Double-end architecture system for mental health inquiry and auxiliary diagnosis

Abstract

The invention belongs to the technical field of psychological health informatization, and discloses a psychological health inquiry and auxiliary diagnosis double-end architecture system which realizes intelligent acquisition, data fusion analysis, auxiliary diagnosis and intervention management of psychological health problems through cooperative operation of a user end and a physician end. The double-end architecture of the system is used for giving consideration to the self-service inquiry experience of the user and the professional diagnosis requirement of the doctor; through multi-source data fusion of natural language, scale and cognitive test, the diagnosis accuracy is improved; structured medical records are automatically generated, intervention suggestions are intelligently output, and the burden of doctors is relieved; the system can realize continuous tracking, provides follow-up visit and effect evaluation functions, and supports long-term mental health management; the system provided by the invention meets the medical data privacy protection requirement and supports local deployment of hospitals.

Description

Technical Field

[0001] This invention relates to the field of mental health information technology, specifically to a dual-end architecture system for mental health consultation and auxiliary diagnosis. Background Technology

[0002] Current mental health assessments primarily rely on human consultations, paper or online scales, and physicians' clinical experience. These current assessment methods have the following shortcomings: Highly subjective: The results of the consultation are easily affected by the doctor's subjective judgment or the patient's unclear expression; Inefficient: Traditional assessments are time-consuming and the results are difficult to standardize; Data fragmentation: There is a lack of integration between patient consultation records, psychological scale results, and cognitive assessment data; Lack of continuity: There is a lack of digital tracking methods for subsequent interventions and follow-ups.

[0003] Therefore, there is an urgent need for a dual-end architecture mental health consultation and auxiliary diagnosis system that can provide users with natural language-friendly interaction, scale completion and cognitive testing support, and provide physicians with intelligent analysis and auxiliary diagnostic suggestions, thereby improving the efficiency and scientific nature of diagnosis and treatment. Summary of the Invention

[0004] The purpose of this invention is to address the aforementioned problems by providing a dual-end architecture system for mental health consultation and auxiliary diagnosis. Through the collaborative operation of the user end and the physician end, it enables intelligent collection of mental health issues, data fusion and analysis, evidence chain organization, auxiliary diagnosis, and intervention management.

[0005] The technical solution adopted in this invention is as follows: A dual-end architecture system for mental health consultation and auxiliary diagnosis, the system including a user end, a physician end and a management platform; The user terminal is used to acquire the user's psychological and physiological data, parse the data provided by the user, and make a preliminary judgment on the user's psychological assessment data based on the parsed data and feed it back to the user, so that the user can obtain preliminary diagnostic data. The physician obtains all data information from the user through the management platform, performs feature fusion and analysis on the psychological and physiological data through the management platform to obtain the user's psychological assessment results, and conducts a comprehensive evaluation of the user's psychological assessment results and psychological and physiological data through the management platform. Based on the evaluation results, an evaluation report is generated and fed back to the user through the management platform.

[0006] Furthermore, the user terminal includes: The natural language consultation module uses natural language processing technology to guide users to conduct psychological questions and answers in the form of voice or text, and to obtain users' psychological question and answer data. The psychological assessment module uses standardized psychological assessment scales to conduct psychological assessments on users. The cognitive function testing module is used to assess users' objective metrics. The physiological indicator data module is used to obtain the user's current physiological information; The image recognition module is used to recognize the content of previous inspection reports uploaded by users; The data upload module is used to upload user information to the management platform; The report viewing module allows users to access evaluation reports sent by the management platform.

[0007] Furthermore, the natural language diagnostic module includes a speech recognition module, an emotion recognition module, and a semantic understanding module. The natural language diagnostic module performs semantic analysis on the user's natural language input and extracts psychological state-related features, including emotional vocabulary, semantic patterns, speech rate and rhythm, and emotional tendencies.

[0008] Furthermore, the psychological scale module includes, but is not limited to, at least one of the following: Self-Rating Depression Scale (SDS), Self-Rating Anxiety Scale (SAS), Beck Depression Scale (BDI), Patient Health Questionnaire-9 (PHQ-9), and Generalized Anxiety Scale (GAD-7).

[0009] Furthermore, the physician terminal includes a multi-source data fusion analysis module and an intelligent diagnostic engine, used to extract semantic features, keyword features, and emotional features from the natural language consultation module, psychological scale module, cognitive function test module, image recognition module, and physiological indicator data module, respectively. The extracted feature vectors are input into a multimodal fusion model to obtain a comprehensive mental health feature vector. The comprehensive mental health feature vector is input into the intelligent diagnostic engine to obtain the assessment results.

