Personalized treatment system and method based on patient's whole-course health management

Abstract

The application relates to the computer technical field and discloses a kind of individualized treatment system and method based on patient whole course health management, method includes: establishing standardization health data record module, support patient to enter emotional state, drug compliance, sleep quality and so on Multidimensional information and generate visual mood mapping;Integrate adaptive psychological training course based on cognitive behavior therapy;Rely on label matching technology to push personalized popular science content;Embedded intelligent follow-up engine automatically triggers follow-up and scale self-evaluation push;Build multi-source data risk monitoring module and start graded early warning response in real time;Build case management end to realize the customization and dynamic adjustment of individualized treatment plan.The application can realize seamless transfer of inpatient and outpatient treatment information and real-time early warning of crisis events by using the above technical solutions, effectively improving the continuity and intelligent level of whole course management of patients with depression.

Description

Technical Field

[0001] This invention belongs to the field of computer technology, specifically relating to a personalized treatment system and method based on the full course of patient health management. Background Technology

[0002] Depression, as a serious mental health disorder, is characterized by high prevalence and relapse rates, and has become a major global public health issue. With the deepening development of the biopsychosocial model of medicine, the treatment philosophy for depression has shifted from simply controlling acute symptoms to long-term management covering the entire course of the disease, emphasizing the organic integration of medication, psychotherapy, and rehabilitation intervention. Against this backdrop, how to leverage digital technology to build a continuous service system that spans in-hospital diagnosis and treatment and out-of-hospital management has become a core direction for the informatization of the mental health field.

[0003] Currently, the management of patients with depression primarily relies on periodic hospitalizations or outpatient treatments, with post-discharge management mainly depending on regular follow-up visits and telephone inquiries. However, this traditional follow-up model has significant drawbacks: the fixed follow-up intervals make it difficult to promptly capture changes in the patient's condition between visits. In particular, fluctuations in depressive symptoms and the emergence of risky behaviors by patients are often sudden and insidious, meaning that by the time healthcare professionals discover these changes through routine follow-ups, the condition may have already deteriorated to a severe stage. Furthermore, there is a lack of continuous data collection methods for key indicators such as medication adherence, sleep quality, and emotional state after discharge. This makes it difficult for healthcare professionals to assess treatment effectiveness and adjust intervention plans based on complete longitudinal data, resulting in an information gap between post-discharge management and in-hospital care, and compromising the continuity and effectiveness of the entire course of treatment.

[0004] While some depression management applications or platforms exist in the current technology landscape, these solutions generally suffer from limited functionality, data silos, and insufficient intelligence. Specifically: patient-side applications primarily offer simple self-assessment scales or medication reminders, lacking systematic collection and visualization of multidimensional health data; there is a lack of automated logical connection between scale assessment results and subsequent intervention plans, making it difficult to achieve dynamic intervention adjustments based on assessment data; the push of popular science content is mostly general, failing to personalize it according to the patient's disease stage, severity of depression, and individual characteristics, making it difficult for patients to obtain accurate and effective health information support; crisis warning mechanisms are limited to single-dimensional threshold triggers, lacking the ability to comprehensively analyze and assess risks from multi-source heterogeneous data (scale scores, physiological indicators, behavioral data), and cannot identify potential crisis signals in the early stages. Current technologies have failed to establish a closed-loop management architecture covering the entire chain of "patient self-management - medical-nursing collaborative intervention - intelligent data analysis." The lack of deep integration of business processes between patients, healthcare professionals, and data analytics hinders case managers from obtaining real-time patient data outside the hospital and responding quickly. Cross-departmental task flows (such as prescription execution, follow-up scheduling, and crisis intervention coordination) lack automated and visualized work order systems, limiting the efficiency of healthcare staff and resulting in insufficiently timely and accurate intervention support for patients. Therefore, how to deeply integrate patient self-management, collaborative intervention with healthcare professionals, and intelligent data analysis through digital technology to build a closed-loop management system covering the entire disease course, achieving seamless flow of in-hospital and out-of-hospital treatment information and real-time early warning of crisis events, has become a pressing technical challenge in the field of digital management of depression. Summary of the Invention

[0005] The purpose of this invention is to provide a personalized treatment system and method based on the patient's health management throughout the entire disease course, which can effectively solve the problems in the background art mentioned above.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A personalized treatment system based on patient health management throughout the entire disease course includes: The user management module is used to establish a standardized health data recording module. This module allows patients to input multidimensional information and simultaneously generate visual mood maps based on that information. The multidimensional information includes emotional state, medication adherence, sleep quality, and life events; The scale management module is used to dynamically adjust the training intensity and content based on the multidimensional information constructed according to the adaptive psychological training courses built based on cognitive behavioral therapy. The adaptive science popularization content module is used to push personalized science popularization content based on tag matching technology. The personalized science popularization content includes health management advice and dietary intervention plans. The intelligent follow-up management module is used to automatically trigger follow-up questions and push professional scale self-assessment. The personalized intervention module is used to generate a personalized treatment plan based on the multidimensional information, the visualized mood map, and the follow-up questions. The multi-source data risk monitoring module is used to perform graded early warning responses based on the multi-dimensional information, the visualized mood map, and the follow-up questions.

