A mental health intelligent evaluation system

Abstract

The application is suitable for the technical field of mental health evaluation and intelligent service, and provides a mental health intelligent evaluation system, which realizes double upgrading from passive evaluation to active protection, from tool empowerment to ecological win-win through multi-dimensional innovation to build a whole-cycle mental health service system, specifically including: a technology fusion module, a service mode module, a scene adaptation module and a data security module; through multi-dimensional optimization of technology fusion, service innovation, scene adaptation and security protection, the system improves the accuracy and efficiency of mental health evaluation, builds a whole-cycle mental health service closed loop, adapts to multi-scene demand, at the same time guarantees the privacy and security of user data, and promotes the development of mental health service towards intelligence, universal access and precision.

Description

Technical Field

[0001] This invention relates to the field of mental health assessment and intelligent service technology, and more specifically, to a mental health intelligent assessment system. Background Technology

[0002] With increasing societal pressures, mental health issues are receiving growing attention. Mental health assessments, as a core means of early screening and risk warning, are widely used in education, business, and healthcare settings.

[0003] Existing mental health assessment systems suffer from the following shortcomings: First, they rely on a single assessment dimension, depending solely on subjective questionnaire responses, making them susceptible to user cognitive biases and concealment, resulting in insufficient assessment accuracy. Second, assessment and intervention are disconnected, only providing basic score reports and failing to offer personalized, tiered intervention services, hindering full-cycle management. Third, they lack adaptability to different scenarios, with general-purpose assessment tools unable to meet the segmented needs of different groups such as adolescents, the elderly, and working professionals. Fourth, data privacy and security are inadequate; mental health data is highly sensitive, and existing systems lack end-to-end encryption and self-control mechanisms, easily leading to privacy leaks. Fifth, it is difficult to balance assessment efficiency and accuracy; fixed question bank models either consume too much time, affecting user completion rates, or the questions are too simple, leading to inaccurate assessments.

[0004] Therefore, there is an urgent need for an intelligent mental health assessment system that integrates multimodal data, AI adaptive technology, full-cycle services, and high security protection to overcome the pain points of existing technologies and achieve an upgrade from passive assessment to proactive protection, and from tool empowerment to ecosystem win-win. Summary of the Invention

[0005] To address the shortcomings of existing technologies, the present invention aims to provide an intelligent mental health assessment system. Through multi-dimensional optimization of technology integration, service innovation, scenario adaptation, and security protection, the system improves the accuracy and efficiency of mental health assessment, constructs a closed loop for full-cycle mental health services, adapts to diverse scenario needs, and protects user data privacy and security, thereby promoting the development of mental health services towards intelligence, universal access, and precision.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A smart mental health assessment system constructs a full-cycle mental health service system through multi-dimensional innovation, achieving a dual upgrade from passive assessment to proactive protection, and from tool empowerment to ecosystem win-win. Specifically, it includes: a technology integration module to break the boundaries of traditional assessments and improve the accuracy and efficiency of evaluations; a service model module to build a closed loop of full-cycle mental health services and achieve seamless integration of assessment and intervention; a scenario adaptation module to expand diverse application scenarios and reconstruct industry value; and a data security module to ensure the privacy, security, and compliant use of sensitive mental health data.

[0007] The present invention is further configured such that: the technology fusion module includes a multimodal data fusion evaluation unit, a dynamic adaptive AI evaluation engine unit, and a privacy and security encryption unit, wherein: the multimodal data fusion evaluation unit is configured to construct a cross-validation system of "subjective questionnaire + objective multimodal data", integrating multi-dimensional data of physiological signals, behavioral features, facial micro-expressions, and voice emotions; the physiological signals are collected by connecting to wearable devices to collect heart rate variability, sleep structure, and exercise frequency; the behavioral features are analyzed using NLP technology to analyze user semantic content and mobile terminal sensors to capture operating habits; the facial micro-expressions and voice emotions are captured and quantified by authorized cameras and voice acquisition devices to solve the problems of user implicit emotion recognition and subjective bias; the dynamic adaptive AI evaluation engine unit is equipped with... To build an adaptive assessment mechanism of "one thousand questions for one thousand people, dynamic iteration," it intelligently focuses on highly relevant questions and skips irrelevant options based on users' previous answers. It dynamically updates the norm database segmented by age, occupation, and region. For users with high-risk tendencies, it adds in-depth probing questions and traces the source of ambiguous answers, balancing assessment efficiency and accuracy. The privacy and security encryption unit is configured to build a four-in-one security system of "end-to-end encryption + hierarchical access control + privacy self-management + compliance audit." It uses national cryptographic algorithms to achieve end-to-end encryption of data collection, transmission, and storage. It supports local priority storage of sensitive data, and users can independently set the data retention time and authorization scope. Group data analysis adopts triple processing of "virtual ID + data desensitization + hash encryption," and at the same time, it establishes a full-process operation audit log.

