Intelligent medical service system based on large language model oneself

Abstract

The invention discloses an intelligent medical service system based on large language model oneself. The system comprises a user information acquisition module; an intelligent recommendation engine; a data analysis and evaluation module; an intelligent algorithm and a big data analysis technology are applied, matched medical resources, treatment schemes or doctors are intelligently recommended according to information provided by a user, through an optimization and feedback mechanism, in the fields of common disease diagnosis and chronic disease management, the diagnosis accuracy can reach 85% or above, and the diagnosis accuracy can reach 85% or above. Personalized health management plans and medical consultation suggestions are customized according to personal information of users, recommendation results are updated and optimized in real time through machine learning and data mining, individuation and precision of services are ensured, a system technical architecture supports self-learning and self-adaption, and by accumulating user feedback and historical data, the health management plan and the medical consultation suggestions are made to be more intelligent. A medical recommendation engine and user experience are continuously optimized, the intelligent level of the platform is gradually improved along with the increase of the number of users, and more accurate and efficient medical recommendation is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent medical services. Specifically, it relates to an intelligent medical service system based on the large language model ERNIE Bot. Background Art

[0002] In the development process of the medical service field, the integrated application of artificial intelligence, natural language processing (NLP), and big data technology has brought new opportunities to improve the medical service experience. Currently, a variety of technologies have been applied in medical scenarios. For example, natural language processing technology is used to interact with patients in conversations, answer basic health questions, and provide personalized health consultations; with the help of the Person recommendation algorithm, current users are provided with selection references based on the behavior of historical similar users; local hospital datasets are collected, and with the help of large models, patient needs are converted into retrieval clues to assist patients in choosing hospitals; health data is analyzed through the interaction between wearable devices and the cloud platform to achieve real-time health monitoring and the push of personalized health suggestions.

[0003] However, existing intelligent medical systems have many defects. On the one hand, the information sources of most systems are single, making it difficult to comprehensively consider the personalized needs of patients, resulting in serious information asymmetry problems, and patients cannot obtain the most suitable medical resources for themselves. On the other hand, traditional intelligent medical systems mostly rely on fixed rules or simple data analysis, and perform poorly in terms of the accuracy and humanization of recommendations, and are difficult to understand and handle complex medical problems and the diverse needs of patients. In addition, these systems usually cannot obtain hospital registration and queuing information in real time, and cannot provide optimized medical advice for patients according to the actual situation, resulting in too long waiting times for patients to seek medical treatment. At the same time, existing systems lack effective self-learning and optimization mechanisms, and it is difficult to continuously improve the recommendation scheme based on the treatment feedback of patients. Summary of the Invention

[0004] In view of the problems in the related art, the present invention proposes an intelligent medical service system and method based on the ERNIE Bot large language model, aiming to optimize the patient's medical treatment selection process, improve the utilization efficiency of medical resources, reduce the time cost of patients, relieve the service pressure of hospitals, and effectively solve the prominent problems existing in current medical services, such as difficult hospital selection, information asymmetry, and low patient information matching degree.

[0005] System Architecture: The intelligent medical service system of the present invention mainly consists of a user information collection module, an intelligent recommendation engine, a data analysis and evaluation module, and a user interaction and feedback module.

[0006] User Information Collection Module: Responsible for comprehensively collecting various basic information of patients, including but not limited to personal health status, geographical location, economic ability, medical insurance reimbursement situation, etc. Patients can either manually input information through smartphones or computers, or the system can automatically obtain relevant data through wearable devices, health records, etc., providing accurate and comprehensive basic data support for subsequent intelligent calculations.

[0007] Intelligent Recommendation Engine: As the core technical module of the system, it relies on the Wenxin Big Language Model and uses deep learning and data analysis technologies to process user information. Specifically, the following key factors are comprehensively considered: First, deeply analyze hospital resources, integrating detailed information such as the levels of hospitals across the country, department advantages, expert resources, special surgeries, and word-of-mouth evaluations; Second, real-time collect the number of registered patients in each hospital and the severity of patients' conditions in each department, providing real-time medical advice for patients; Third, precisely match user needs, based on the condition and needs provided by the patient, and combine with hospital resources for precise matching to recommend the most suitable medical service options.