[0007] Preferably, the user management module further includes a file management unit, a message center unit, and a personal center unit. The file management unit forms a patient registration database through the patient registration service and a medical staff information management database through the registration of medical and health service personnel. Both the patient registration database and the medical staff information management database adopt a role-based access control mechanism to ensure data security.

[0008] Preferably, the message center unit is used to classify messages into medical appointment notifications, examination and test report notifications, payment notifications, health information pushes, and doctor-patient communication messages, and uses different visual identifiers to distinguish message types.

[0009] Preferably, the personal center unit is also used to edit the patient's name, gender, age, and contact information, as well as upload avatars and complete real-name authentication; The personal center unit is also used to view medical records and display medical records in the form of a timeline. It can also establish health files for several patients based on the patient registration database and display the health data in the health files in a visual trend.

[0010] Preferably, the scale management module integrates several commonly used depression assessment scales, including the Patient Health Questionnaire, Hamilton Depression Rating Scale, Beck Depression Rating Scale, Depression-Anxiety-Stress Rating Scale, and Childhood Depression Rating Scale, and recommends suitable scales for different patients to fill out based on user profiles.

[0011] Preferably, the adaptive science popularization content module is used to classify science popularization content according to knowledge type, presentation form, applicable population and disease stage in multiple dimensions and label it accordingly. Based on the user profile, it recommends science popularization content to patients through collaborative filtering algorithm and deep learning model, and dynamically adjusts the recommendation strategy according to the patient's real-time behavioral feedback.

[0012] Preferably, the intelligent follow-up management module includes a multi-channel follow-up task scheduling and message push unit, used for online remote video follow-up via mobile terminal; The personalized intervention module also includes a case manager database, in which several managers correspond to the patients. The managers develop personalized treatment plans based on the patient's disease stage using preset templates.

[0013] Preferably, the multi-source data risk monitoring module identifies potential crisis risks in real time based on the scores of risk behavior items in the patient scale, abnormal physiological data from smart devices, and high-frequency access records of crisis intervention content in the adaptive science popularization content module, and classifies the crisis risks into three risk levels to trigger corresponding response measures; wherein, The three risk levels mentioned include low risk, medium risk, and high risk.

[0014] Preferably, the multi-source data risk monitoring module is also used to automatically push soothing audio and articles to patients when the crisis risk is at a low risk level, remind the administrator to intervene and communicate when the crisis risk is at a medium risk level, and immediately trigger emergency contact notification and generate a crisis management suggestion list when the crisis risk is at a high risk level.

[0015] Based on the same technical concept, the present invention also provides a personalized treatment method based on the patient's health management throughout the entire course of disease, applied to the aforementioned personalized treatment system based on the patient's health management throughout the entire course of disease, comprising: A standardized health data recording module is established, which is used for patients to enter multidimensional information and simultaneously generate a visual mood map based on the multidimensional information; The multidimensional information includes emotional state, medication adherence, sleep quality, and life events; An adaptive psychological training course based on cognitive behavioral therapy dynamically adjusts the training intensity and content based on the aforementioned multidimensional information; Personalized science popularization content is pushed based on tag matching technology. The personalized science popularization content includes health management advice and dietary intervention plans. Automatically triggers follow-up questions and professional scale self-assessment push notifications; A personalized treatment plan is generated based on the multidimensional information, the visualized mood map, and the follow-up questions. A tiered early warning response is implemented based on the multidimensional information, the visualized mood map, and the follow-up questions.

[0016] Compared with existing technologies, the present invention has the following beneficial effects: by constructing a closed-loop management system covering the entire chain of patient self-management, medical and nursing collaborative intervention and data intelligent analysis, it realizes seamless flow of diagnosis and treatment information inside and outside the hospital and real-time early warning of crisis events, effectively improving the continuity and intelligence level of the whole course management of patients with depression. Attached Figure Description

[0017] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 A functional block diagram of a personalized treatment system based on patient health management throughout the entire disease course, provided in an embodiment of the present invention; Figure 2 A flowchart illustrating a personalized treatment method based on the patient's health management throughout the entire disease course, provided as an embodiment of the present invention. Detailed Implementation

[0018] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0019] See Figure 1 As shown, this invention provides a personalized treatment system based on patient health management throughout the entire disease course, comprising: The user management module is used to establish a standardized health data recording module. This module allows patients to input multidimensional information and simultaneously generates visual mood maps based on that information. Multidimensional information includes emotional state, medication adherence, sleep quality, and life events; The scale management module is used to dynamically adjust the intensity and content of training based on multidimensional information in adaptive psychological training courses built according to cognitive behavioral therapy. The adaptive science popularization content module is used to push personalized science popularization content based on tag matching technology. Personalized science popularization content includes health management advice and dietary intervention plans. The intelligent follow-up management module is used to automatically trigger follow-up questions and push professional scale self-assessment. The personalized intervention module is used to generate personalized treatment plans based on multidimensional information, visualized mood maps, and follow-up questions. The multi-source data risk monitoring module is used to conduct graded early warning responses based on multi-dimensional information, visualized mood maps, and follow-up questions.