[0008] The present invention is further configured such that the service mode module includes a hierarchical intervention unit, a full-cycle dynamic tracking unit, and a cross-scenario linkage service unit, wherein: the hierarchical intervention unit is configured to construct a four-level service system of "self-help intervention - light consultation - professional diagnosis and treatment - crisis intervention" based on the risk level assessment. Low-risk users are pushed with AI-customized self-help toolkits and smart devices are linked to provide feedback on the intervention effect; medium-risk users are provided with online light consultation and group psychological courses; high-risk users trigger a two-level emergency intervention mechanism, connecting with medical institutions to open a green channel for medical treatment and arranging a full-time psychological crisis interventionist to follow up within 24 hours; the full-cycle dynamic tracking unit is configured to construct a user full-cycle dynamic tracking unit. The lifecycle mental health growth record establishes a closed-loop management system encompassing assessment, intervention, retesting, optimization, and data retention. It sets personalized retesting cycles based on user risk levels and proactively pushes reminders using daily behavioral data. It generates mental health trend maps and dynamically optimizes intervention plans and quantifies effectiveness indicators based on retest results, intervention implementation data, and objective behavioral data. The cross-scenario collaborative service unit is configured to build a four-party collaborative mechanism involving individuals, institutions, medical institutions, and non-profit organizations. This breaks down data barriers to achieve resource complementarity and deeply integrates with school academic affairs systems, enterprise OA systems, and medical institution HIS and EMR systems, enabling multi-scenario service collaboration and data interoperability.

[0009] The present invention is further configured such that the scenario adaptation module includes a segmented scenario customization unit, a data value mining unit, and a low-threshold inclusive service unit. Specifically: the segmented scenario customization unit is configured to create customized solutions for different groups such as teenagers, professionals, and the elderly. For teenagers, it adds parent-child relationship assessments, academic pressure segmentation dimensions, and a dedicated interactive interface; for professionals, it develops segmented assessments of burnout and workplace interpersonal relationships and provides a white paper on enterprise management; for the elderly, it adapts to the operating habits of elderly users, adds loneliness assessments and cognitive function screening, and supports remote authorization for children to view the data. The data value mining unit is configured to build a big data analysis platform based on massive amounts of anonymized assessment data, providing quantitative decision-making basis for institutions, collaborating with universities and research institutions to open anonymized data to support industry research, constructing an early warning model for abnormal mental health trends, and providing accurate early warning and prevention suggestions for public health departments. The low-threshold inclusive service unit is configured to support lightweight use across multiple terminals such as mini-programs, web pages, and apps, adopting a tiered pricing strategy of "free basic services + paid value-added services," providing free customized services to public welfare organizations and grassroots schools, and reducing the professional usage threshold and labor costs by generating standardized report interpretations, intervention suggestions, and communication templates through AI.

[0010] The present invention is further configured such that: the multimodal data fusion evaluation unit supports multilingual adaptation, can cover diverse user scenarios, facial micro-expression recognition focuses on features such as frowning, drooping corners of the mouth, and wandering eyes, and voice emotion analysis quantifies the intensity of emotions based on speech rate, tone, and pause frequency, which is suitable for groups such as teenagers and working professionals who are unwilling to express themselves actively.

[0011] The present invention is further configured such that the dynamic adaptive AI assessment engine unit can shorten the traditional 10-20 minute assessment time to 5-8 minutes, improve the user completion rate, and at the same time eliminate the interference of wrong answers by tracing the source of wrong questions, so as to avoid the inaccuracy of assessment caused by a single answer deviation.

[0012] The present invention is further configured such that: the privacy and security encryption unit complies with the Personal Information Protection Law and the mental health service specifications, supports users to delete all assessment records with one click, and records data access, modification and export behaviors in the entire process operation audit log to ensure that data use is traceable and adapts to the compliance requirements of industries such as medical care and education.

[0013] The present invention is further configured such that the AI-customized self-service toolkit pushed by the hierarchical intervention unit can adjust the content according to the user's schedule and preferences, including nighttime sleep-aid meditation, exercise-based mood regulation programs, mindfulness drawing, music therapy, etc., dynamically optimizing the frequency and content of reminders, and cultivating the user's ability to manage emotions independently.

