Intelligent outpatient service auxiliary method and system

Abstract

The invention belongs to the technical field of diagnosis and treatment assistance, and discloses an intelligent outpatient service assistance method and system, and the method comprises the steps: enabling a patient to enter a pre-inquiry interface for questionnaire filling after registration, and generating a pre-inquiry report based on the filling content; based on the pre-inquiry report, doctors of the corresponding department are distributed to carry out formal inquiry; in the formal inquiry process, voice content of a patient and a doctor is collected in real time, the voice content dialogue content is converted into characters, the converted content is analyzed in combination with a large language model, and an outpatient electronic medical record is generated; generating a preliminary diagnosis suggestion based on the outpatient service electronic medical record; and the doctor checks and modifies the preliminary diagnosis suggestions and determines final diagnosis suggestions and follow-up visit plans. According to the technical scheme, the big language model technology and the outpatient service process are fused, the outpatient service efficiency and quality are remarkably improved, the burden of doctors is relieved, and the medical experience of patients is improved.

Description

Technical Field

[0001] This invention belongs to the field of diagnostic and treatment assistance technology, and in particular relates to an intelligent outpatient assistance method and system. Background Technology

[0002] Outpatient clinics are a crucial link in the healthcare system. However, in the current outpatient model, patients experience long waiting times after registration but limited time to communicate with doctors, making it difficult to fully describe their symptoms. Doctors, on the other hand, must complete multiple tasks within a limited time, including consultation, diagnosis, medical record writing, and prescription writing. The standardized writing of electronic medical records consumes a significant amount of time, forcing doctors to reduce their time communicating with patients or extend their working hours to complete paperwork, resulting in considerable workload. While some existing technologies attempt to assist doctors through voice transcription or templated questionnaires, these are mostly isolated tools that fail to organically connect the entire consultation process and lack understanding of natural language semantics, resulting in insufficient intelligence. Therefore, there is an urgent need for an intelligent assistance solution that can be integrated throughout the entire outpatient process. Summary of the Invention

[0003] The purpose of this invention is to provide an intelligent outpatient assistance method and system to solve the problems existing in the prior art.

[0004] To achieve the above objectives, this invention provides an intelligent outpatient assistance method, comprising: S1: After registering, patients enter the pre-consultation interface to fill out a questionnaire, and a pre-consultation report is generated based on the completed questionnaire. The completed questionnaire includes symptoms, duration, aggravating and relieving factors, medical history, and allergy history. The pre-consultation interface is built based on a large language model. S2: Based on the pre-consultation report, doctors from the corresponding departments are assigned to conduct formal consultations; S3: During the formal consultation, the voice content between the patient and the doctor is collected in real time, the voice dialogue is converted into text, and the converted content is analyzed in combination with a large language model to generate an outpatient electronic medical record; preliminary diagnostic suggestions are generated based on the outpatient electronic medical record. S4: The doctor reviews and modifies the preliminary diagnostic recommendations to determine the final diagnostic recommendations and follow-up plan.

[0005] Optionally, in step S1, after registering, the patient enters the pre-consultation interface via a mobile device.

[0006] Optionally, the processing procedure of the pre-consultation interface specifically includes: Based on the pre-set questionnaire requirements, a large language model is invoked to generate a questionnaire list; the content entered by the user according to the questionnaire list is received; the large language model is invoked to understand and reason about the content to obtain a pre-diagnosis report, and a doctor from the corresponding department is matched.

[0007] Optionally, step S3 specifically includes: During the formal consultation, the voice content from the direction of the doctor and the patient is picked up by a microphone array. The collected voice content is then denoised and converted into a text stream. Input the text stream into the medical entity recognition model and label the key entities; The system uses a large language model to understand and reason about the labeled text, and outputs outpatient electronic medical records in a preset format.

[0008] Optionally, step S3 further includes: Doctors review and modify outpatient electronic medical records to obtain revised outpatient electronic medical records. Based on the revised outpatient electronic medical records and preset prompts, they use a large language model to infer preliminary diagnostic suggestions.

[0009] Optionally, the process for determining the final diagnostic recommendation specifically includes: The doctor reviews and modifies the preliminary diagnostic recommendations, takes into account the patient's medication history and allergy history, conducts a safety review of the proposed medications, and determines the final diagnostic recommendations and follow-up plan.