[0010] Furthermore, the evaluation and analysis of the intelligent diagnostic engine specifically includes obtaining the probability of the assessed person's mental illness and constructing the following discriminant formula: (1) In the formula, SDS is the comprehensive discrimination score.

[0011] Furthermore, the comprehensive discrimination score SDS specifically includes: The comprehensive discriminant score SDS specifically includes: establishing a path-dependent consultation method based on the DSM-5 diagnostic criteria by training and fine-tuning the Qwen3 series models of the natural language consultation module and the Deepseek large model on their own datasets. The consultation content includes various existing types of mental illnesses. Weight ratios are set for the consultation content items, and the comprehensive discriminant score SDS is obtained through standardized calculation of the weight scores of the consultation content items.

[0012] Furthermore, the system also includes a report generation and export module. Physicians obtain the user's disease probability through the intelligent diagnostic engine and output a structured diagnostic report, including a medical history summary, scale results, physician diagnostic suggestions, and intervention plan, to the user's report viewing module through the management platform.

[0013] Furthermore, the physician's terminal also includes an intervention and follow-up module, used to set up personalized intervention plans, including psychotherapy, drug treatment recommendations, and family guidance; Set up follow-up tasks and provide periodic feedback to the user's device. The user's device will receive a reminder and complete the follow-up tasks.

[0014] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: The dual-end architecture of this invention balances the user's self-service consultation experience with the professional diagnostic needs of physicians; multi-source data fusion, through natural language processing, scales, cognitive tests, previous examination results, and physiological parameters, improves diagnostic accuracy; intelligent output allows the system to automatically generate structured medical records and intervention suggestions, reducing the burden on physicians; the system provides continuous tracking, offering follow-up and effect evaluation functions to support long-term mental health management; and the system meets medical data privacy protection requirements and supports local deployment in hospitals. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating the usage of a dual-end architecture system for mental health consultation and auxiliary diagnosis according to the present invention. Detailed Implementation

[0016] The present invention will now be described in detail with reference to the accompanying drawings.

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0018] Example This embodiment provides a dual-end architecture system for mental health consultation and auxiliary diagnosis. The dual-end architecture specifically includes: The user-facing interface is used by users with psychological needs to complete natural language consultations, fill out psychological scales, perform cognitive tests, upload past examination reports, and receive personalized feedback. The physician side is used by psychiatrists, psychotherapists, or counselors to view patient data, conduct auxiliary diagnoses, and develop intervention and follow-up plans.

[0019] Both ends interact with the cloud / local server through encrypted data transmission to ensure data security and privacy compliance.

[0020] In this embodiment, the client specifically includes the following: Natural Language Consultation Module: Based on natural language processing technology, this module guides users to conduct psychological Q&A sessions in the form of voice or text. Psychological scale module: Provides a variety of standardized psychological assessment scales; Cognitive function testing module: used to assess objective indicators such as attention, executive function, and memory; Image recognition module: Used to recognize the content of previous inspection reports uploaded by users; Physiological indicator data collection: After the user completes relevant examinations such as EEG, near-infrared brain functional imaging, eye movement data, and wearable portable devices, the system collects physiological indicator data.

[0021] Data upload module: Encrypts and transmits user consultation, scale, and test data to the server and management platform; Report viewing module: Users can view automatically generated preliminary mental health analysis and intervention suggestions.

[0022] The physician's side of this embodiment specifically includes the following: User management module: Supports user profile management, medical record recording, and follow-up tracking; Multi-source data fusion analysis module: performs comprehensive analysis of natural language features, scale scores, and cognitive test results; Intelligent diagnostic engine: Based on machine learning / deep learning models, it outputs auxiliary diagnostic results and disease risk probabilities. The specific model construction is as follows: The Comprehensive Discriminant Score (SDS) specifically includes: a path-dependent questioning method based on the DSM-5 diagnostic criteria, established through fine-tuning of proprietary datasets from the Qwen3 series of natural language processing models and the Deepseek model. The questioning content includes common mental illnesses such as depression, anxiety disorders, bipolar I, bipolar II, schizophrenia, attention deficit hyperactivity disorder (ADHD), autism spectrum disorder (ASD), academic difficulties, internet addiction, post-traumatic stress disorder (PTSD), and personality disorders. Weighting is applied to the items within the questioning content for these illnesses.

