Psychological stress state assessment method based on multi-knowledge domain information
By collecting physiological and psychological information, constructing a key factor association network and combining it with a deep learning model, the shortcomings of traditional assessment methods are addressed, enabling accurate assessment and personalized intervention support for psychological stress.
Patent Information
- Application Number
- CN202510956785.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2024-12-19
- Filing Date
- 2025-07-11
- Publication Date
- 2025-10-28
AI Technical Summary
Traditional psychological stress state assessment methods are difficult to achieve objective and continuous qualitative and quantitative assessments in the absence of clear stressors, and there are large individual differences, making it difficult to accurately match subjective psychological perceptions with objective physiological manifestations.
By collecting physiological and psychological information, introducing experts from multiple fields to formulate key factors, constructing a key factor association network and performing topological analysis, and combining deep learning models to perform multimodal feature representation and classification, a framework for assessing psychological stress state is constructed.
It enables precise assessment of psychological stress, improves the accuracy and reliability of the assessment, and provides strong support for psychological intervention and treatment.
Smart Images
Figure CN120853887A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of signal analysis technology, specifically to a method for assessing psychological stress state based on multi-knowledge domain information. Background Technology
[0002] Since psychological phenomena and physiological reactions are corresponding and interdependent, changes in psychological state directly affect changes in physiological indicators. Therefore, interpreting this barometer of physiological characteristic changes, and clearly or statistically accurately identifying one or a combination of physiological characteristics that represents changes in psychological stress, can effectively identify psychological stress states. This not only helps to understand the severity and incidence of psychological trauma in stressed individuals, providing a basis for decision-making regarding resource allocation for psychological assistance work, but also enables more precise identification of key targets for psychological assistance work, providing personalized intervention and treatment, and providing a basis for subsequent intervention work and consolidation of therapeutic effects.
[0003] However, the human body is an extremely complex nonlinear dynamic system, constantly influenced by various internal environmental regulations and external environmental disturbances, exhibiting significant individual differences. Physiological and psychological state changes are coupled and influence each other on both temporal and spatial scales. Therefore, traditional stress assessment methods based on psychological scales or psychophysiological measurements are difficult to objectively and continuously provide qualitative and quantitative assessments of psychological stress. With the establishment of behaviorist theory and the development of systems theory, cybernetics, and information theory, the acquisition of comprehensive life information perception by individuals has broken through the limitations of space and time. How to improve and perfect the theory and comprehensive evaluation system for estimating human psychosomatic states based on the theory of biofeedback closed-loop control systems, and how to deterministically assess psychological stress states in the absence of clear stressors, and achieve a deterministic match between subjective psychological perception and objective physiological representation during feedback intervention, remains an urgent problem to be solved in the psychosomatic field. Summary of the Invention
[0004] To address the shortcomings of existing technologies, this invention provides a method for assessing psychological stress states based on multi-knowledge domain information, thus solving the problems mentioned in the background section.
[0005] To achieve the above objectives, the present invention provides the following technical solution: a method for assessing psychological stress state based on multi-knowledge domain information, specifically including the following steps:
[0006] S1. Collect physiological and psychological information from several subjects, and obtain a dataset after preprocessing;
[0007] S2. Introduce experts from multiple fields to formulate key factors according to different fields;
[0008] S3. Extract key factors from the dataset, generate a key factor set, construct a key factor association network, and perform topology analysis.
[0009] S4. Utilize deep learning models to fuse physiological information, psychological information, and key factors to achieve a unified multimodal feature representation;
[0010] S5. Construct a psychological stress state classifier based on multimodal features, and after training and optimization, complete the construction of a multi-knowledge domain psychological stress state assessment framework.
[0011] S6. Input the patient's physiological and psychological information into the multi-knowledge domain psychological stress state assessment framework, and output the patient's psychological stress state assessment results.
[0012] The present invention is further configured such that: the physiological information in S1 includes individual characteristic information, electrocardiogram information, skin conductance information, respiratory rate, and blood oxygen concentration;
[0013] The individual characteristics information includes age and gender.
[0014] The present invention is further configured such that the method of acquiring psychological information in S1 includes:
[0015] Psychological information of the subjects was obtained through psychological scales and questionnaires.
[0016] The present invention is further configured such that the multi-domain experts in S2 include psychology experts, medical experts, biological experts and information science experts.
[0017] The present invention is further configured such that the key factor in S2 is a factor that has a significant impact on the assessment of psychological stress state.
[0018] The present invention is further configured such that the method for constructing the key factor association network in S3 includes:
[0019] Statistical methods are used to calculate the correlation between key factors, and the correlation values are used to construct the edges and weights of the association network, where nodes represent key factors and edges represent the correlation between factors.
