Doctor-patient service engineering system built by virtual digital doctor and model training method thereof

Abstract

The invention discloses a doctor-patient service engineering system built by a virtual digital doctor and a model training method of the doctor-patient service engineering system, and belongs to the technical field of doctor-patient services. The doctor-patient service management system is composed of a digital doctor model processing system, a doctor-oriented doctor diagnosis module, a nurse-oriented nursing service module, a patient-oriented patient terminal module and a monitoring management and control module. According to the system, information interaction among a patient, a doctor and a nurse is realized, and the doctor can obtain identity data of the patient, historical illness state data of the patient and a detection result of the patient through the system, so that the doctor is helped to analyze the illness state of the patient; a nurse can guide food requirements, movement indexes and health management of a patient in different stages through a database according to a doctor's advice issued by a doctor, the patient informs the nurse and the doctor of abnormal physical conditions in time, the condition of the patient can be tracked in real time according to abnormal condition information of the patient, and the safety of the patient is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of doctor-patient services, and particularly to a virtual digital doctor for building a doctor-patient service engineering system and its model training method. Background Art

[0002] Doctor-patient service refers to various medical services provided by medical institutions and their medical staff to patients, covering multiple links such as diagnosis, treatment, nursing, and rehabilitation. Its core goal is to ensure the health of patients, improve the quality of medical services, and promote the harmony of doctor-patient relationships;

[0003] With the continuous development of technology, based on digital human technology, it is relatively common to create an AI digital doctor image, and the model that the doctor image provides services covering pre-diagnosis, in-diagnosis, and post-diagnosis for patients. However, the current digital doctor model can only achieve single docking, that is, the communication between the digital doctor and the patient, and cannot achieve the purpose of mutual communication among patients, doctors, and nursing staff. This makes it difficult for various medical services provided by medical institutions and their medical staff to patients to meet the expectations of patients, and the lack of information interaction among patients, doctors, and nursing staff is also not conducive to the medical management of patients. Based on this, a virtual digital doctor for building a doctor-patient service engineering system and its model training method are proposed. Summary of the Invention

[0004] The purpose of the present invention is to solve the problem in the prior art that the digital doctor model can only achieve single docking, that is, the communication between the digital doctor and the patient, and cannot achieve the purpose of mutual communication among patients, doctors, and nursing staff. This makes it difficult for various medical services provided by medical institutions and their medical staff to patients to meet the expectations of patients, and the lack of information interaction among patients, doctors, and nursing staff is also not conducive to the medical management of patients. A virtual digital doctor for building a doctor-patient service engineering system and its model training method are proposed.

[0005] In order to achieve the above purpose, the present invention adopts the following technical solutions:

[0006] A virtual digital doctor for building a doctor-patient service engineering system and its model training method, including a doctor-patient service management system established based on a digital doctor. The doctor-patient service management system is composed of a digital doctor model processing system, a doctor diagnosis module for doctors, a nursing service module for nursing staff, a patient terminal module for patients, and a monitoring and control module;

[0007] The digital doctor model processing system is established based on an AI model, data modeling, and basic algorithms, processes data through a database, and exchanges and transfers data among patients, nursing staff, and doctors through a data platform;

[0008] The doctor diagnosis module is composed of outpatient service, inpatient service, and medication guidance;

[0009] The patient terminal module consists of disease diagnosis explanation, real-time disease tracking, and medication guidance;

[0010] The nursing service module consists of medication management, disease data interaction, and rehabilitation services;

[0011] The monitoring and supervision module is established based on the digital doctor model processing system, and the monitoring and supervision module consists of medication supervision, monitoring management and services, and user complaints and suggestions.

[0012] Preferably, the doctor diagnosis module;

[0013] Outpatient service: Obtain the patient's identity data through networking, obtain the patient's previous disease data, obtain the patient's test results synchronously, and assist the doctor in making outpatient judgments;

[0014] Inpatient service and medication guidance: The doctor designates the daily treatment plan during hospitalization, and classifies and identifies the daily treatment plan according to the nursing staff and patients through the digital doctor model processing system, and interacts the corresponding plans after identification to the corresponding nursing staff and patients respectively

[0015] The daily treatment plan includes but is not limited to doctor's order adjustment, medical record data, examinations and tests, and surgical arrangements.