[0014] In the medical setting, the cross-scenario linkage service unit can synchronize assessment reports to the doctor's workstation, and discharged patients can complete follow-up assessments through the system. It also links with community health service centers to achieve seamless integration of "in-hospital treatment + out-of-hospital follow-up" and reduce the recurrence rate.

[0015] The early warning model constructed by the data value mining unit can monitor group data in the region in real time and promptly detect abnormal changes in mental health risks in specific groups and regions, including a sharp rise in the incidence of adolescent depression and abnormally high rates of occupational burnout in industries, thus providing support for precise prevention and control.

[0016] The advantages of this invention are: 1. This invention effectively eliminates subjective bias and the limitations of single data by using multimodal data fusion and weighting algorithms combined with an AI adaptive assessment engine. The assessment accuracy is improved by more than 40% compared with traditional questionnaire assessments, while the assessment time is shortened to 5-8 minutes, balancing efficiency and accuracy.

[0017] 2. This invention constructs a four-level hierarchical intervention system and a full-cycle dynamic tracking mechanism. It optimizes the intervention plan in real time through an effect quantification algorithm, achieving seamless connection from assessment to intervention and follow-up, and solving the pain point of traditional systems that "only assess but do not intervene".

[0018] 3. This invention creates customized solutions for specific scenarios such as teenagers, workplaces, and the elderly, adapting to the core needs of different groups and expanding the application scope of the system.

[0019] 4. This invention achieves end-to-end data protection, autonomous control, and compliance auditing through a four-in-one security system and encryption algorithm, thereby mitigating privacy risks and enhancing user trust.

[0020] 5. This invention provides support for institutional decision-making, industry research, and public health prevention and control through a data value mining unit, realizing a value upgrade from personal services to industry empowerment. Attached Figure Description

[0021] Figure 1 This is a block diagram of a mental health intelligent assessment system according to the present invention.

[0022] Figure 2 This is a block diagram of the technical integration module of the present invention.

[0023] Figure 3 This is a block diagram of the service mode module composition of the present invention.

[0024] Figure 4 This is a block diagram of the scenario adaptation module of the present invention. Detailed Implementation

[0025] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0026] It should be noted that, unless otherwise specified, all technical and scientific terms used in this application have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0027] In this invention, unless otherwise stated, the directional terms such as "up" and "down" generally refer to the directions shown in the accompanying drawings, or to the vertical, perpendicular, or gravitational direction; similarly, for ease of understanding and description, "left" and "right" generally refer to the left and right shown in the accompanying drawings; "inner" and "outer" refer to the inner and outer contours of each component itself, but the above directional terms are not intended to limit this invention.

[0028] Example 1, please refer to Figure 1-4 The present invention provides the following technical solutions: A mental health intelligent assessment system, specifically, constructs a full-cycle mental health service system through multi-dimensional innovation, achieving a dual upgrade from passive assessment to proactive protection, and from tool empowerment to ecosystem win-win, specifically including: The technology integration module is used to break down the boundaries of traditional assessments and improve the accuracy and efficiency of evaluations; The service model module is used to build a closed loop of full-cycle mental health services, achieving seamless integration of assessment and intervention. The scenario adaptation module is used to expand diverse application scenarios and reconstruct industry value; The data security module is used to protect the privacy and compliant use of sensitive mental health data.

[0029] The data security module works in conjunction with the privacy and security encryption unit in the technology integration module to implement autonomous data privacy management and compliance audit requirements. It supports users to delete assessment records with one click, establishes a full-process operation audit log, records data access, modification, and export behaviors, ensures traceability of data use, and complies with the Personal Information Protection Law and mental health service standards.

[0030] The technology fusion module includes a multimodal data fusion evaluation unit, a dynamic adaptive AI evaluation engine unit, and a privacy and security encryption unit, wherein: The multimodal data fusion assessment unit is configured to construct a cross-validation system of "subjective questionnaire + objective multimodal data", integrating multi-dimensional data such as physiological signals, behavioral characteristics, facial micro-expressions and voice emotions.

[0031] Data fusion algorithms are used to eliminate the limitations of single data points and improve assessment accuracy. The subjective questionnaire combines standardized scales (such as the SDS Depression Scale and the SAS Anxiety Scale) with customized scales to obtain users' self-perceived mental health status. Objective data collection and processing methods are as follows: Physiological signals: Connect to wearable devices (smartwatches, wristbands) to collect data such as heart rate variability (HRV), sleep structure (sleep onset time, percentage of deep sleep), and exercise frequency. Use filtering algorithms to remove noise interference and extract effective feature values.