[0020] The present invention is further configured such that the topology analysis method in S3 includes:
[0021] Obtain the degree, clustering coefficient, and path length of nodes in the topology of an associated network. The degree of a node is the number of edges directly connected to it. The clustering coefficient is an indicator of the tightness of the connection between the node and its neighboring nodes. The path length is the number of shortest edges that need to be traversed from one node to another in the associated network.
[0022] Obtain the weighted sum of degree, clustering coefficient and path length of each node in the interconnected network, and use it as the influence score of that node. Sort all nodes according to the influence score to obtain the influence of different nodes in the interconnected network.
[0023] The present invention is further configured such that the methods for fusing physiological information, psychological information and key factors in S4 include: feature splicing, additive fusion and multiplicative fusion.
[0024] This invention provides a method for assessing psychological stress state based on multi-knowledge domain information. It has the following beneficial effects:
[0025] This invention proposes key factors through multiple knowledge domains, combines physiological and psychological information with key factors to generate multimodal data, and improves the confidence of data analysis. By constructing key factor association networks and topological analysis, it not only helps to reveal the relationships between factors, optimize feature selection, and improve model performance, but also, in conjunction with deep learning models for feature fusion and classification, achieves accurate assessment of psychological stress state. This not only improves the accuracy and reliability of psychological stress state assessment, but also provides strong support for subsequent psychological intervention and treatment. Attached Figure Description
[0026] Figure 1 This is a schematic diagram of the process of the present invention. Detailed Implementation
[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0028] See also Figure 1 The present invention provides the following technical solutions:
[0029] Example 1
[0030] The method for assessing psychological stress based on multi-knowledge domain information includes the following steps:
[0031] S1. Collect physiological and psychological information from several subjects. Physiological information includes individual characteristics, electrocardiogram information, skin conductance information, respiratory rate and blood oxygen concentration. Individual characteristics include age and gender. Psychological information is obtained through psychological scales and questionnaires. After filtering, noise reduction and standardization preprocessing, a dataset is obtained.
[0032] S2. Introduce psychology experts to identify key factors that have a significant impact on the assessment of psychological stress.
[0033] S3. Extract key factors from the dataset, generate a key factor set, and construct a key factor association network. Methods for constructing the key factor association network include:
[0034] Using statistical methods, including but not limited to Pearson correlation coefficient and Spearman rank correlation coefficient, the correlation between key factors is calculated, and the edges and weights of the association network are constructed using the correlation values, where nodes represent key factors and edges represent the correlation between factors.
[0035] Topology analysis can be performed, and the methods include:
[0036] Obtain the degree, clustering coefficient, and path length of nodes in the topology of an interconnected network. The degree of a node is the number of edges directly connected to it, the clustering coefficient is an indicator of the tightness of the connection between the node and its neighbors, and the path length is the number of shortest edges that need to be traversed from one node to another in the interconnected network.
[0037] Obtain the weighted sum of degree, clustering coefficient and path length of each node in the interconnected network, and use it as the influence score of that node. Sort all nodes according to the influence score to obtain the influence of different nodes in the interconnected network.
[0038] S4. Utilize deep learning models to fuse physiological information, psychological information, and key factors. The deep learning models include, but are not limited to, convolutional neural networks and deep belief networks. The fusion methods include feature concatenation, additive fusion, and multiplicative fusion to achieve a unified multimodal feature representation.
[0039] S5. Construct a psychological stress state classifier based on multimodal features. After training and optimization, complete the construction of a multi-knowledge domain psychological stress state assessment framework.
[0040] S6. Input the patient's physiological and psychological information into the multi-knowledge domain psychological stress state assessment framework, and output the patient's psychological stress state assessment results.
[0041] In this embodiment, psychological stress status is assessed by combining physiological information, psychological information, and key factors proposed by a single psychology expert.
[0042] Example 2
[0043] This embodiment, as an improvement on the previous embodiment, is a psychological stress state assessment method based on multi-knowledge domain information, specifically including the following steps:
[0044] S1. Collect physiological and psychological information from several subjects. Physiological information includes individual characteristics, electrocardiogram information, skin conductance information, respiratory rate and blood oxygen concentration. Individual characteristics include age and gender. Psychological information is obtained through psychological scales and questionnaires. Psychological scales can quantify the psychological state of the subjects, while questionnaires can collect the subjects' subjective descriptions and feelings about their own psychological state.
[0045] The physiological and psychological information is preprocessed by filtering, denoising, and standardization to obtain the dataset.
[0046] S2. Introduce experts in psychology, medicine, biology, and information science to identify key factors that have a significant impact on the assessment of psychological stress, according to their respective fields. These key factors include, but are not limited to, physiological indicators such as heart rate variability, skin conductance, respiratory rate, and blood oxygen concentration, as well as psychological indicators such as anxiety, depression, and stress perception.