[0016] Preferably, the patient terminal module;

[0017] Disease diagnosis explanation: Use the digital doctor through the digital doctor model processing system to explain the doctor's disease diagnosis, and retrieve the corresponding disease data in the database for the keywords of the disease diagnosis, and conduct a video explanation of the comprehensive disease and diagnosis results for the patient;

[0018] Real-time disease tracking: According to the daily treatment plan prescribed by the doctor, corresponding to the different disease symptoms in each stage to form a general disease condition, obtain the patient's own disease recovery situation by comparing the disease conditions in each stage, and when the disease condition is abnormal, promptly interact the abnormal information with the doctor, and the doctor can realize real-time tracking of the patient's disease based on the patient's disease abnormal information;

[0019] Medication guidance: Use the digital doctor through the digital doctor model processing system to conduct a video explanation of the doctor's order for medication guidance. The patient can query the doctor's order for medication guidance in real time, and conduct timed medication reminders and rehabilitation management through the digital doctor model processing system.

[0020] Preferably, the nursing service module:

[0021] The rehabilitation service: Based on the digital doctor model processing system, it identifies the condition according to the doctor's order, and through the database, it guides the patient's food requirements, exercise indicators, and health management at different stages to form a guidance guideline to serve the patient;

[0022] For the disease data interaction and medication management, according to the digital doctor model processing system, it strictly conducts daily care according to the doctor's order data. Facing the patient, it can obtain the patient's health data in the first time, and can synchronously transmit the patient's health data to the doctor through the digital doctor model processing system for judgment, providing professional and efficient care for the patient.

[0023] Preferably, the monitoring and supervision module audits the drugs prescribed by the doctor, checks them against the drug list and its usage instructions in the database, rejects the drugs with excessive use, and requires the doctor to conduct a review to ensure the patient's medication safety;

[0024] Process the patient's in-hospital instrument data, conduct real-time evaluation of the patient's health data, and give an alarm in time when the health data exceeds the threshold;

[0025] Supervise the services between the patient and the doctor, and between the patient and the nursing staff. The patient can file complaints and suggestions about the doctor and the nursing staff through the health supervision module, and it will be corresponded to the relevant personnel through the digital doctor model processing system for review.

[0026] Preferably, the doctor diagnosis module, the nursing service module, and the patient terminal module all establish user interfaces based on the APP server. There are three types of user interfaces, which are respectively used by doctors, nursing staff, and patients.

[0027] The model training method includes the following modules:

[0028] S1. Data training experiment: Transmit the same type of experimental medical samples to the digital doctor model processing system, summarize and check the instructions generated after the processing of the digital doctor model processing system. The patient instructions and nursing staff instructions issued by the same type of experimental medical samples are required to meet the qualification rate, and the model is optimized by minimizing the error between the predicted value and the true label;

[0029] S2. Data storage experiment: Input the medical data into the database to form a new database, select the medical models in the new and old databases for diagnosis, and compare the new and old data processed by the digital doctor model processing system to ensure the identification and verification of the stored data for use

[0030] S3. Virtual digital doctor training: Use the training dataset to train the generator and discriminator of the model for the digital doctor image respectively, and determine the trained image translation model as the digital doctor model. Among them, the image building model is a generative adversarial network model, and the training dataset includes the target frame image, the contour line data of the target frame image, the distance image data of the target frame image, and the first N frame image data of the target frame image.

[0031] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0032] 1. The present invention builds a system based on the existing digital doctor model, conducts data communication among patients, doctors, nurses, and patients. Doctors can obtain patients' identity data, past medical conditions, and test results through the system to help analyze patients' conditions. Nurses can identify patients' conditions based on the doctor's prescriptions and guide patients' food requirements, exercise indicators, and health management at different stages through the database. Patients can understand their conditions, promptly inform nurses and doctors of abnormal physical conditions, and realize real-time tracking of patients' conditions based on abnormal information to ensure patients' safety.