[0008] Data Analysis and Evaluation Module: Screen, sort, and evaluate various data of hospitals and their services. For example, generate a comprehensive score for each hospital and department based on information such as the technical strength of the hospital, patient feedback, and medical insurance policies, helping patients make more rational and scientific medical choices.

[0009] User Interaction and Feedback Module: After receiving the recommendation results, patients can further interact with the intelligent agent to confirm the recommended medical plan or modify it according to their own needs. At the same time, the system uses the Person algorithm to optimize and adjust according to the patient's feedback, providing more personalized recommendation services for subsequent users.

[0010] Data Processing Method (Workflow): Step 1: User Information Input and Storage: The user inputs medical problems and relevant personal information into the system, and the system stores this information in the workflow for subsequent use.

[0011] Step 2: Generation of Personalized Medical Advice: Based on the stored user information, the system uses the intelligent recommendation engine, combining information from multiple aspects such as the Wenxin Big Language Model, hospital resource analysis, registration queue situation, and user need matching, to generate personalized medical advice, including recommending suitable hospitals and departments.

[0012] Step 3: Output and Display of Recommendation Results: Format the generated recommendation information and present it to the user in a clear and understandable manner for the patient to refer to.

[0013] Step 4: User Feedback and System Optimization: The system asks the user if there are any other medical problems. If so, it returns to Step 1 to continue serving the user. If not, it collects the user's feedback information during this service. The system uses the Person algorithm based on the feedback to optimize the recommendation engine, improving the accuracy and personalization of subsequent recommendations.

[0014] Advantages of the Present Invention Social Effects: Improve the accessibility of medical services: Break geographical restrictions through online medical recommendation services, provide convenient medical consultation channels for users in remote areas, rural areas, and those with limited mobility, expand the coverage of social medical resources, and narrow the urban-rural medical gap.

[0015] Relieve the pressure on the medical system: With the help of online consultations, disease screening, and health management services, reduce patients' excessive dependence on hospital outpatient services, relieve the tense situation of medical resources, especially during special periods, and effectively reduce the pressure on hospitals.

[0016] Economic Effects: Reduce medical costs: With accurate medical recommendations, reduce unnecessary examinations and treatments for patients. For example, recommend appropriate diagnosis and treatment plans based on the user's basic information and symptoms, avoid repeated examinations and treatments caused by misdiagnosis or unclear symptoms, and save medical expenses.

[0017] Increase the revenue and efficiency of medical institutions: By cooperating with this platform, medical institutions can expand the coverage of medical services, attract more users for online consultations, and increase revenue. At the same time, use the accurate patient data provided by the platform to optimize service processes and resource allocation, improving service efficiency.

[0018] Technical Effects: Improve the diagnostic accuracy: Use intelligent algorithms and big data analysis technologies to intelligently recommend matching medical resources, treatment plans, or doctors based on the information provided by users. Through multiple rounds of optimization and feedback mechanisms, in the field of common disease diagnosis and chronic disease management, the diagnostic accuracy can reach over 85%.

[0019] Achieve personalized services: Customize personalized health management plans and medical consultation suggestions according to the user's personal information (such as age, gender, health status, living habits, etc.). Through machine learning and data mining, real-time update and optimize the recommendation results to ensure the personalization and precision of services.

[0020] Promote the intelligent upgrade of the platform: The system technical architecture supports self-learning and self-adaptation. By accumulating user feedback and historical data, continuously optimize the medical recommendation engine and user experience. As the number of users increases, the intelligent level of the platform gradually improves, achieving more accurate and efficient medical recommendations. Description of the Drawings

[0021] The accompanying drawings are used to provide a further understanding of the present invention and form a part of the specification. They are used together with the embodiments of the present invention to explain the present invention and do not constitute a limitation to the present invention. In the accompanying drawings: Figure 1 is a schematic flow chart of the intelligent medical service system based on the large language model Wenxin of the present invention; Detailed implementation manners

[0022] Please refer to Figure 1 as shown, which is an embodiment of the present invention.