[0023] The SDS score of DSM_Score is obtained through standardized calculation of the weighted scores of the consultation content items, as follows: The standardized calculation method for the depressive episode discrimination score is as follows: Depression SDS = Sum of weighted scores of the number of items that meet the DSM-5 diagnostic criteria / 51; Strict adherence to the rule that symptoms must last for ≥14 days is required to calculate the depression SDS score; Item count weight score: Low mood, such as feeling sad, empty, or hopeless; children and adolescents may exhibit irritability = 22 points; A significant decrease in interest or pleasure, such as a decline in interest or enjoyment in almost all activities, = 22 points; Significantly insufficient energy or physical strength = 1 point; Sleep disorder (insomnia or hypersomnia) = 1 point; Psychomotor agitation or retardation (e.g., restlessness, slowed movements, reduced speech) = 1 point; Feelings of worthlessness or excessive / inappropriate guilt = 1 point; Difficulty concentrating and indecisiveness = 1 point; Repeated suicidal thoughts = 1 point; Significant changes in weight or appetite, such as a weight loss of ≥5% or a weight gain without dieting, = 1 point; If the duration of the illness is less than 14 days, the SDS for depression is 0, which means there are no depressive episodes. If the SDS of depression is ≤0.45, the diagnosis of depressive episode is not valid; The SDS is converted into a discriminant probability using the logistic function, as follows: (1) In the formula, SDS is the comprehensive discrimination score, which is used to determine the probability of disease occurrence.

[0024] Report generation and export module: Outputs structured diagnostic reports, including medical history summary, scale results, AI diagnostic suggestions, and intervention plans; Follow-up and intervention module: Supports multiple assessments and comparisons, personalized interventions, and effect tracking.

[0025] This embodiment of the system supports multi-terminal adaptation: it supports deployment in multiple forms such as Web, APP, and mini-program.

[0026] The implementation process of this embodiment is as follows: Figure 1 As shown, the specific process includes: The user-side process is as follows: Login and identity verification: Users log in to the system using their account, mobile phone number, student ID, etc.; single sign-on and facial recognition verification are supported to ensure data uniqueness and security.

[0027] The natural language consultation module is based on the DSM-5 / SCID-5 consultation standard and uses NLP natural language processing technology to achieve intelligent questioning. Users can answer questions through voice or text input, and the system automatically recognizes emotional words and keywords to generate preliminary symptom labels.

[0028] The psychological scale assessment module allows users to complete standardized psychological scales, such as PHQ-9, GAD-7, MMPI-A, SCARED, and CBCL, by following system prompts.

[0029] The system automatically scores and stores the results.

[0030] The cognitive function testing module includes attention tests, working memory tests, reaction time tests, task switching tests, etc. The system automatically records indicators such as reaction time, accuracy, and error rate.

[0031] The physiological indicator acquisition module includes EEG, near-infrared brain functional imaging, eye-tracking devices, and wearable portable vital signs devices.

[0032] Initial feedback and risk warnings: Based on user self-assessment and objective test results, the system generates a brief psychological report, providing "psychological state visualization + risk level + next steps suggestions".

[0033] The physician's workflow is as follows: Patient data reception and integration: The physician receives natural language records, psychological scale data and cognitive test data from the user terminal. The data is fused using a multi-source fusion algorithm (multimodal analysis) for unified modeling.

[0034] The intelligent diagnostic engine, also known as the AI-assisted diagnostic engine, calls a machine learning model and combines a knowledge graph with the DSM-5 standard to output auxiliary diagnostic results. The results are presented in the form of "probability distribution + diagnostic hints", for example: 72% probability of depressive disorder, 55% probability of generalized anxiety disorder, and 33% probability of attention deficit hyperactivity disorder.

[0035] A structured diagnostic report is generated, which includes the following sections: Basic information (gender, age, academic / work status, etc.); symptom presentation (natural language extraction and physician supplementation); scale results (raw scores compared with norms); cognitive function performance (reaction time, attention fluctuations, etc.); physiological indicator data (EEG waveform, near-infrared integral value / center of gravity value, eye movement data, vital signs, etc.); AI-assisted analysis conclusions; treatment and intervention recommendations; the report supports PDF export and can be stored in an electronic medical record system.

[0036] The intervention and follow-up module allows physicians to set up personalized intervention plans based on reports, including psychotherapy, drug treatment recommendations, and family guidance. The system supports setting follow-up tasks (such as monthly scale assessments and cognitive training tasks), and users will receive reminders to complete the follow-up content.