[0010] Optionally, model feedback optimization may also be included, specifically: fine-tuning and optimizing the large language model by collecting feedback from doctors.

[0011] On the other hand, to achieve the above objectives, the present invention provides an intelligent outpatient auxiliary system, comprising: The pre-consultation module is used by patients to fill out a questionnaire after registration, and a pre-consultation report is generated based on the completed questionnaire. The completed questionnaire includes symptoms, duration, aggravating and relieving factors, medical history, and allergy history. The pre-consultation interface is built based on a large language model. The consultation assignment module is used to assign doctors from the corresponding departments to conduct formal consultations based on the pre-consultation report; The diagnostic suggestion generation module is used to collect the voice content between patients and doctors in real time during formal consultations, convert the voice dialogue into text, analyze the converted content using a large language model, and generate outpatient electronic medical records; based on the outpatient electronic medical records, preliminary diagnostic suggestions are generated; doctors review and modify the preliminary diagnostic suggestions to determine the final diagnostic suggestions and follow-up plans.

[0012] The technical effects of this invention are as follows: This invention provides an intelligent outpatient assistance method, comprising: after a patient registers, they enter a pre-consultation interface to complete a questionnaire, and a pre-consultation report is generated based on the completed questionnaire; based on the pre-consultation report, a doctor from the corresponding department is assigned to conduct a formal consultation; during the formal consultation, the voice content between the patient and the doctor is collected in real time, the voice dialogue is converted into text, and the converted content is analyzed using a large language model to generate an electronic medical record; a preliminary diagnostic suggestion is generated based on the electronic medical record; the doctor reviews and modifies the preliminary diagnostic suggestion to determine the final diagnostic suggestion. This invention, by integrating a large language model into the entire outpatient process, significantly improves the efficiency and safety of diagnosis and treatment while freeing doctors from heavy workloads, thus realizing intelligent outpatient services. Attached Figure Description

[0013] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart illustrating the implementation of the method in this embodiment of the invention; Figure 2 This is a schematic diagram of the system structure in an embodiment of the present invention. Detailed Implementation

[0015] Various exemplary embodiments of the present invention will now be described in detail. This detailed description should not be considered as a limitation of the present invention, but rather as a more detailed description of certain aspects, features, and embodiments of the present invention.

[0016] It should be understood that the terminology used in this invention is merely for describing particular embodiments and is not intended to limit the invention. Furthermore, with respect to numerical ranges in this invention, it should be understood that each intermediate value between the upper and lower limits of the range is also specifically disclosed. Every smaller range between any stated value or intermediate value within a stated range, and any other stated value or intermediate value within said range, is also included in this invention. The upper and lower limits of these smaller ranges may be independently included or excluded from the range.

[0017] Various modifications and variations can be made to the specific embodiments described in this specification without departing from the scope or spirit of the invention, as will be apparent to those skilled in the art. Other embodiments derived from this specification will also be obvious to those skilled in the art. This application specification and embodiments are merely exemplary.

[0018] The terms “include,” “including,” “have,” “contain,” etc., used in this article are all open-ended terms, meaning that they include but are not limited to.

[0019] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0020] like Figure 1 - Figure 2 As shown, this embodiment provides an intelligent outpatient assistance method, including: S1: After registering, patients enter the pre-consultation interface to fill out a questionnaire, and a pre-consultation report is generated based on the completed questionnaire. The completed questionnaire includes symptoms, duration, aggravating and relieving factors, medical history, and allergy history. The pre-consultation interface is built based on a large language model. S2: Based on the pre-consultation report, doctors from the corresponding departments are assigned to conduct formal consultations; S3: During the formal consultation, the voice content between the patient and the doctor is collected in real time, the voice dialogue is converted into text, and the converted content is analyzed in combination with a large language model to generate an outpatient electronic medical record; preliminary diagnostic suggestions are generated based on the outpatient electronic medical record. S4: The doctor reviews and modifies the preliminary diagnostic recommendations to determine the final diagnostic recommendations and follow-up plan.

