HIV / AIDS patient comprehensive disease management system based on multi-modal fusion machine learning

Abstract

The invention provides an HIV / AIDS patient comprehensive disease management system based on multi-modal fusion machine learning, and belongs to the technical field of medical data processing. According to the system, multi-modal data acquisition, processing, fusion and intelligent decision making are integrated, the system architecture is modularly designed, the function expansibility is high, and the deployment and maintenance cost is low; multi-agent collaborative management and personalized precise intervention can be realized on the same platform; the intelligent monitoring and intervention method provided by the invention is based on multi-modal data fusion and a machine learning algorithm, the health state of a patient is dynamically identified, and an accurate health management scheme is generated; the distributed storage and encryption technology is combined, efficient and safe data management and sharing are achieved, and the intelligence and reliability of disease management are comprehensively improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of medical data processing, and in particular relates to a comprehensive disease management system for HIV / AIDS patients based on multimodal fusion machine learning. Background Art

[0002] In recent years, artificial intelligence (AI) has driven the widespread application of intelligent disease management systems in chronic disease management. HIV / AIDS, a major global public health issue, requires patient management involving medication guidance, symptom monitoring, psychological support, and follow-up plans.

[0003] Currently, disease management for HIV / AIDS patients mainly relies on the following technical solutions: (1) SMS reminder systems: These use timed SMS reminders to remind patients to take medication and undergo follow-up examinations, aiming to improve their treatment compliance. However, these systems are only one-way reminders and lack personalized management and intervention. (2) Mobile applications: Some applications provide patients with health knowledge, disease management guidelines, and basic record keeping functions, but they tend to focus on information transmission and have relatively simple functions, making it difficult to meet the diverse needs of patients due to individual differences in their daily lives. (3) Electronic medical record systems: These facilitate clinical follow-up and treatment decision-making by collecting and storing patients' diagnosis and treatment information. However, these systems focus on recording static patient data and do not actively integrate non-clinical data for comprehensive analysis, resulting in a lack of dynamic monitoring, proactive management, and precise intervention capabilities.

[0004] The existing systems still have five major deficiencies: (1) Limited service targets: There is a lack of systems that can be used by multiple entities; (2) Functional limitations: Most can only achieve a single function, some can only achieve two functions, and the functional coverage is narrow; (3) Inability to intelligently identify user needs: It is difficult to automatically provide personalized solutions; (4) Lack of information interaction: Existing HIV / AIDS patient disease management systems mainly rely on medical staff to provide PLWHA with unilateral support, and do not effectively obtain PLWHA feedback, resulting in a lack of interaction. In addition, the various systems are currently unmaintained and iterated, and the internal knowledge base of the system is not updated in a timely manner, resulting in poor information timeliness.

[0005] In view of this, there is an urgent need to develop an AIDS disease management system that is comprehensive in functionality, can intelligently identify user needs, has an efficient interactive mechanism, allows for multi-subject participation, and has high acceptability. Summary of the Invention

[0006] The present invention aims to solve the problems of the existing HIV / AIDS patient disease management system, such as insufficient utilization of multimodal data, single service objects, functional limitations, weak intelligent recognition ability of user needs, lack of information interaction, and delayed system maintenance and knowledge updating. It integrates heterogeneous data through multimodal data fusion technology to comprehensively construct patient health portraits; supports the use of multiple subjects such as patients, medical staff and prevention and control management personnel to achieve resource collaboration; integrates monitoring, early warning, intervention and evaluation functions to cover the entire process of disease management; accurately identifies patients' personalized needs based on deep learning algorithms and provides dynamically adjusted intervention suggestions; realizes real-time communication between supply and demand parties through two-way interactive design; and has a built-in automatic update mechanism to ensure the real-time nature of the knowledge base and the timeliness of decision-making, thereby significantly improving the comprehensiveness, intelligence and precision of the disease management system, improving patient management effects and reducing the burden on medical staff.