[0047] S3. Extract key factors from the dataset, generate a key factor set, and construct a key factor association network. Methods for constructing the key factor association network include:
[0048] Using statistical methods, including but not limited to Pearson correlation coefficient and Spearman rank correlation coefficient, the correlation between key factors is calculated, and the edges and weights of the association network are constructed using the correlation values, where nodes represent key factors and edges represent the correlation between factors.
[0049] Topology analysis can be performed, and the methods include:
[0050] Obtain the degree, clustering coefficient, and path length of nodes in the topology of an interconnected network. The degree of a node is the number of edges directly connected to it, the clustering coefficient is an indicator of the tightness of the connection between the node and its neighbors, and the path length is the number of shortest edges that need to be traversed from one node to another in the interconnected network.
[0051] Obtain the weighted sum of degree, clustering coefficient and path length of each node in the interconnected network, and use it as the influence score of that node. Sort all nodes according to the influence score to obtain the influence of different nodes in the interconnected network.
[0052] S4. Utilize deep learning models to fuse physiological information, psychological information, and key factors. The deep learning models include, but are not limited to, convolutional neural networks and deep belief networks. The fusion methods include feature concatenation, additive fusion, and multiplicative fusion to achieve a unified multimodal feature representation.
[0053] S5. Construct a psychological stress state classifier based on multimodal features, including but not limited to SVM classifier and Naive Bayes classifier. After training and optimization, complete the construction of a multi-knowledge domain psychological stress state assessment framework.
[0054] S6. Input the patient's physiological and psychological information into the multi-knowledge domain psychological stress state assessment framework, and output the patient's psychological stress state assessment results.
[0055] The advantage of Example 2 over Example 1 is that it combines physiological and psychological information, as well as key factors proposed by psychology experts, medical experts, biological experts and information science experts to conduct a comprehensive assessment of psychological stress state. The integration of multiple knowledge domains effectively improves the confidence of the assessment results.
[0056] As an extension, the classifier's output serves as an assessment of the patient's psychological stress state. This result can be a specific classification label, such as high stress, moderate stress, and low stress, or it can be a continuous score or probability value.
[0057] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.
[0058] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. A method for assessing psychological stress state based on multi-knowledge domain information, characterized in that: Specifically, the following steps are included: S1. Collect physiological and psychological information from several subjects, and obtain a dataset after preprocessing; S2. Introduce experts from multiple fields to formulate key factors according to different fields; S3. Extract key factors from the dataset, generate a key factor set, construct a key factor association network, and perform topology analysis. S4. Utilize deep learning models to fuse physiological information, psychological information, and key factors to achieve a unified multimodal feature representation; S5. Construct a psychological stress state classifier based on multimodal features, and after training and optimization, complete the construction of a multi-knowledge domain psychological stress state assessment framework. S6. Input the patient's physiological and psychological information into the multi-knowledge domain psychological stress state assessment framework, and output the patient's psychological stress state assessment results.
2. The method for assessing psychological stress state based on multi-knowledge domain information according to claim 1, characterized in that: The physiological information in S1 includes individual characteristic information, electrocardiogram information, skin conductance information, respiratory rate, and blood oxygen concentration; The individual characteristics information includes age and gender.
3. The method for assessing psychological stress state based on multi-knowledge domain information according to claim 1, characterized in that: The methods for obtaining psychological information in S1 include: Psychological information of the subjects was obtained through psychological scales and questionnaires.
4. The psychological stress state assessment method based on multi-knowledge domain information according to claim 1, characterized in that: The multidisciplinary experts in S2 include psychology experts, medical experts, biological experts, and information science experts.
5. The method for assessing psychological stress state based on multi-knowledge domain information according to claim 1, characterized in that: The key factors in S2 are those that have a significant impact on the assessment of psychological stress.
6. The method for assessing psychological stress state based on multi-knowledge domain information according to claim 1, characterized in that: The methods for constructing the key factor association network in S3 include: Statistical methods are used to calculate the correlation between key factors, and the correlation values are used to construct the edges and weights of the association network, where nodes represent key factors and edges represent the correlation between factors.
7. The method for assessing psychological stress state based on multi-knowledge domain information according to claim 1, characterized in that: The topology analysis methods in S3 include: Obtain the degree, clustering coefficient, and path length of nodes in the associated network topology. The degree of a node is the number of edges directly connected to it. The clustering coefficient is an indicator of the tightness of the connection between the node and its neighboring nodes. The path length is the number of shortest edges that need to be traversed from one node to another in the associated network. Obtain the weighted sum of degree, clustering coefficient and path length of each node in the interconnected network, and use it as the influence score of that node. Sort all nodes according to the influence score to obtain the influence of different nodes in the interconnected network.
8. The method for assessing psychological stress state based on multi-knowledge domain information according to claim 1, characterized in that: The methods for fusing physiological information, psychological information, and key factors in S4 include: feature splicing, additive fusion, and multiplicative fusion.