[0021] This embodiment provides an intelligent outpatient assistance method, aiming to solve the problems of long patient waiting times, short effective communication time, and heavy paperwork burden for doctors in the existing outpatient process. The overall process is as follows: Figure 1 As shown, after registering, patients access a pre-consultation interface driven by a large language model via their mobile devices to complete an intelligent questionnaire, generating a structured pre-consultation report. Based on the generated pre-consultation report, patients are accurately assigned to doctors in the corresponding departments. After assignment, they wait for the formal consultation. During the formal consultation, the doctor-patient voice is collected and transcribed in real time. Combined with the large language model, the text is analyzed to automatically generate an outpatient electronic medical record and preliminary diagnostic suggestions. Doctors review and modify the diagnostic suggestions to determine the final diagnosis. This embodiment integrates the entire outpatient process through large language model technology, which can significantly improve outpatient efficiency and quality, reduce the burden on doctors, and improve the patient's medical experience.

[0022] The specific implementation process of this embodiment is as follows: Step S1: Generate a pre-consultation report After patients complete their registration through the hospital's official app or mini-program, a link is pushed to guide them to the pre-consultation interface.

[0023] The pre-consultation interface generates a dynamic, interactive questionnaire list based on preset questionnaire requirements (such as the need to cover symptoms, duration, aggravating and relieving factors, past medical history, allergy history, etc.) by calling a large language model.

[0024] This list is not entirely fixed; the large language model can intelligently follow up with questions based on the patient's initial complaints, thus personalizing the questionnaire process.

[0025] Patients fill out the questionnaire as guided. After completion, the system calls the large language model again to deeply understand and reason about the content based on preset prompts, identify the logical connections between information, and generate a pre-diagnosis report. In addition to the original information, the report also includes preliminary symptom analysis and triage suggestions.

[0026] By conducting a pre-consultation, patients can calmly fill out their medical records, avoiding the tension and forgetfulness that may occur during a face-to-face consultation. This also gives doctors more time to explain and communicate, improving efficiency.

[0027] Step S2: Assign outpatient doctors Based on the pre-consultation report generated by S1, it is assigned to the most suitable specialist physician. For example, if the report analysis indicates angina, it is preferentially assigned to a cardiologist.

[0028] Step S3: Real-time consultation and medical record generation The microphone array deployed in the clinic picks up the voices of doctors and patients in a directional manner, performs noise reduction processing, and then converts the clean voice stream into a text stream in real time through a speech recognition engine.

[0029] The converted text stream is input in real time into a pre-trained medical entity recognition model (BERT-BiLSTM-CRF) to identify and label key medical entities in the text, such as symptoms (S), signs (O), drugs (M), examinations (E), and diagnoses (D).

[0030] The system utilizes a large language model, taking as input the text labeled with medical entities, relevant information from the pre-consultation report, and information entered by the doctor during the consultation. Combined with prompts, the system generates an outpatient electronic medical record according to the basic format of medical record writing. This automatic generation of compliant outpatient electronic medical records significantly reduces the amount of manual writing and time costs for doctors.

[0031] Doctors check and revise the generated draft of the outpatient electronic medical record to confirm the final version; the system inputs the revised final version of the outpatient electronic medical record into a large language model for reasoning and outputs structured preliminary diagnostic suggestions.

[0032] Step S4: Review of Diagnostic Recommendations Doctors will adopt, modify, or reject recommendations based on their clinical experience and the patient's actual situation, and will also review drug interactions and allergy history before determining the final diagnosis and treatment plan.

[0033] This embodiment implements the above process using a large language model and different instructions. The specific training process of the model is as follows: Offline training process: The training data includes medical text data, de-identified historical electronic medical record data, and doctor-patient dialogue texts.

[0034] First, domain pre-training is required to enable the model to understand the professional knowledge, terminology, and expressions in the medical field. Self-supervised learning is then used, where a portion of the words in the text are randomly masked, and the model is asked to predict the masked words. Through this process, the model learns the internal structure and logical relationships of medical knowledge.

[0035] Then, supervised training is used. The text of the doctor-patient dialogue is input into the model, the model generates results, and these results are compared with standard medical records. The loss function is calculated, and the model parameters are adjusted accordingly.

[0036] Online training: The system establishes a feedback mechanism. When doctors modify medical records or diagnostic suggestions, these modifications are anonymized and collected as training data. This data is periodically used to fine-tune the large language model, making the output increasingly consistent with clinical practice and achieving continuous system optimization.

[0037] In summary, this embodiment significantly improves diagnostic and treatment efficiency and safety by integrating a large language model into the entire outpatient process, while freeing doctors from heavy workloads and realizing intelligent outpatient services. Furthermore, this embodiment generates a corresponding follow-up plan based on the final diagnosis and sends reminders to both patients and doctors via SMS or telephone, thereby extending a single outpatient visit into continuous health management and further ensuring medical safety.