[0007] The purpose of the present invention is to provide a comprehensive disease management system for HIV / AIDS patients based on multimodal fusion machine learning.

[0008] The present invention provides a comprehensive disease management system for HIV / AIDS patients based on multimodal fusion machine learning, comprising:

[0009] An input module configured to input multimodal data of a patient;

[0010] The data processing module is configured to process the multimodal data of the input patient in combination with the database and the knowledge graph;

[0011] an output module configured to combine the output information of the data processing module with the patient's health feature vector and the disease management knowledge base to generate personalized early warning and intervention suggestions;

[0012] Wherein, the data processing module includes a data fusion module, a feature extraction module and an information processing module.

[0013] Furthermore, the multimodal data includes examination results, treatment plans, psychological conditions, positioning information and diagnosis and treatment needs.

[0014] Furthermore, in the data processing module, multimodal data requires preprocessing, which includes data cleaning, standardization, and feature extraction.

[0015] Furthermore, the information processing module connects the information processed by the feature extraction module with the machine learning module and renders it in combination with a visualization method.

[0016] Furthermore, the machine learning module is selected from convolutional neural networks and Transformer.

[0017] Furthermore, the output module compares the health feature vector with the normal range value in the knowledge base to generate a health status report; when an abnormality is detected, an early warning signal is generated based on the machine learning algorithm and intervention measures are recommended; the system displays the patient's health status trend chart, early warning information and personalized intervention suggestions to medical staff through an interactive interface.

[0018] The present invention has achieved the following beneficial effects:

[0019] The present invention is based on a comprehensive disease management system for HIV / AIDS patients based on multimodal fusion machine learning. By integrating multimodal data, supporting multi-subject collaboration, and realizing intelligent personalized management, it significantly improves the efficiency and accuracy of disease management for HIV / AIDS patients. The system can comprehensively integrate heterogeneous data from different sources, provide accurate patient health portraits, and solve the problem of insufficient data utilization in traditional systems; at the same time, it supports multi-subject collaboration such as patients, medical staff, and disease control management personnel, thereby improving management efficiency; personalized and precise intervention is achieved through deep learning algorithms, solving the problem that existing technologies cannot intelligently identify patient needs; the system's two-way interactive design enhances information feedback and communication, promotes interaction between patients and medical staff, and further improves disease management effects; in addition, the built-in automatic update and security encryption mechanism solves the existing system maintenance and data security problems, ensuring the long-term stability of the system and the real-time nature of the data. Therefore, the present invention greatly improves the intelligent level of disease management, solves multiple technical bottlenecks, and has important practical application value.

[0020] Obviously, based on the above contents of the present invention, according to common technical knowledge and customary means in this field, without departing from the above basic technical ideas of the present invention, other various forms of modifications, replacements or changes can be made.

[0021] The following is a detailed description of the present invention through specific embodiments in the form of examples. However, this should not be construed as limiting the scope of the present invention to the following examples. All technologies implemented based on the above-mentioned content of the present invention fall within the scope of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 The comprehensive disease management system for HIV / AIDS patients based on multimodal fusion machine learning consists of: a user input system 10, a data processing system 20, and a function system 30.

[0023] Figure 2 Workflow diagram of the comprehensive disease management system for HIV / AIDS patients based on multimodal fusion machine learning. DETAILED DESCRIPTION

[0024] It should be noted that the algorithms for data collection, transmission, storage and processing steps not specifically described in the embodiments, as well as the hardware structures, circuit connections, etc. not specifically described can all be implemented through the disclosed content of the prior art.

[0025] In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application for which protection is claimed, but merely represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0026] Example 1: Comprehensive disease management system for HIV / AIDS patients based on multimodal fusion machine learning

[0027] The present invention provides a comprehensive disease management system for HIV / AIDS patients based on multimodal fusion machine learning.

[0028] 1. System Overview

[0029] Figure 1 The system is composed of a comprehensive disease management system for HIV / AIDS patients based on multimodal fusion machine learning, including a user input system 10, a data processing system 20, and a function system 30.