[0020] In one specific embodiment of this application, the user management module further includes a file management unit, a message center unit, and a personal center unit. The file management unit forms a patient registration database through the patient registration service and a medical staff information management database through the registration of medical and health service personnel. Both the patient registration database and the medical staff information management database adopt a role-based access control mechanism to ensure data security.

[0021] In one specific embodiment of this application, the message center unit is used to classify messages into medical appointment notifications, examination and test report notifications, payment notifications, health information pushes, and doctor-patient communication messages, and uses different visual identifiers to distinguish message types.

[0022] In one specific embodiment of this application, the personal center unit is also used to edit the patient's name, gender, age, and contact information, as well as upload avatars and complete real-name authentication; The personal center unit is also used to view medical records and display medical records in a timeline format. It establishes health files for several patients based on the patient registration database and displays the health data in the health files in a visual trend.

[0023] In one specific embodiment of this application, the scale management module integrates several commonly used depression assessment scales, including the Patient Health Questionnaire, Hamilton Depression Rating Scale, Beck Depression Rating Scale, Depression-Anxiety-Stress Rating Scale, and Childhood Depression Rating Scale. The module also recommends suitable scales for different patients to complete based on their user profiles.

[0024] In one specific embodiment of this application, the adaptive science popularization content module is used to classify science popularization content according to knowledge type, presentation form, applicable population and disease stage in multiple dimensions and label it accordingly. Based on user profile, it recommends science popularization content to patients through collaborative filtering algorithm and deep learning model, and dynamically adjusts the recommendation strategy according to the patient's real-time behavioral feedback.

[0025] In one specific embodiment of this application, the intelligent follow-up management module includes a multi-channel follow-up task scheduling and message push unit, which is used to conduct online remote video follow-up via mobile terminal; The personalized intervention module also includes a case manager database, where several managers correspond to patients. The managers develop personalized treatment plans based on the patient's disease stage using preset templates.

[0026] In one specific embodiment of this application, the multi-source data risk monitoring module identifies potential crisis risks in real time based on the scores of risk behavior items in the patient scale, abnormal physiological data from smart devices, and high-frequency access records of crisis intervention content in the adaptive science popularization content module, and classifies the crisis risks into three risk levels to trigger corresponding response measures; wherein, The three risk levels are low risk, medium risk, and high risk.

[0027] In one specific embodiment of this application, the multi-source data risk monitoring module is also used to automatically push soothing audio and articles to patients when the crisis risk is at a low risk level, remind administrators to intervene and communicate when the crisis risk is at a medium risk level, and immediately trigger emergency contact notification and generate a crisis management suggestion list when the crisis risk is at a high risk level.

[0028] Based on the same technical concept, see [reference] Figure 2 As shown, a personalized treatment method based on patient health management throughout the entire disease course, applied in a personalized treatment system based on patient health management throughout the entire disease course, includes: Establish a standardized health data recording module, which is used by patients to enter multidimensional information and simultaneously generate a visual mood map based on the multidimensional information; The multidimensional information includes emotional state, medication adherence, sleep quality, and life events; An adaptive psychological training course based on cognitive behavioral therapy dynamically adjusts the intensity and content of training based on multidimensional information. Personalized science content is pushed based on tag matching technology. The personalized science content includes health management advice and dietary intervention plans. Automatically triggers follow-up questions and professional scale self-assessment push notifications; Personalized treatment plans are generated based on multidimensional information, visualized mood maps, and follow-up questions. A tiered early warning response is implemented based on multidimensional information, visualized mood maps, and follow-up questions.

[0029] In this patent's technical solution, the system is built using a multi-tiered B / S software architecture based on J2EE. Application developers can focus on encapsulating business logic and rules, while runtime attributes of application components such as transactions, persistence, and security are customized in the deployment environment using a declarative approach. The system is designed and built entirely based on cloud computing technology, enabling elastic resource allocation according to platform load and user scale. This ensures both platform flexibility and scalability while enhancing stability and disaster recovery capabilities. The system supports distributed deployment, provides management services for the cluster system, and achieves smooth platform expansion.