[0038] Alternatively, this embodiment also provides an intelligent outpatient auxiliary system, including: The pre-consultation module is used by patients to fill out a questionnaire after registration, and a pre-consultation report is generated based on the completed questionnaire. The completed questionnaire includes symptoms, duration, aggravating and relieving factors, medical history, and allergy history. The pre-consultation interface is built based on a large language model. The consultation assignment module is used to assign doctors from the corresponding departments to conduct formal consultations based on the pre-consultation report; The diagnostic suggestion generation module is used to collect the voice content between patients and doctors in real time during formal consultations, convert the voice dialogue into text, analyze the converted content using a large language model, and generate outpatient electronic medical records; based on the outpatient electronic medical records, preliminary diagnostic suggestions are generated; doctors review and modify the preliminary diagnostic suggestions to determine the final diagnostic suggestions and follow-up plans.

[0039] The above description is merely a preferred embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An intelligent outpatient assistance method, characterized in that, include: S1: After registering, patients enter the pre-consultation interface to fill out a questionnaire, and a pre-consultation report is generated based on the completed questionnaire. The completed questionnaire includes symptoms, duration, aggravating and relieving factors, past medical history, and allergy history. The pre-consultation interface is built based on a large language model. S2: Based on the pre-consultation report, doctors from the corresponding departments are assigned to conduct formal consultations; S3: During the formal consultation, the voice content between the patient and the doctor is collected in real time, the voice content is converted into text, and the converted content is analyzed in combination with a large language model to generate an outpatient electronic medical record; preliminary diagnostic suggestions are generated based on the outpatient electronic medical record. S4: The doctor reviews and modifies the preliminary diagnostic recommendations to determine the final diagnostic recommendations and follow-up plan.

2. The method according to claim 1, characterized in that, In step S1, after registering, the patient enters the pre-consultation interface via a mobile device.

3. The method according to claim 2, characterized in that, The processing procedure of the pre-consultation interface specifically includes: Based on the pre-set questionnaire requirements, a large language model is invoked to generate a questionnaire list; the content entered by the user according to the questionnaire list is received; the large language model is invoked to understand and reason about the content to obtain a pre-diagnosis report, and a doctor from the corresponding department is matched.

4. The method according to claim 1, characterized in that, Step S3 specifically includes: During the formal consultation, the voice content from the direction of the doctor and the patient is picked up by a microphone array. The collected voice content is then denoised and converted into a text stream. Input the text stream into the medical entity recognition model and label the key entities; The system uses a large language model to understand and reason about the labeled text, and outputs outpatient electronic medical records in a preset format.

5. The method according to claim 4, characterized in that, Step S3 further includes: Doctors review and modify outpatient electronic medical records to obtain revised outpatient electronic medical records. Based on the revised outpatient electronic medical records and preset prompts, they use a large language model to infer preliminary diagnostic suggestions.

6. The method according to claim 1, characterized in that, The process of determining the final diagnostic recommendation specifically includes: The doctor reviews and modifies the preliminary diagnostic recommendations, takes into account the patient's medication history and allergy history, conducts a safety review of the proposed medications, and determines the final diagnostic recommendations and follow-up plan.

7. The method according to claim 1, characterized in that, It also includes model feedback optimization, specifically: fine-tuning and optimizing the large language model by collecting feedback from doctors.

8. An intelligent outpatient auxiliary system, characterized in that, include: The pre-consultation module is used by patients to complete a questionnaire after registration, and a pre-consultation report is generated based on the completed questionnaire. The completed questionnaire includes symptoms, duration, aggravating and relieving factors, medical history, and allergy history. The pre-consultation interface is built based on a large language model. The consultation assignment module is used to assign doctors from the corresponding departments to conduct formal consultations based on the pre-consultation report; The diagnostic suggestion generation module is used to collect the voice content between patients and doctors in real time during formal consultations, convert the voice dialogue into text, analyze the converted content using a large language model, generate outpatient electronic medical records, and generate preliminary diagnostic suggestions based on the outpatient electronic medical records. The doctor reviews and modifies the initial diagnostic recommendations to determine the final diagnostic recommendations and follow-up plan.

Images