[0032] Behavioral characteristics: NLP technology is used to analyze users' emotional diaries and open-ended responses to extract emotional keywords and semantic tendencies; mobile sensors are used to capture behavioral data such as typing rhythm (standard deviation of key interval), screen operation frequency (number of page switches / minute), and dwell time.

[0033] Facial micro-expressions and voice emotions: After authorization, micro-expression features such as frowning, drooping corners of the mouth, and wandering eyes are captured by the camera, and voice features such as speech rate, tone, and pause frequency are captured by the voice acquisition device. These features are then converted into calculable emotion values ​​through feature quantification algorithms.

[0034] To achieve effective fusion of multimodal data, this unit uses a weighted fusion algorithm to calculate the comprehensive sentiment assessment value, as shown in the following formula:

[0035] in: E is the comprehensive emotion assessment value, ranging from [0,10]. A higher value indicates a more significant negative emotion. ω is the weight coefficient of the i-th type of data, satisfying Σω=1, where the weight of subjective questionnaire is 0.4, the weight of physiological signal is 0.25, the weight of behavioral feature is 0.15, and the weights of facial micro-expression and voice emotion are each 0.1 (which can be dynamically adjusted according to the scenario). E is the standardized sentiment assessment value of the i-th data, with a value range of [0,10], obtained through the feature quantization algorithm of the corresponding dimension.

[0036] The physiological signals are collected via a wearable device, including heart rate variability, sleep structure, and exercise frequency. The behavioral characteristics are analyzed using NLP technology to examine user semantic content and mobile sensor capture of operating habits. Facial micro-expressions and vocal emotions are captured and quantified using authorized cameras and voice acquisition devices to address issues of implicit emotion recognition and subjective bias. The dynamic adaptive AI assessment engine unit is configured to build an adaptive assessment mechanism of "one thousand questions for one thousand people, dynamic iteration." Based on the user's previous answers, it intelligently focuses on highly relevant questions and skips irrelevant options, dynamically updates the norm database segmented by age, occupation, and region, adds in-depth probing questions for users with high-risk tendencies, and traces and questions the source of ambiguous answers to balance assessment efficiency and accuracy. The privacy and security encryption unit is configured to build a four-in-one security system of "end-to-end encryption + hierarchical permission + privacy self-management + compliance audit." It uses national cryptographic algorithms to achieve end-to-end encryption of data collection, transmission, and storage, supports local priority storage of sensitive data, allows users to set data retention time and authorization scope, and uses "virtual ID + data desensitization + hash encryption" for group data analysis, while establishing a full-process operation audit log.

[0037] This unit constructs an adaptive assessment mechanism of "one thousand questions for one thousand people, dynamic iteration," which adjusts subsequent questions based on users' previous answers and normative data through AI algorithms to balance assessment efficiency and accuracy. The core implementation logic is as follows: Question adaptation logic: Based on the user's answers to the first 5 questions, calculate the preliminary scores for each psychological health dimension, filter highly relevant questions through a question relevance algorithm, and skip questions that are not significantly related to the preliminary scores to shorten the assessment time.

[0038] Normative data updates: A rolling update algorithm is used to incorporate new user assessment data in real time, and the norm database is optimized by age, occupation, and region. The formula is as follows:

[0039] in: This is the updated norm mean; The mean of the norm before the update; This represents the average of the newly added user evaluation data; α is a smoothing coefficient, with a value of 0.8-0.9, to ensure the stability and timeliness of the norm data.

[0040] For users with high risk tolerance, supplementary assessments are conducted through in-depth question probing. Simultaneously, for ambiguous answers (such as selecting "uncertain" for three consecutive questions), further investigation is performed to trace the source of the errors. An error tracing algorithm is used to eliminate interference from incorrect answers, as shown in the following formula:

[0041] in: C is the credibility coefficient of the answer, with a value range of [0,1]. When the value is lower than 0.6, it triggers further questioning. The number of questions with identical answers; This represents the total number of questions answered. β is the influence coefficient of fuzzy responses, with a value of 0.1; The number of questions to be answered is ambiguous.