[0030] The user management module, serving as the system's foundational support layer, is located on the server side and establishes a persistent connection with the database cluster. This module comprises three functional sub-modules: a file management unit, a message center unit, and a personal center unit. The file management unit is configured to perform patient registration services, medical staff registration services, and role-based access control functions. Specifically, it receives patient identity information entry requests from mobile terminals through the patient registration service module, verifies the integrity of the identity information according to preset verification rules, and generates a globally unique patient identifier upon successful verification, writing it to the patient registration database. The file management unit also includes a patient identity matching service sub-module, configured to merge duplicate patient identities in the data center using a fuzzy identity matching algorithm. This fuzzy identity matching algorithm comprehensively judges based on multi-dimensional features such as name similarity calculation, ID number check digit verification, and contact information association analysis. The medical and health service personnel registration service sub-module is configured to receive registration applications from medical staff, record their professional information, professional qualifications, and department data, and assign a unique employee ID to each medical staff member. The access control submodule is configured to implement a role-based access control mechanism, which divides system resources into categories such as medical records, scale data, follow-up tasks, and intervention plans, and assigns corresponding access permissions according to user roles (patients, nurses, doctors, case managers, and administrators).

[0031] The scale management module is integrated into the application server and is configured to manage the item data, scoring rules, and result interpretation logic of various commonly used depression assessment scales. This module includes a built-in item library for 9 patient health questionnaires, a 24-item Hamilton Depression Rating Scale, a Beck Depression Rating Scale, a Depression-Anxiety-Stress Rating Scale, and a Childhood Depression Rating Scale. Each scale has its own independent item data table and scoring calculation algorithm. The scale management module also features a scale recommendation engine, configured to automatically match suitable scale types for patients based on user profile data such as age, gender, and previous diagnostic records in the patient's file, and push the recommendations to the task queue of the patient's mobile application.

[0032] The adaptive science popularization content module is deployed on a content server cluster and comprises four subsystems: a content management subsystem, a tagging system subsystem, a recommendation algorithm subsystem, and a content distribution subsystem. The content management subsystem is configured to store and manage various forms of health education materials, including science popularization articles, text and image information, video content, audio files, and interactive games. Each piece of content is associated with knowledge type tags, presentation format tags, target audience tags, and disease stage tags. The tagging system subsystem maintains a multi-dimensional tag classification tree structure, supporting cross-indexing of content based on knowledge type dimensions (disease cognition, treatment methods, rehabilitation techniques, psychological adjustment, etc.), presentation format dimensions (text, image, video, audio, animation, interactive games, etc.), target audience dimensions (adolescents, adults, elderly, etc.), and disease stage dimensions (acute phase, remission phase, consolidation phase, etc.). The recommendation algorithm subsystem is configured to use collaborative filtering algorithms and deep learning models to provide personalized content recommendations to patients. This subsystem includes a user feature extraction module, an item feature extraction module, a similarity calculation module, and a ranking output module. The content distribution subsystem is configured to push popular science content to patients' mobile devices based on recommendation results, and after patients perform behaviors such as browsing, liking, collecting, and sharing, the behavioral data is sent back to the recommendation algorithm subsystem for model updates.

[0033] The intelligent follow-up management module is located on the follow-up scheduling server and includes a follow-up task scheduling subsystem, a message push subsystem, a video call subsystem, and a record management subsystem. The follow-up task scheduling subsystem is configured to automatically generate follow-up task work orders based on preset follow-up plan templates and the patient's actual situation, and distribute the work orders to the corresponding execution queues. The message push subsystem is configured to achieve precise delivery of follow-up reminders through mobile messaging channels, supporting multiple push channels such as SMS, push notifications, and in-app messages. The video call subsystem is configured to provide real-time video communication capabilities, including a signaling server, a media stream forwarding server, and a bandwidth adaptive adjustment module, which can automatically adjust video resolution and frame rate when network conditions change. The record management subsystem is configured to perform structured storage of materials such as voice-to-text data, video call records, and follow-up conclusions generated during the follow-up process, and provides a time-axis-based visual retrieval function.

[0034] The personalized intervention module's case manager interface is deployed on healthcare workstations and includes a patient data aggregation subsystem, a treatment plan management subsystem, a work order workflow subsystem, and an intervention record subsystem. The patient data aggregation subsystem is configured to synchronize patient information from multiple data sources, including the user management module, scale management module, adaptive science popularization content module, and smart device data interfaces, constructing a multi-dimensional patient profile dataset encompassing basic information, assessment results, behavioral trajectories, and physiological indicators. The treatment plan management subsystem is configured to provide functions for creating, editing, version management, and execution tracking of treatment plans, including a pre-set treatment plan template library covering medication adjustment plans, cognitive behavioral therapy course arrangements, exercise rehabilitation plans, and meditation training courses. The work order workflow subsystem is configured to enable cross-departmental task collaboration, supporting the transfer and signing of work orders such as medical order execution tasks, follow-up tasks, and crisis intervention tasks among nurses, doctors, and case managers. The intervention record subsystem is configured to record the specific content of each intervention operation, participating personnel, execution time, and intervention effect evaluation data.