[0023] The intelligent medical service system based on the large language model Wenxin.

[0024] System construction and deployment: According to the system architecture design, develop and deploy a user information collection module, an intelligent recommendation engine, a data analysis and evaluation module, and a user interaction and feedback module. Select a suitable hardware server or cloud computing platform to ensure the stable operation of the system, and configure a database to store user information, hospital data, and data generated during the operation of the system. During the construction process, fully consider the scalability and compatibility of the system so that subsequent function upgrades and optimizations can be carried out conveniently.

[0025] Data collection and maintenance: The user information collection module continuously collects user information and regularly updates the hospital data set, including hospital resource information, registration queue data, etc. Clean, sort, and preprocess the collected data to remove duplicate data and correct incorrect data to ensure the accuracy and integrity of the data. At the same time, establish a data update mechanism to obtain the latest hospital information and user feedback in a timely manner to ensure the timeliness of system recommendations.

[0026] Model training and optimization: The Wenxin large language model in the intelligent recommendation engine is regularly trained and optimized using a large amount of medical text data, user interaction records, etc. to improve the model's understanding and processing ability of medical problems. At the same time, optimize the deep learning algorithm and data analysis model, adjust the model parameters, and improve the accuracy of recommendations. During the training process, adopt methods such as cross-validation to ensure the stability and reliability of the model.

[0027] System operation and service provision: The system goes online and receives the medical problems and personal information input by users, and provides personalized medical recommendation services for users according to the data processing flow. Monitor the operation status of the system in real time, process user feedback in a timely manner, and ensure the stability and reliability of the service. Establish a user service mechanism to respond to and process users' questions and complaints in a timely manner to improve user satisfaction.

[0028] System evaluation and continuous improvement: Regularly evaluate the system and analyze indicators such as recommendation accuracy and user satisfaction. According to the evaluation results, adjust system parameters and optimize algorithms to continuously improve system performance and service quality. At the same time, pay attention to the latest technological developments in the industry, introduce new technologies and methods in a timely manner, upgrade and improve the system, and maintain the system's advancement and competitiveness.

[0029] Multi-dimensional data fusion innovation: Break through the limitations of the single information source of the existing smart medical system, integrate data from multiple dimensions such as geographic location, severity of illness, medical insurance policy, etc., recommend the most suitable hospital and department for patients through intelligent calculation, fully consider the personalized needs of patients, and effectively solve the problem of information asymmetry.

[0030] Advantages of natural language processing capabilities: By leveraging the powerful natural language processing capabilities of the Wenxin large language model, the system can understand and process medical terms, and naturally generate suggestions that are consistent with patient understanding during conversations with users, thereby improving the humanity and accuracy of recommendations and enhancing the interactivity and intelligence of the system.

[0031] Real-time data optimization service: The system obtains the number of hospital registrations and department queues in real time, and provides real-time optimization suggestions for patients based on the severity of their illness, helping patients avoid peak medical treatment periods, saving waiting time for medical treatment, and significantly improving medical efficiency.

[0032] Construction of self-learning and optimization mechanism: The intelligent agent has long-term memory function. According to the patient's treatment feedback, historical medical records, treatment effects, hospital scores and other information, it continuously optimizes the recommendation algorithm to avoid recommending undesirable hospitals or departments to other patients, and realizes continuous optimization and self-learning of system recommendations.

[0033] The workflow of the entire system is that first, users enter their medical problems into the system, and the system stores this information in a workflow to ensure that it can be used later. Next, the system generates personalized medical advice based on the stored user information. These recommendations will be formatted and printed for the user's reference. The system will then ask the user if they have other medical questions to consult. If so, the process will return to the beginning and continue to generate relevant advice; if not, the process ends. The entire system is designed to provide users with accurate medical recommendations and ensure the continuous use and updating of information. Alternative solutions: Alternative language model solutions: If the Wenxin language model cannot be used, the system can choose other advanced natural language processing models, such as the GPT series, BERT, etc. Although different models differ in processing speed and naturalness of conversation, they can generally provide similar intelligent services to the system.