[0042] The service model module includes a tiered intervention unit, a full-cycle dynamic tracking unit, and a cross-scenario linkage service unit. The tiered intervention unit is configured to build a four-level service system based on risk assessment: self-help intervention, light consultation, professional diagnosis and treatment, and crisis intervention. Low-risk users receive AI-customized self-help toolkits and smart devices provide feedback on intervention effectiveness. Medium-risk users receive online light consultation and group psychological courses. High-risk users trigger a two-level emergency intervention mechanism, connecting with medical institutions to open green channels for treatment and arranging dedicated psychological crisis intervention specialists for 24-hour follow-up. The full-cycle dynamic tracking unit is configured to build a user's full life-cycle psychological... The system manages health growth records, establishing a closed-loop management system encompassing assessment, intervention, retesting, optimization, and data retention. It sets personalized retesting cycles based on user risk levels and proactively pushes reminders using daily behavioral data. It generates mental health trend maps and dynamically optimizes intervention plans and quantifies effectiveness indicators based on retest results, intervention execution data, and objective behavioral data. The system also includes a cross-scenario collaborative service unit, configured to build a four-party collaborative mechanism involving individuals, institutions, medical institutions, and non-profit organizations. This breaks down data barriers to achieve resource complementarity and deeply integrates with school academic affairs systems, enterprise OA systems, and medical institution HIS and EMR systems, enabling multi-scenario service collaboration and data interoperability.

[0043] The full-cycle dynamic tracking unit constructs a user's full-lifecycle mental health growth profile, establishing a closed loop of "assessment-intervention-retesting-optimization-retention," and setting personalized retesting cycles based on risk levels. The cycle calculation formula is as follows:

[0044] Where: T is the retesting period (unit: days), and R is the risk level assessment value. To round down, the retesting period is 30-90 days for low-risk users, 10-30 days for medium-risk users, and 7-10 days for high-risk users. A mental health trend map is generated simultaneously, and the intervention plan is dynamically optimized based on the retest results. The effectiveness of the intervention is evaluated using an effectiveness quantification algorithm.

[0045] Where: E is the intervention effectiveness rate, R is the risk assessment value before intervention, and R is the risk assessment value after intervention. An effectiveness rate of ≥30% indicates that the intervention is effective, and the intervention plan should be adjusted if the effectiveness rate is <10%.

[0046] The scenario adaptation module includes a segmented scenario customization unit, a data value mining unit, and a low-threshold inclusive service unit. Specifically: the segmented scenario customization unit is configured to create customized solutions for different groups such as teenagers, working professionals, and the elderly. For teenagers, it adds parent-child relationship assessments, academic pressure subdivisions, and a dedicated interactive interface; for the working professional scenario, it develops subdivisions for burnout and workplace interpersonal relationship assessments and provides a white paper on enterprise management; for the elderly scenario, it adapts to the operating habits of elderly users, adds loneliness assessments, cognitive function screening, and supports remote authorization for children to view the data. The data value mining unit is configured to build a big data analysis platform based on massive amounts of anonymized assessment data, providing quantitative decision-making basis for institutions, collaborating with universities and research institutions to open anonymized data to support industry research, constructing a mental health abnormality trend early warning model, and providing accurate early warning and prevention suggestions for public health departments. The low-threshold inclusive service unit is configured to support lightweight use across multiple terminals such as mini-programs, web pages, and apps, adopting a tiered pricing strategy of "free basic services + paid value-added services," providing free customized services to public welfare organizations and grassroots schools, and using AI to generate standardized report interpretations, intervention suggestions, and communication templates to reduce the professional usage threshold and labor costs.

[0047] The scenario adaptation module is used to expand into diverse application scenarios, including a segmented scenario customization unit, a data value mining unit, and a low-threshold inclusive service unit. It creates customized solutions for different scenarios such as teenagers, workplaces, and the elderly, mining value from massive amounts of data while lowering the barriers to use and achieving inclusive services.

[0048] The multimodal data fusion evaluation unit supports multilingual adaptation and can cover diverse user scenarios. Facial micro-expression recognition focuses on features such as frowning, drooping corners of the mouth, and wandering eyes. Voice emotion analysis quantifies emotion intensity based on speech rate, tone, and pause frequency, making it suitable for groups such as teenagers and working professionals who are unwilling to express themselves actively.

[0049] The dynamic adaptive AI assessment engine unit can shorten the traditional 10-20 minute assessment time to 5-8 minutes, improve the user completion rate, and at the same time eliminate the interference of incorrect answers by tracing the source of wrong questions, avoiding inaccurate assessment caused by a single answer deviation.

[0050] The privacy and security encryption unit complies with the Personal Information Protection Law and mental health service standards. It supports users to delete all assessment records with one click. The entire process operation audit log records data access, modification, and export behavior to ensure data use is traceable and meets the compliance requirements of industries such as medical care and education.