[0035] The multi-source data risk monitoring module is integrated into the data analysis server and includes a data acquisition pipeline, a risk calculation engine, and an early warning distribution module. The data acquisition pipeline is configured to receive in real-time risk behavior item scores from the scale assessment module, abnormal physiological indicator data from smart device data interfaces (such as sleep duration below the threshold and heart rate variability exceeding the normal range), and user behavior logs from the adaptive science popularization content module (especially the frequency of access to crisis intervention-related content). The risk calculation engine is configured to execute a preset risk algorithm model that comprehensively considers the weighted scores of multi-dimensional input features to calculate a comprehensive crisis risk score. The early warning distribution module is configured to trigger corresponding tiered early warning response processes based on the risk score.

[0036] The tiered early warning response, as a downstream processing module of the risk monitoring module, is configured to execute three different levels of early warning response actions. In low-risk situations, the system automatically pushes soothing audio files and stress-reduction articles to the patient's mobile device; the content is provided by the stress-reduction content library in the content management subsystem. In medium-risk situations, the system sends a reminder notification to the case manager's workstation via the message center, marks the patient as high-concern in the case manager's patient list, and automatically opens an instant messaging window with the patient. In high-risk situations, the system immediately triggers the emergency contact notification process, simultaneously sending crisis warning information to pre-set family contacts via SMS and telephone calls, and simultaneously generating a work order containing the patient's basic information, current location, recent medical records, and a crisis management suggestion list, which is then pushed to the to-do queues of the case manager and the crisis intervention team.

[0037] The patented technical solution of this invention will be further explained through specific implementation and application scenarios. Example 1

[0038] The system workflow in this embodiment is as follows: The standardized health data recording module is configured to receive multi-dimensional health information entered by patients via a mobile application, including self-assessment data of emotional state, medication adherence records, sleep quality scores, and life event records. The module's data entry interface is configured to provide a structured form input method, supporting patients in selecting preset emotion options (such as happy, neutral, depressed, frustrated, and hopeless) and a slider-style sleep quality score. After data submission, the standardized health data recording module stores the data in the patient's health record database and simultaneously triggers the visualization mood mapping module to generate a mood time series chart. This chart uses date as the horizontal axis and mood score as the vertical axis, displaying the patient's recent mood fluctuation trends in a line graph format. The chart data supports aggregation and display by weekly, monthly, and quarterly time granularities.

[0039] The adaptive psychological training module is configured to build a psychological training content system based on the theoretical framework of cognitive behavioral therapy, including course units such as cognitive restructuring training, behavioral activation exercises, and emotion regulation techniques. The module's treatment engine is configured to dynamically adjust the intensity and content combination of subsequent training courses based on the patient's scores on scale assessments, recent mood mapping trends, and the patient's progress in the psychological training courses. Specifically, when the system detects a continuous downward trend in the patient's recent mood score, the treatment engine automatically adjusts the course difficulty to an introductory level and adds basic exercises in emotion recognition and naming; when the patient completes multiple courses consecutively and the assessment results show symptom improvement, the treatment engine gradually increases the training intensity, adding advanced content such as cognitive distortion recognition and Socratic questioning.

[0040] The personalized science popularization content push process is triggered. The recommendation algorithm subsystem of the adaptive science popularization content module first extracts the patient's multi-dimensional feature vector from the patient profile database, including age group, gender, duration of diagnosis, current disease stage (acute phase, consolidation phase, maintenance phase), recent scale score range, and historical interaction data on various types of science popularization content. Then, the recommendation algorithm subsystem inputs this feature vector into a collaborative filtering model and a deep neural network recommendation model for inference calculation, producing two recommendation candidate sets respectively. The collaborative filtering model recommends content that is "generally preferred by similar user groups and has not yet been accessed by the target user" to the target patient by analyzing the content preferences of similar user groups; the deep neural network recommendation model predicts the probability of the patient's interest in each candidate content by learning the non-linear mapping relationship between user features and content features. Finally, the two candidate sets are merged through a weighted fusion strategy, sorted by score, and a final recommendation list is generated and pushed to the science popularization content recommendation section on the patient's mobile device. During the content display process, the system continuously collects real-time feedback data such as the patient's browsing time, click behavior, and collection behavior, and sends it back to the model training pipeline for online learning and updates, ensuring that the recommendation strategy can be dynamically adjusted according to the patient's real-time behavior.