[0033] 2. The present invention simplifies the steps of disease interaction among doctors, nurses, and patients, improves professionalism in the doctor-patient service process, and supervises doctors' prescribing, nurses' medication use, and patients' drug use through the monitoring and supervision module to ensure the safety of drug use. Brief Description of the Drawings

[0034] Figure 1 It is a schematic framework diagram of the doctor-patient service engineering system built by the virtual digital doctor proposed by the present invention;

[0035] Figure 2 It is a model training method diagram of the doctor-patient service engineering system built by the virtual digital doctor proposed by the present invention. Detailed Embodiments

[0036] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0037] In the description of the present invention, it should be noted that the orientation or positional relationship indicated by terms such as "upper", "lower", "inner", "outer", "top / bottom end", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention. In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance.

[0038] Example, refer to Figures 1 to 2 , a virtual digital doctor builds a doctor-patient service engineering system and its model training method. The digital doctor model is an existing large model technology and will not be elaborated here. It includes a doctor-patient service management system established based on the digital doctor. The doctor-patient service management system consists of a digital doctor model processing system, a doctor diagnosis module for doctors, a nursing service module for nursing staff, a patient terminal module for patients, and a monitoring and control module;

[0039] The digital doctor model processing system is established based on the AI model, data modeling, and basic algorithms. It processes data through a database and interacts and transfers the data among patients, nursing staff, and doctors through a data platform;

[0040] The doctor diagnosis module consists of outpatient service, inpatient service, and medication guidance; Outpatient service of the doctor diagnosis module: Obtain the patient's identity data through the network, obtain the patient's previous medical history data, and obtain the patient's test results synchronously to assist doctors in making outpatient judgments;

[0041] Inpatient service and medication guidance: The doctor specifies the daily treatment plan during hospitalization, and through the digital doctor model processing system, classifies and identifies the daily treatment plan according to nursing staff and patients, and respectively interacts the identified corresponding plans to the corresponding nursing staff and patients. The daily treatment plan includes but is not limited to doctor's order adjustment, medical record data, examinations and tests, and surgical arrangements.

[0042] The patient terminal module consists of disease diagnosis explanation, real-time disease tracking, and medication guidance; Disease diagnosis explanation of the patient terminal module: Through the digital doctor model processing system, use the digital doctor to explain the doctor's disease diagnosis, and retrieve the corresponding disease data in the database according to the disease diagnosis keywords to conduct a comprehensive video explanation of the disease and diagnosis results for the patient, establish a basic sense of trust with the patient before the diagnosis, and prepare for efficient medical treatment; During the diagnosis, further popularize medical knowledge to promote face-to-face communication between doctors and patients; After the diagnosis, provide guidance on medication management, rehabilitation nursing, etc. to help the patient recover health as soon as possible;

[0043] Real-time condition tracking: According to the daily treatment plan prescribed by the doctor, corresponding to the different symptoms of the condition in each stage to form the general condition of the disease. By comparing the condition status in each stage, the patient can obtain the recovery situation of their own condition. When the condition status is abnormal, the abnormal information is promptly interacted with the doctor. The doctor can achieve real-time tracking of the patient's condition based on the abnormal information of the patient's condition;

[0044] Medication guidance: Based on the digital doctor model processing system, the digital doctor is used to conduct video explanations of the medical advice for medication guidance. Patients can query the medical advice for medication guidance in real time, and through the digital doctor model processing system, they can receive timed reminders to take medicine and rehabilitation management.

[0045] The nursing service module consists of medication management, condition data interaction, and rehabilitation services;

[0046] Rehabilitation services of the nursing service module: Based on the digital doctor model processing system, the condition is identified according to the doctor's prescribed medical advice, and through the database, guidance is provided for the patient's food requirements, exercise indicators, and health management at different stages, forming a guidance guide to serve the patient;

[0047] Condition data interaction and medication management are carried out strictly according to the doctor's prescribed medical advice based on the digital doctor model processing system. Facing the patient, the health data of the patient can be obtained in the first time, and at the same time, through the digital doctor model processing system, the patient's health data can be transmitted to the doctor for judgment, providing professional and efficient nursing for the patient.