[0051] The privacy and security encryption unit constructs a four-in-one security system of "end-to-end encryption + hierarchical access control + self-management of privacy + compliance auditing" to ensure data privacy and compliant use. End-to-end encryption: The data is encrypted using the Chinese national cryptographic algorithm (SM4). The encryption formula is as follows:

[0052] Where: C is the encrypted data, K is the key, and P is the original data. The key is updated periodically through a dynamic generation algorithm to ensure encryption security.

[0053] Group data anonymization: A triple process of "virtual ID + data anonymization + hash encryption" is employed. The hash encryption formula is as follows:

[0054] Where: ID is the user's real identity identifier, and Salt is a random salt value, ensuring that the same ID is unique and cannot be reversed after hashing.

[0055] The AI-customized self-service toolkit pushed by the hierarchical intervention unit can adjust the content according to the user's schedule and preferences, including nighttime sleep-aid meditation, exercise-based mood regulation programs, mindfulness drawing, music therapy, etc., dynamically optimizing the frequency and content of reminders and cultivating the user's ability to manage emotions independently.

[0056] In the medical setting, the cross-scenario linkage service unit can synchronize assessment reports to the doctor's workstation, and discharged patients can complete follow-up assessments through the system. It also links with community health service centers to achieve seamless integration of "in-hospital treatment + out-of-hospital follow-up" and reduce the recurrence rate.

[0057] The cross-scenario collaborative service unit constructs a four-party collaborative mechanism involving individuals, institutions, medical institutions, and public welfare organizations, breaking down data barriers and deeply integrating with school academic affairs systems, enterprise OA systems, and medical institution HIS and EMR systems. Data interoperability is achieved through data interface protocols, and encrypted transmission is used to ensure data security, while also enabling multi-scenario service collaboration.

[0058] The early warning model constructed by the data value mining unit can monitor group data in the region in real time and promptly detect abnormal changes in mental health risks in specific groups and regions, including a sharp rise in the incidence of adolescent depression and abnormally high rates of occupational burnout in industries, thus providing support for precise prevention and control.

[0059] The working principle of this embodiment is as follows:

[0060] This system adopts a collaborative deployment architecture of "cloud-device-edge": the cloud deploys a big data analysis platform, a normative database, and core algorithm models, which are responsible for data storage, computing, and service scheduling; the device side includes mini-programs, web pages, apps, and wearable devices, providing user interaction and data collection entry points; the edge side deploys local encryption and data processing modules to achieve local priority storage and real-time processing of sensitive data, reducing transmission pressure and privacy risks.

[0061] The specific implementation process is as follows: 1. Customer registration and permission activation Users register accounts through the client-side entry point and select their user type (individual / organization / professional). The system assigns corresponding permissions based on the user type: individual users gain access to assessment, report viewing, and intervention plan execution; organization administrators gain access to batch task creation, group report viewing, and data export; and professionals gain access to report interpretation, customized intervention plans, and user follow-up. Users also complete privacy authorization settings, choosing the data retention period and authorization scope.

[0062] 2. Multimodal data acquisition and evaluation After a user initiates an assessment request, the system pushes initial questionnaire questions and, upon authorization, collects physiological signals, behavioral characteristics, facial micro-expressions, and voice emotion data: wearable devices upload physiological data in real time, mobile sensors capture behavioral data, and cameras and microphones capture micro-expressions and voice data. The dynamic adaptive AI assessment engine adjusts subsequent questions based on the user's answers to the first 5 questions, triggers deep probing questions for users with high-risk tendencies, and conducts follow-up questions for ambiguous answers.

[0063] 3. Data processing and risk assessment The edge processing unit preprocesses the collected multimodal data (filtering, denoising, feature extraction), transmits it to the cloud, calculates the comprehensive sentiment assessment value through a weighted fusion algorithm, combines it with standardized scale scores, determines the risk level through a risk assessment algorithm, and generates personalized assessment reports and group analysis reports (exclusively for institutional users).

[0064] 4. Stratified intervention and dynamic tracking The system pushes corresponding intervention services based on risk level: low-risk users receive AI-customized self-help toolkits (such as nighttime sleep-aid meditation and exercise-based mood regulation programs), with wearable devices providing feedback on intervention effectiveness; medium-risk users are matched with online light consultation services and group courses, with regular follow-up reminders; high-risk users trigger an emergency intervention mechanism, connecting them with a green channel for medical institutions and arranging for intervention specialists to follow up within 24 hours. The system pushes follow-up reminders according to personalized follow-up cycles, and evaluates the intervention effect based on the follow-up results using an effectiveness quantification algorithm, dynamically optimizing the intervention plan.

