art therapy assistance system according to claim 7, wherein: After the average score in the model training module is calculated, a display threshold will be established, where the display threshold is set by the staff, and let the display threshold be Z Y , when D J ≥Z Y , the corresponding verification result of the average score will be displayed. When D J <Z Y , the corresponding verification result of the average score will not be displayed.

9. An application method of an immersive art therapy assistance system applicable to the immersive art therapy assistance system according to any one of claims 1-8, characterized in that, Include the following steps: S100. Collect a large - scale dataset covering various art styles and mental health states, use VR technology to create a virtual space, and obtain a treatment space; S200. Collect the patient's behavioral data and biological signal data in real time, construct a psychological analysis model based on the convolutional neural network CNN combined with the patient's behavioral data and biological signal data, and use the psychological analysis model to analyze the patient's psychological state and the change trend of the psychological state; S300. Generate a personalized treatment plan, obtain verification cases, generate a dynamic regulation strategy for dopamine and serotonin, and record it in the verification table. Analyze the types of behavioral data in the verification cases, extract the corresponding values for different types, and calculate the distance index of different verification cases through the Euclidean distance algorithm; S400. Number the verification cases, extract two verification cases with similar distance indices from the verification table, delete them from the verification table, establish a comparison group, record the two verification cases in the comparison group, obtain X comparison groups, and the X comparison groups constitute the verification dataset; S500. Conduct X-1 rounds of training and verification on the psychoanalysis model. In each round of training, one of the comparison groups is extracted as the verification set, and the remaining X-1 comparison groups are used as the training set. Use the training set to train the psychoanalysis model, use the verification set to conduct verification processing on the psychoanalysis model, record the verification result information, and use the blockchain to record the entire data processing process.

Images