[0048] The monitoring and supervision module is established based on the digital doctor model processing system. The monitoring and supervision module consists of medication supervision, monitoring management and services, and user complaints and suggestions. The monitoring and supervision module reviews the drugs prescribed by the doctor, and checks them against the drug list and its usage instructions in the database. For drugs used beyond the limit, they are rejected and the doctor is required to conduct a review to ensure the safety of the patient's medication;

[0049] Process the patient's in-hospital instrument data, conduct real-time evaluation of the patient's health data, and give an alarm in time when the health data exceeds the threshold;

[0050] Supervise the services between the patient and the doctor, and between the patient and the nursing staff. The patient can file complaints and suggestions about the doctor and the nursing staff through the health supervision module, and through the digital doctor model processing system, it will be corresponded to the relevant personnel for review.

[0051] The doctor diagnosis module, the nursing service module, and the patient terminal module all establish user interfaces based on the APP server. There are three types of user interfaces, which are respectively used by doctors, nursing staff, and patients.

[0052] The present invention establishes a system based on the existing digital doctor model, and exchanges data between patients and doctors, doctors and nursing staff, and nursing staff and patients. Doctors can obtain the patient's identity data, the patient's historical medical data, and the patient's test results through the system to help doctors analyze the patient's condition. Nursing staff can identify the condition according to the doctor's orders, and guide the patient's food requirements, exercise indicators and health management at different stages through the database. Patients can understand their condition and inform nursing staff and doctors of abnormal physical conditions in a timely manner. Real-time tracking of the patient's condition can be achieved based on the patient's abnormal condition information to ensure the patient's safety.

[0053] The present invention simplifies the steps of medical condition interaction between doctors, nurses and patients, improves the professionalism of the medical service process, and supervises the doctor's prescription, nurses' medication and patients' medication through the monitoring and supervision module to ensure safe medication.

[0054] The model training method includes the following modules: Data training test: The same type of experimental medical samples are delivered to the digital doctor model processing system, and the instructions generated after processing by the digital doctor model processing system are summarized and checked. The patient instructions and nursing staff instructions issued by the same type of experimental medical samples are required to meet the qualification rate, and the model is optimized by minimizing the error between the predicted value and the true label;

[0055] Data storage test: input medical data into the database to form a new database, select the medical models in the new and old databases for diagnosis, and compare the new and old data processed by the digital doctor model processing system to ensure the storage data is recognized and used

[0056] Virtual digital doctor training: The generator and discriminator of the digital doctor image model are trained separately using the training data set, and the trained image translation model is determined as the digital doctor model, where the image model is a generative adversarial network model, and the training data set includes the target frame image, the contour data of the target frame image, the distance image data of the target frame image, and the first N frames of the target frame image. There are various model training methods, and the choice depends on the task requirements and data characteristics. The optimization algorithms in this scheme include gradient descent and Adam. During the training process, attention should be paid to overfitting and underfitting problems. Regularization methods such as L2 regularization and Dropout are commonly used for correction. It can identify and judge the reliability of the model through the data aggregation of the nursing service module and the patient terminal module, so as to achieve the purpose of data training experiment.

[0057] The above are only the preferred specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, making equivalent substitutions or changes, shall be covered by the protection scope of the present invention.

Claims

1. Virtual digital doctors build a medical service engineering system, including a medical service management system based on digital doctors, which is characterized by: The medical service management system is composed of a digital doctor model processing system, a doctor diagnosis module for doctors, a nursing service module for nursing staff, and a patient terminal module and a monitoring and control module for patients; The digital doctor model processing system is established based on AI models, data modeling and basic algorithms, performs data processing through a database, and interactively transmits data between patients, caregivers and doctors through a data platform; The doctor diagnosis module consists of outpatient services, inpatient services and medication guidance; The patient terminal module consists of disease diagnosis explanation, real-time disease tracking and medication guidance; The nursing service module consists of medication management, disease data interaction and rehabilitation services; The monitoring and supervision module is established based on the digital doctor model processing system, and the monitoring and supervision module consists of medication supervision, monitoring management and services, and user complaints and suggestions.