[0065] 5. Data security and compliance management Data is encrypted during transmission and storage using national cryptographic algorithms. Data is anonymized during group data analysis, and a full-process operation audit log records data access and modification activities in real time. Users can adjust privacy settings and delete evaluation records with a single click through their personal center, ensuring self-management of data privacy.

[0066] Specific implementation examples are as follows: Implementation in youth scenarios A middle school deployed this system, customizing assessment programs for its student population. New features included parent-child relationship assessments, detailed dimensions of academic stress, and a cartoon-style interactive interface specifically for teenagers. The system integrates with the school's academic affairs system to synchronize student academic data and, combined with multimodal assessment data, generates student mental health development profiles and group reports.

[0067] For low-risk students, we provide parent-child communication skills and academic stress management tools; for medium-risk students, we arrange for school counselors to provide mild counseling; for high-risk students, we work with local mental health centers to open green channels for medical treatment, and at the same time, we work with parents to develop family intervention plans to achieve collaborative protection among families, schools and medical institutions.

[0068] Workplace scenario implementation A company introduced this system to conduct annual mental health checkups for its employees, customized assessments of job burnout and workplace interpersonal relationships, and integrated the system with the company's OA system. By combining employee attendance and performance data, the system analyzes the correlation between job burnout and work intensity and provides the company with suggestions for optimizing human resource management (such as adjusting shift schedules and adding team building activities).

[0069] For low-risk employees, workplace emotion management tools are provided; for medium- and high-risk employees, online psychological counseling services are offered, and a white paper on corporate employee mental health management is generated to help companies improve employee well-being and work efficiency.

[0070] Obviously, the embodiments described above are merely some, not all, embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort should fall within the scope of protection of the present invention.

[0071] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this application. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0072] It should be noted that the terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in sequences other than those illustrated or described herein.

[0073] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

[0074] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A mental health intelligent assessment system, characterized in that: By constructing a full-cycle mental health service system through multi-dimensional innovation, we achieve a dual upgrade from passive assessment to proactive protection, and from tool empowerment to ecosystem win-win, specifically including: The technology integration module is used to break down the boundaries of traditional assessments and improve the accuracy and efficiency of evaluations; The service model module is used to build a closed loop of full-cycle mental health services, achieving seamless integration of assessment and intervention. The scenario adaptation module is used to expand diverse application scenarios and reconstruct industry value; The data security module is used to protect the privacy and compliant use of sensitive mental health data.

2. The intelligent psychological health assessment system according to claim 1, characterized in that: The technology fusion module includes a multimodal data fusion evaluation unit, a dynamic adaptive AI evaluation engine unit, and a privacy and security encryption unit, wherein: The multimodal data fusion assessment unit is configured to construct a cross-validation system of "subjective questionnaire + objective multimodal data", integrating multi-dimensional data such as physiological signals, behavioral characteristics, facial micro-expressions and voice emotions; The physiological signals are collected via a wearable device, including heart rate variability, sleep structure, and exercise frequency. The behavioral characteristics are analyzed using NLP technology to examine user semantic content and mobile sensor capture of operating habits. The facial micro-expressions and voice emotions are captured and quantified by the authorized camera and voice acquisition device to solve the problems of implicit emotion recognition and subjective bias. The dynamic adaptive AI assessment engine unit is configured to build an adaptive assessment mechanism of "thousands of questions for thousands of people and dynamic iteration". Based on the user's previous answers, it intelligently focuses on highly relevant questions and skips irrelevant options. It dynamically updates the norm database segmented by age, occupation and region. For users with high risk tendencies, it adds in-depth probing questions and traces the source of ambiguous answers, balancing assessment efficiency and accuracy. The privacy and security encryption unit is configured to build a four-in-one security system of "end-to-end encryption + hierarchical access control + privacy self-management + compliance audit". It adopts national cryptographic algorithms to achieve end-to-end encryption of data collection, transmission and storage, supports local priority storage of sensitive data, and allows users to set the data retention time and authorization scope. Group data analysis adopts triple processing of "virtual ID + data desensitization + hash encryption" and establishes a full-process operation audit log.