[0041] The intelligent follow-up management module is configured to automate the follow-up task scheduling and questionnaire delivery process. The module's follow-up scheduling engine first generates a follow-up task work order based on the follow-up plan set by the patient administrator on the case manager's end. This work order includes attributes such as the follow-up purpose, a list of recommended questionnaires, and estimated time. Subsequently, the intelligent follow-up management module's message push subsystem is configured to send follow-up reminder notifications to the patient's mobile device. The reminder includes the follow-up time, follow-up method (video follow-up / telephone follow-up / text self-assessment), and materials to be prepared in advance. After receiving the reminder, the patient can choose to accept the follow-up appointment or modify the follow-up time on their mobile device. After the follow-up appointment is confirmed, the scheduling engine automatically synchronizes the follow-up task to the case manager's schedule module and sends a reminder again 30 minutes before the follow-up. During the follow-up phase, the video call subsystem is configured to establish a real-time video connection between the patient and the case manager. During the follow-up, the case manager can view the patient's historical assessment data, recent health records, and the case summary from the last follow-up on the system interface. After the follow-up, the case manager completes the follow-up record in the system, including follow-up conclusions, identified issues, and subsequent recommendations. The follow-up record is automatically archived in the patient's health file.

[0042] The multi-source data risk monitoring module operates continuously. This module's data acquisition pipeline receives event streams from various data sources around the clock. When a patient completes a scale assessment containing risky behavior items in the scale management module, the assessment results are immediately pushed to the risk calculation engine. When the smart device data interface detects abnormal patterns in sleep data transmitted from a patient's smart bracelet, such as sleep duration less than 4 hours for three consecutive days or a nighttime resting heart rate exceeding 120 beats / min, these abnormal events are pushed to the risk calculation engine. When the adaptive science popularization content module detects frequent access to crisis intervention-related science popularization content (such as behavioral risk identification, how to seek help, and psychological crisis hotlines) within a short period, the access log is pushed to the risk calculation engine. After receiving these multi-source events, the risk calculation engine executes a risk algorithm model to calculate a comprehensive score. The core logic of this model can be expressed as: the comprehensive score equals the weighted score of the risky behavior items multiplied by a first weight coefficient, plus the weighted score of the physiological abnormality event multiplied by a second weight coefficient, plus the weighted score of the crisis content access behavior multiplied by a third weight coefficient. Based on the range of comprehensive scores, the system classifies crisis risk into three levels: low (0-30 points), medium (31-70 points), and high (71-100 points).

[0043] The personalized treatment plan customization and dynamic adjustment process is executed on the case manager's end. After logging into the system, the case manager first views the patient's complete profile data through the patient data aggregation subsystem, including basic information synchronized from the user management module, assessment results from the scale management module, learning records summarized from the adaptive science popularization module, follow-up records archived from the intelligent follow-up management module, and recent physiological indicator trends obtained from the smart device data interface. Then, based on the patient's disease stage, the case manager selects the corresponding treatment plan template from the preset template library of the treatment plan management subsystem. The template content covers medication adjustment suggestions, psychotherapy arrangements (such as weekly cognitive behavioral therapy), rehabilitation training tasks (such as three aerobic exercises per week and 15 minutes of mindfulness meditation daily), sleep improvement plans, etc. After the treatment plan is created, the system automatically breaks down the plan content into executable task orders and distributes them to the task center on the patient's mobile device. The patient performs the tasks according to the plan, and the generated data is sent back to the system. When a case manager discovers that the treatment plan needs to be adjusted based on the patient's follow-up visit feedback or daily monitoring data, the existing plan can be edited and modified in the treatment plan management subsystem. The system automatically records the changes and generates a new version snapshot, forming a complete record of the treatment plan change process.

[0044] Based on Example 1, this example provides differentiated system configuration schemes for special populations (such as adolescent patients with depression).

[0045] Adolescent patients with depression differ significantly from adult patients in terms of symptom presentation, disease course, and treatment needs. In this embodiment, the scale management module is configured to prioritize the Childhood Depression Scale and the 9-item Patient Health Questionnaire as primary assessment tools, while using the 24-item Hamilton Depression Rating Scale as a supplementary option for professional assessment. The scale recommendation engine incorporates adolescent patient characteristic recognition rules; when the age field in the patient's profile falls within the 6-18 age range, the system automatically adjusts the scale recommendation priority to prioritize the Childhood Depression Scale, followed by the 9-item Patient Health Questionnaire.

[0046] The adaptive science popularization content module constructs a dedicated content tagging system for teenagers. In terms of target audience tags, the system sets exclusive tags for teenagers (such as "junior high school students," "high school students," and "university students"). In terms of presentation format tags, the system adds content formats preferred by teenagers, such as interactive games, short videos, and comics. The recommendation algorithm subsystem trains a dedicated collaborative filtering model for teenagers. This model incorporates features such as similar age group and grade level into the definition of similar users, making the recommendation results more aligned with teenagers' information acquisition habits and interests.

[0047] The video call subsystem of the intelligent follow-up management module is configured to support the collaborative participation of guardians in follow-up visits. In follow-up scenarios for adolescent patients, the system automatically identifies the age characteristics of the patient and allows guardians to join the meeting when a video call is established. The follow-up record module is configured to record behavioral observation information provided by guardians and self-reported information from the child, providing case managers with more comprehensive assessment data.