2. The virtual digital doctor construction medical service engineering system according to claim 1 is characterized in that: The doctor diagnosis module; Outpatient services: Access the patient's identity data, historical medical data, and test results through the Internet to assist doctors in making outpatient decisions; Inpatient services and medication guidance: The doctor specifies the daily treatment plan during hospitalization, and classifies and identifies the daily treatment plan by nurses and patients through the digital doctor model processing system, and interacts with the corresponding plans to the corresponding nurses and patients respectively; The daily treatment plan includes but is not limited to doctor's order adjustments, medical record data, examinations and tests, and surgical arrangements.

3. The virtual digital doctor construction medical service engineering system according to claim 1 is characterized in that: The patient terminal module; Explanation of disease diagnosis: The digital doctor model processing system uses a digital doctor to explain the doctor's disease diagnosis, and retrieves the corresponding disease data in the database according to the disease diagnosis keywords, and provides a comprehensive video explanation of the patient's disease and diagnosis results; Real-time tracking of the condition: According to the daily treatment plan prescribed by the doctor, the general condition of the condition is formed according to the different symptoms at different stages. By comparing the condition at each stage, the patient's own recovery status can be obtained. When the condition is abnormal, the abnormal information is exchanged with the doctor in time. The doctor can track the patient's condition in real time based on the abnormal information of the patient's condition; Medication guidance: Based on the digital doctor model processing system, the digital doctor will explain the medical instructions of medication guidance through video. Patients can query the medical instructions of medication guidance in real time, and use the digital doctor model processing system for regular medication reminders and rehabilitation management.

4. The virtual digital doctor construction medical service engineering system according to claim 1 is characterized in that: The nursing service module: The rehabilitation service: based on the digital doctor model processing system, the condition is identified according to the doctor's prescription, and the patient's food requirements, exercise indicators and health management at different stages are guided through the database to form a guidance guide to serve the patient; The medical condition data interaction and medication management are based on the digital doctor model processing system and strictly perform daily care according to the doctor's order data according to the doctor's orders. The patient can obtain the patient's health data in the first time, and can simultaneously transmit the patient's health data to the doctor for judgment through the digital doctor model processing system, providing patients with professional and efficient care.

5. The virtual digital doctor construction medical service engineering system according to claim 1 is characterized in that: The monitoring and supervision module reviews the drugs prescribed by doctors and compares them with the drug list and instructions in the database, rejects drugs that exceed the limit, and requires doctors to review them to ensure the safety of patients' medication; Process the patient's inpatient instrument data, conduct real-time evaluation of the patient's health data, and issue an alarm in a timely manner when the health data exceeds the threshold; The services between patients and doctors, and between patients and caregivers are supervised. Patients can make complaints and suggestions to doctors and caregivers through the health supervision module, and the complaints are then processed by the digital doctor model system and directed to relevant personnel for review.

6. The virtual digital doctor construction medical service engineering system according to claim 1 is characterized in that: The doctor diagnosis module, nursing service module and patient terminal module all establish user interfaces based on the APP server. The user interfaces are divided into three types, one for doctors, one for nursing staff and one for patients.

7. The model training method proposed by any one of claims 1 to 6 for building a medical and patient service engineering system using a virtual digital doctor is characterized in that: Includes the following modules: S1. Data training test: The same type of experimental medical samples are sent to the digital doctor model processing system. The instructions generated after processing by the digital doctor model processing system are summarized and checked. The patient instructions and nursing staff instructions issued by the same type of experimental medical samples are required to meet the qualification rate. The model is optimized by minimizing the error between the predicted value and the true label. S2. Data storage test: input medical data into the database to form a new database, select the medical models in the new and old databases for diagnosis, and compare the new and old data processed by the digital doctor model processing system to ensure the storage data is recognized and used S3. Virtual digital doctor training: using the training data set to train the generator and the discriminator of the digital doctor image model respectively, and determine the trained image translation model as the digital doctor model, wherein the image model is a generative adversarial network model, and the training data set includes the target frame image, the contour data of the target frame image, the distance image data of the target frame image and the first N frames of image data of the target frame image.

Images