3. The intelligent psychological health assessment system according to claim 1, characterized in that: The service mode module includes a hierarchical intervention unit, a full-cycle dynamic tracking unit, and a cross-scenario linkage service unit, wherein: The tiered intervention unit is configured to build a four-level service system based on the risk level assessment: "self-help intervention - light consultation - professional diagnosis and treatment - crisis intervention". Low-risk users are pushed AI-customized self-help toolkits and smart devices are linked to provide feedback on the intervention effect. Medium-risk users are provided with online light consultation and group psychological courses. High-risk users trigger a two-level emergency intervention mechanism, which connects with medical institutions to open a green channel for medical treatment and arranges a full-time psychological crisis interventionist to follow up within 24 hours. The full-cycle dynamic tracking unit is configured to build a user's full life cycle mental health growth profile, connect the "assessment-intervention-retesting-optimization-retention" closed-loop management, set personalized retesting cycles according to the user's risk level and actively push reminders based on daily behavior data, generate mental health trend maps, and dynamically optimize intervention plans and quantify effect indicators based on retest results, intervention execution data and objective behavior data; The cross-scenario collaborative service unit is configured to build a four-party collaborative mechanism of "individual-institution-medical institution-public welfare organization", break down data barriers to achieve resource complementarity, and deeply connect with school academic affairs system, enterprise OA system, medical institution HIS and EMR system to achieve multi-scenario service collaboration and data interoperability.

4. The intelligent psychological health assessment system according to claim 1, characterized in that: The scenario adaptation module includes a segmented scenario customization unit, a data value mining unit, and a low-threshold inclusive service unit, wherein: The segmented scenario customization unit is configured to create customized solutions for different groups such as teenagers, working professionals, and the elderly. The teenager scenario adds parent-child relationship assessment, academic pressure subdivision and exclusive interactive interface; the working professional scenario develops occupational burnout and workplace interpersonal relationship subdivision assessment and provides enterprise management white paper; the elderly scenario is adapted to the operating habits of elderly users, adds loneliness sense assessment, cognitive function screening and supports remote authorization viewing by children. The data value mining unit is configured to build a big data analysis platform based on massive amounts of desensitized assessment data, provide quantitative decision-making basis for institutions, cooperate with universities and research institutions to open up desensitized data to support industry research, build an early warning model for abnormal mental health trends, and provide accurate early warning and prevention and control suggestions for public health departments; The low-threshold, inclusive service unit is configured to support lightweight use across multiple terminals, including mini-programs, web pages, and apps. It adopts a tiered pricing strategy of "free basic services + paid value-added services" to provide free customized services to non-profit organizations and grassroots schools. It uses AI to generate standardized report interpretations, intervention suggestions, and communication templates, thereby reducing the professional usage threshold and labor costs.

5. The intelligent psychological health assessment system according to claim 2, characterized in that: The multimodal data fusion evaluation unit supports multilingual adaptation and can cover diverse user scenarios. Facial micro-expression recognition focuses on features such as frowning, drooping corners of the mouth, and wandering eyes. Voice emotion analysis quantifies emotion intensity based on speech rate, tone, and pause frequency, making it suitable for groups such as teenagers and working professionals who are unwilling to express themselves actively.

6. The intelligent psychological health assessment system according to claim 2, characterized in that: The dynamic adaptive AI assessment engine unit can shorten the traditional 10-20 minute assessment time to 5-8 minutes, improve the user completion rate, and at the same time eliminate the interference of incorrect answers by tracing the source of wrong questions, avoiding inaccurate assessment caused by a single answer deviation.

7. The intelligent psychological health assessment system according to claim 2, characterized in that: The privacy and security encryption unit complies with the Personal Information Protection Law and mental health service standards. It supports users to delete all assessment records with one click. The entire process operation audit log records data access, modification, and export behavior to ensure data use is traceable and meets the compliance requirements of industries such as medical care and education.

8. The intelligent psychological health assessment system according to claim 3, characterized in that: The AI-customized self-service toolkit pushed by the hierarchical intervention unit can adjust the content according to the user's schedule and preferences, including nighttime sleep-aid meditation, exercise-based mood regulation programs, mindfulness drawing, music therapy, etc., dynamically optimizing the frequency and content of reminders and cultivating the user's ability to manage emotions independently.

9. The intelligent psychological health assessment system according to claim 3, characterized in that: In the medical setting, the cross-scenario linkage service unit can synchronize the assessment report to the doctor's workstation, and discharged patients can complete follow-up assessments through the system. It can also link with community health service centers to achieve seamless integration of "in-hospital treatment + out-of-hospital follow-up" and reduce the recurrence rate.

10. A mental health intelligent assessment system according to claim 4, characterized in that: The early warning model constructed by the data value mining unit can monitor group data in the region in real time and promptly detect abnormal changes in mental health risks in specific groups and regions, including a sharp rise in the incidence of adolescent depression and abnormally high rates of occupational burnout in industries, thus providing support for precise prevention and control.

Images