[0037] This article uses specific embodiments to illustrate the principles and implementation methods of the present invention. The descriptions of the embodiments above are only for the purpose of helping to understand the method and core ideas of the present invention. It should be noted that those skilled in the art can make several improvements and modifications to the present invention without departing from the principles of the present invention, and these improvements and modifications also fall within the protection scope of the claims of the present invention.

Claims

1. A dual-end architecture system for mental health consultation and auxiliary diagnosis, characterized in that, The system includes a user terminal, a physician terminal, and a management platform; The user terminal is used to acquire the user's psychological and physiological data, parse the data provided by the user, and make a preliminary judgment on the user's psychological assessment data based on the parsed data and feed it back to the user, so that the user can obtain preliminary diagnostic data. The physician obtains all data information from the user through the management platform, performs feature fusion and analysis on the psychological and physiological data through the management platform to obtain the user's psychological assessment results, and conducts a comprehensive evaluation of the user's psychological assessment results and psychological and physiological data through the management platform. Based on the evaluation results, an evaluation report is generated and fed back to the user through the management platform.

2. The dual-end architecture system for mental health consultation and auxiliary diagnosis according to claim 1, characterized in that, The user terminal includes: The natural language consultation module uses natural language processing technology to guide users to conduct psychological questions and answers in the form of voice or text, and to obtain users' psychological question and answer data. The psychological scale module uses standardized psychological assessment scales to conduct psychological assessments on users. The cognitive function testing module is used to assess users' objective metrics. The physiological indicator data module is used to obtain the user's current physiological information; The data upload module is used to upload user information to the management platform; The report viewing module allows users to access evaluation reports sent by the management platform.

3. The dual-end architecture system for mental health consultation and auxiliary diagnosis according to claim 2, characterized in that, The natural language diagnostic module includes a speech recognition module, an emotion recognition module, and a semantic understanding module. The natural language diagnostic module performs semantic analysis on the user's natural language input and extracts psychological state-related features, including emotional vocabulary, semantic patterns, speech rate and rhythm, and emotional tendencies.

4. The dual-end architecture system for mental health consultation and auxiliary diagnosis according to claim 2, characterized in that, The psychological scale module includes, but is not limited to, at least one of the following: Self-Rating Depression Scale (SDS), Self-Rating Anxiety Scale (SAS), Beck Depression Scale (BDI), Patient Health Questionnaire-9 (PHQ-9), and Generalized Anxiety Scale (GAD-7).

5. The dual-end architecture system for mental health consultation and auxiliary diagnosis according to claim 2, characterized in that, The physician terminal includes a multi-source data fusion and analysis module and an intelligent diagnostic engine, which are used to extract semantic features, keyword features and emotional features from the natural language consultation module, psychological scale module, cognitive function test module and physiological indicator data module, respectively. The extracted feature vectors are input into a multimodal fusion model to obtain a comprehensive mental health feature vector. The comprehensive mental health feature vector is input into the intelligent diagnostic engine to obtain the assessment results.

6. The dual-end architecture system for mental health consultation and auxiliary diagnosis according to claim 5, characterized in that, The evaluation and analysis of the intelligent diagnostic engine specifically includes obtaining the probability of mental illness in the person being evaluated and constructing the following discriminant formula: （1） In the formula, SDS is the comprehensive discrimination score.

7. A dual-end architecture system for mental health consultation and auxiliary diagnosis according to claim 6, characterized in that, The comprehensive discriminant score SDS specifically includes: The comprehensive discriminant score SDS specifically includes: establishing a path-dependent consultation method based on the DSM-5 diagnostic criteria by training and fine-tuning the Qwen3 series models of the natural language consultation module and the Deepseek large model on their own datasets. The consultation content includes various existing types of mental illnesses. Weight ratios are set for the consultation content items, and the comprehensive discriminant score SDS is obtained through standardized calculation of the weight scores of the consultation content items.

8. A dual-end architecture system for mental health consultation and auxiliary diagnosis according to claim 6, characterized in that, The system also includes a report generation and export module. Physicians obtain the user's disease probability through the intelligent diagnostic engine and output a structured diagnostic report through the management platform, including medical history summary, scale results, cognitive assessment results, physiological indicators, previous examination report content, physician diagnostic suggestions and intervention plans to the user's report viewing module.

9. A dual-end architecture system for mental health consultation and auxiliary diagnosis according to claim 1, characterized in that, The physician's terminal also includes an intervention and follow-up module, which is used to set up personalized intervention plans, including psychotherapy, drug treatment recommendations, physical therapy plans, and family guidance; Set up follow-up tasks and provide periodic feedback to the user's device. The user's device will receive a reminder and complete the follow-up tasks.

Images