[0048] The case manager's interface for the personalized intervention module is configured to provide treatment plan templates specifically for adolescents. These templates emphasize engagement and participation, transforming traditional written exercises in cognitive behavioral therapy into more interactive digital tasks, such as completing cognitive restructuring exercises through gamified interfaces and learning emotion regulation skills through animated videos. The frequency and duration of each treatment session are also adjusted based on the attentional characteristics of adolescent patients, employing a short-duration, high-frequency task segmentation strategy.

[0049] The tiered early warning response is configured with more sensitive trigger thresholds for adolescent patients. Considering the sudden onset and insidious nature of risky behaviors in adolescents, the system increases the weighting of risk behavior questions by 20% and shortens the detection window for abnormal physiological events to two consecutive days. When a high-risk level is triggered, the system not only notifies emergency contacts but also automatically generates a suggestion form prompting the case manager to contact the school counselor or community mental health doctor.

[0050] Based on Example 1, this example provides a remote closed-loop management solution for patients with depression who are recovering at home after being discharged from the hospital.

[0051] For patients who have completed acute treatment and entered the home rehabilitation phase, the system in this embodiment establishes a data connection with the patient's self-purchased smart wearable device to continuously collect out-of-hospital physiological indicators. The smart device's data interface is configured to support data access protocols for various mainstream smart bracelets and smartwatches, including heart rate monitoring data, sleep stage data, and activity step count data. The data acquisition pipeline pulls the latest data from the device's cloud every 15 minutes, performs data cleaning and outlier filtering locally, and then stores it in the patient's health record database.

[0052] In home-based rehabilitation scenarios, the follow-up scheduling strategy of the intelligent follow-up management module has been adjusted to a more flexible combination of on-demand and regular follow-up. Regular follow-ups are automatically triggered every two weeks, focusing on self-assessment using scales and reviewing health data. On-demand follow-ups are initiated by patients via mobile devices. When a patient's mood score is below a threshold for three consecutive days or smart device data shows a significant decline in sleep quality, the system automatically sends a notification to the patient requesting an appointment. After the patient selects to schedule an appointment, the system automatically recommends available time slots based on the availability of the case manager's schedule, and the patient confirms and completes the appointment.

[0053] The adaptive science popularization content module's content recommendation strategy during the home rehabilitation phase focuses on consolidating rehabilitation skills and improving quality of life. The recommendation algorithm subsystem adds rehabilitation stage tags to patient profiles and dynamically adjusts the difficulty gradient of recommended content based on the length of time a patient has been in the consolidation or maintenance phase. For patients in the early stages of rehabilitation, the system prioritizes recommending basic disease knowledge and content on emotional self-monitoring skills; for patients in the stable rehabilitation phase, the system gradually adds guidance on social function rehabilitation (such as interpersonal skills and preparation for vocational ability recovery) and relapse prevention knowledge.

[0054] The personalized intervention module's case manager interface enhances remote monitoring and timely intervention capabilities during the home-based rehabilitation phase. The system automatically generates weekly and monthly patient reports, including multi-dimensional data such as scale score trends, statistics of abnormal physiological events, completion of learning content, and follow-up adherence indicators, presented to the case manager in visual charts. When the system detects signs of declining patient adherence (such as failure to complete rehabilitation tasks for a consecutive week, extended intervals between scale assessments, or reduced frequency of data collection from smart devices), it automatically sends an alert to the case manager, allowing for timely intervention, communication, understanding of the patient's condition, and provision of targeted support.

[0055] In home-based rehabilitation scenarios, the multi-source data risk monitoring module is configured to access data from more dimensions. In addition to the three types of data sources described in Example 1, the system also integrates patient medication check-in records (collected through confirmation operations of the mobile medication reminder function), exercise data (steps and exercise duration collected through smart devices), and social interaction data (collected through patient posting and commenting behavior in the application's community). The risk calculation engine employs more diversified feature engineering methods to transform the aforementioned multi-dimensional behavioral data into auxiliary features for risk assessment, further improving the accuracy and timeliness of crisis early warning.

[0056] The tiered early warning response system works in conjunction with community mental health prevention and control efforts in home-based rehabilitation scenarios. When the system triggers a medium-risk warning, it not only alerts the case manager to intervene but also automatically sends a request for assistance to the mental health doctor in the patient's community, providing the patient's recent medical records and risk assessment results. When a high-risk warning is triggered, the system notifies emergency contacts and automatically generates an emergency rescue package containing location information, past medical history, medication details, and crisis management suggestions. This package is then integrated with the local 120 emergency dispatch system and the mental health center's crisis intervention hotline via an interface, enabling a one-click rapid response.

[0057] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the present invention.