[0034] Traditional data mining algorithm solution: When resources are limited, the system can adopt a recommendation solution based on traditional data mining algorithms (such as decision trees, KNN, etc.). Although this solution is not as accurate as the deep learning model, it can still provide relatively reliable medical treatment recommendations for patients when the amount of data is small.

[0035] Manual intervention solution: In some special cases, if the system recommendation results cannot fully meet the user's needs, the system provides a manual intervention option for the user to communicate with medical experts one-on-one to obtain personalized professional advice.

[0036] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A smart medical service system based on the Wenxin language model, characterized in that: include: The user information collection module is used to collect various basic information of patients, including personal health status, geographic location, financial ability, and medical insurance reimbursement. Patients can input the information manually through smartphones or computers, and the system can also automatically obtain relevant data through wearable devices and health records to provide basic data support for subsequent intelligent calculations; The intelligent recommendation engine relies on the Wenxin language model, processes user information through deep learning and data analysis, comprehensively considers hospital resource analysis, and integrates the grades, department advantages, expert resources, special surgeries, and word-of-mouth evaluation information of hospitals across the country; Collect the number of patients registered in each hospital and the severity of each department's condition in real time, and provide real-time medical advice; accurately match the condition and needs provided by the patient with the hospital's resources and recommend the most appropriate medical service options; The data analysis and evaluation module screens, sorts and evaluates various types of data on hospitals and their services, and generates comprehensive scores for each hospital and department based on the hospital's technical strength, patient feedback, and medical insurance policy information; User interaction and feedback module: After receiving the recommendation results, patients can further interact with the intelligent agent to confirm the recommended medical plan or modify it according to their own needs. The system uses the Person algorithm to optimize and adjust based on patient feedback to provide more personalized recommendation services to subsequent users.

2. A smart medical service method based on the Wenxin language model, characterized in that: The following steps are involved: User information input and storage: Users input medical issues and related personal information into the system, and the system stores this information in the workflow; Personalized medical advice generation: Based on the stored user information, the system uses an intelligent recommendation engine, combined with Wenxin language model, hospital resource analysis, registration queue situation and user demand matching information to generate personalized medical advice, including recommendations for appropriate hospitals and departments; Output and display of recommendation results: format the generated recommendation information and present it to users in a clear and understandable way; User feedback and system optimization: The system asks the user whether he has other medical problems. If so, it returns to step 1 to continue serving the user. If not, it collects feedback from users during this service, and the system uses the Person algorithm to optimize the recommendation engine based on the feedback.

3. According to claim 1, the intelligent medical service system based on the Wenxin language model is characterized in that: The intelligent recommendation engine processes user information based on the Wenxin language model and makes recommendations by comprehensively considering hospital resources, registration queue conditions and user needs.

4. According to claim 1, the intelligent medical service system based on the Wenxin language model is characterized in that: The data analysis and evaluation module generates comprehensive scores for hospitals and departments based on the hospital's technical strength, patient feedback, medical insurance policies, etc.

5. According to claim 2, the intelligent medical service method based on the Wenxin language model is characterized in that: The user interaction and feedback module uses the Person algorithm to optimize the recommendation engine based on user feedback.

6. According to claim 1, the intelligent medical service system based on the Wenxin language model is characterized in that: When the Wenxin large language model cannot be used, the system uses other advanced natural language processing models, such as the GPT series and BERT as substitutes.

7. The intelligent medical service system based on the Wenxin language model according to claim 1 is characterized in that: When resources are limited, the system uses recommendation schemes based on traditional data mining algorithms, such as decision trees and KNN.

8. The intelligent medical service system based on the Wenxin language model according to claim 1 is characterized in that: In special circumstances where the system recommendation results cannot fully meet user needs, a manual intervention option is provided to allow users to communicate one-on-one with medical experts to obtain personalized professional advice.

Images