[0030] The input module (i.e., the user input system 10) is configured to collect multimodal data of the patient, including but not limited to: laboratory test data; clinical text records (such as medical record summaries, medical advice, etc.); imaging data and mental health assessment data (such as questionnaire scores, voice emotion analysis results, etc.).

[0031] The data processing module (ie, the data processing system 20 ) is configured to pre-process the collected multimodal data, including data cleaning, standardization, and feature extraction.

[0032] The output module (ie, functional system 30 ) is configured to monitor the patient's health status in real time and generate personalized early warning and intervention recommendations based on the patient's health feature vector and disease management knowledge base.

[0033] 2. Workflow

[0034] Figure 2This is a workflow diagram for a comprehensive disease management system for HIV / AIDS patients that uses multimodal machine learning. The steps include:

[0035] S1. Users use the system to upload and modify their relevant information (including examination results, treatment plans, psychological conditions, location information, and diagnosis and treatment needs) through various media such as text, voice, and pictures;

[0036] S2. The system performs modal fusion processing on the input information of various media based on the multi-module fusion algorithm and converts it into a unified text field format for subsequent use;

[0037] S3. The extracted key information is used as input to a machine learning-based recommendation algorithm, which processes it and outputs the recommendation results. The results are then rendered using a visualization method to ensure a good user experience.

[0038] S4. Compare the health feature vector with the normal range values in the knowledge base to generate a health status report. When an abnormality is detected, a warning signal is generated based on the machine learning algorithm and intervention measures are recommended;

[0039] S5. The system displays patient health status trend charts, early warning information, and personalized intervention suggestions to medical staff through an interactive interface;

[0040] S6. The corresponding node releases computing resources and the process ends.

[0041] In summary, the present invention provides a comprehensive disease management system for HIV / AIDS patients based on multimodal fusion machine learning. By integrating multimodal data acquisition, processing, fusion, and intelligent decision-making, the system features a modular architecture, strong functional scalability, and low deployment and maintenance costs. It enables multi-agent collaborative management and personalized precision intervention on the same platform. The intelligent monitoring and intervention method provided by the present invention, based on multimodal data fusion and machine learning algorithms, dynamically identifies a patient's health status and generates precise health management plans. Combined with distributed storage and encryption technologies, it enables efficient and secure data management and sharing, comprehensively enhancing the intelligence and reliability of disease management.

Claims

1. A comprehensive disease management system for HIV / AIDS patients based on multimodal fusion machine learning, characterized by: include: An input module configured to input multimodal data of a patient; The data processing module is configured to process the multimodal data of the input patient in combination with the database and the knowledge graph; an output module configured to combine the output information of the data processing module with the patient's health feature vector and the disease management knowledge base to generate personalized early warning and intervention suggestions; The data processing module includes a data fusion module, a feature extraction module and an information processing module.

2. The comprehensive disease management system for HIV / AIDS patients according to claim 1, characterized in that: The multimodal data includes examination results, treatment plans, psychological conditions, positioning information and diagnosis and treatment needs.

3. The comprehensive disease management system for HIV / AIDS patients according to claim 1, characterized in that: In the data processing module, multimodal data requires preprocessing, which includes data cleaning, standardization, and feature extraction.

4. The comprehensive disease management system for HIV / AIDS patients according to claim 1, characterized in that: The information processing module connects the information processed by the feature extraction module with the machine learning module and renders it in combination with a visualization method.

5. The comprehensive disease management system for HIV / AIDS patients according to claim 4, characterized in that: The machine learning module is selected from convolutional neural network and Transformer.

6. The comprehensive disease management system for HIV / AIDS patients according to claim 1, characterized in that: The output module compares the health feature vector with the normal range values in the knowledge base to generate a health status report; when an abnormality is detected, an early warning signal is generated based on the machine learning algorithm and intervention measures are recommended; the system displays the patient's health status trend chart, early warning information and personalized intervention suggestions to medical staff through an interactive interface.

Images