Claims

1. A personalized treatment system based on patient health management throughout the entire disease course, characterized in that, include: The user management module is used to establish a standardized health data recording module. This module allows patients to input multidimensional information and simultaneously generate visual mood maps based on that information. The multidimensional information includes emotional state, medication adherence, sleep quality, and life events; The scale management module is used to dynamically adjust the training intensity and content based on the multidimensional information constructed according to the adaptive psychological training courses built based on cognitive behavioral therapy. The adaptive science popularization content module is used to push personalized science popularization content based on tag matching technology. The personalized science popularization content includes health management advice and dietary intervention plans. The intelligent follow-up management module is used to automatically trigger follow-up questions and push professional scale self-assessment. The personalized intervention module is used to generate a personalized treatment plan based on the multidimensional information, the visualized mood map, and the follow-up questions. The multi-source data risk monitoring module is used to perform graded early warning responses based on the multi-dimensional information, the visualized mood map, and the follow-up questions.

2. The personalized treatment system based on the patient's health management throughout the entire disease course according to claim 1, characterized in that, The user management module also includes a file management unit, a message center unit, and a personal center unit. The file management unit forms a patient registration database through the patient registration service and a medical staff information management database through the registration of medical and health service personnel. Both the patient registration database and the medical staff information management database adopt a role-based access control mechanism to ensure data security.

3. The personalized treatment system based on the patient's health management throughout the entire disease course according to claim 2, characterized in that, The message center unit is used to classify messages into medical appointment notifications, examination and test report notifications, payment notifications, health information pushes, and doctor-patient communication messages, and uses different visual identifiers to distinguish message types.

4. The personalized treatment system based on the patient's health management throughout the entire disease course according to claim 2, characterized in that, The personal center unit is also used to edit the patient's name, gender, age, and contact information, as well as upload avatars and complete real-name authentication; The personal center unit is also used to view medical records and display medical records in the form of a timeline. It can also establish health files for several patients based on the patient registration database and display the health data in the health files in a visual trend.

5. A personalized treatment system based on patient health management throughout the entire disease course according to claim 1, characterized in that, The scale management module integrates several commonly used depression assessment scales, including the Patient Health Questionnaire, Hamilton Depression Rating Scale, Beck Depression Rating Scale, Depression-Anxiety-Stress Rating Scale, and Childhood Depression Rating Scale. It also recommends suitable scales for different patients to complete based on their user profiles.

6. A personalized treatment system based on patient health management throughout the entire disease course, as described in claim 5, is characterized in that... The adaptive science popularization content module is used to classify science popularization content according to knowledge type, presentation form, applicable population and disease stage in multiple dimensions and label it accordingly. Based on the user profile, it recommends science popularization content to patients through collaborative filtering algorithm and deep learning model, and dynamically adjusts the recommendation strategy according to the real-time behavior feedback of patients.

7. A personalized treatment system based on full-course patient health management according to claim 1, characterized in that, The intelligent follow-up management module includes a multi-channel follow-up task scheduling and message push unit, which is used to conduct online remote video follow-up via mobile terminal; The personalized intervention module also includes a case manager database, in which several managers correspond to the patients. The managers develop personalized treatment plans based on the patient's disease stage using preset templates.

8. A personalized treatment system based on patient health management throughout the entire disease course according to claim 1, characterized in that, The multi-source data risk monitoring module identifies potential crisis risks in real time based on the scores of risk behavior items in the patient scale, abnormal physiological data from smart devices, and high-frequency access records of crisis intervention content in the adaptive science popularization content module, and classifies the crisis risks into three risk levels to trigger corresponding response measures; wherein, The three risk levels mentioned include low risk, medium risk, and high risk.

9. A personalized treatment system based on the patient's health management throughout the entire disease course, as described in claim 8, is characterized in that... The multi-source data risk monitoring module is also used to automatically push soothing audio and articles to patients when the crisis risk is at a low risk level, remind the administrator to intervene and communicate when the crisis risk is at a medium risk level, and immediately trigger emergency contact notification and generate a crisis management suggestion list when the crisis risk is at a high risk level.

10. A personalized treatment method based on patient health management throughout the entire disease course, applied to the personalized treatment system based on patient health management throughout the entire disease course as described in any one of claims 1-9, characterized in that, include: A standardized health data recording module is established, which is used for patients to enter multidimensional information and simultaneously generate a visual mood map based on the multidimensional information; The multidimensional information includes emotional state, medication adherence, sleep quality, and life events; An adaptive psychological training course based on cognitive behavioral therapy dynamically adjusts the training intensity and content based on the aforementioned multidimensional information; Personalized science popularization content is pushed based on tag matching technology. The personalized science popularization content includes health management advice and dietary intervention plans. Automatically triggers follow-up questions and professional scale self-assessment push notifications; A personalized treatment plan is generated based on the multidimensional information, the visualized mood map, and the follow-up questions. A tiered early warning response is implemented based on the multidimensional information, the visualized mood map, and the follow-up questions.

Images