Traditional chinese medicine auxiliary diagnosis and treatment system based on large model
By calling a large model on the server side to process TCM diagnosis and treatment information, the problem of high computing resource requirements for TCM intelligent robots is solved, enabling rapid deployment and efficient application.
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- CAPITALBIO CORP
- Filing Date
- 2025-11-26
- Publication Date
- 2026-06-18
AI Technical Summary
Deploying large-scale models requires significant computing resources, hindering the rapid deployment and easy application of TCM intelligent robots and increasing costs.
By calling the target large model on the server side, the computational resource requirements of the TCM robot are reduced. The large model fine-tuning module and the calling module are used to process the data on the server side to generate TCM diagnosis and treatment information, and the TCM robot performs the corresponding diagnosis and treatment operations.
This enabled the rapid deployment and efficient application of large-scale models on TCM robots, reducing computing resource requirements and improving data processing efficiency.
Smart Images
Figure CN2025137760_18062026_PF_FP_ABST
Abstract
Description
A Traditional Chinese Medicine Auxiliary Diagnosis and Treatment System Based on a Large Model
[0001] This application claims priority to Chinese Patent Application No. 202411813206.6, filed on December 10, 2024, entitled "A Traditional Chinese Medicine Auxiliary Diagnosis and Treatment System Based on a Large Model", the entire contents of which are incorporated herein by reference. Technical Field
[0002] This application relates to the field of data processing technology, and more specifically to a TCM auxiliary diagnosis and treatment system based on a large model. Background Technology
[0003] The increasing aging population has led to a growing demand for Traditional Chinese Medicine (TCM), driving the intelligent development of TCM robots. These robots can assist patients in treatment by executing corresponding instructions; for example, an acupuncture robot can simulate traditional TCM techniques to treat patients based on specific acupoints. The rise of large-scale modeling technology has provided conditions for TCM robots to develop towards intelligent diagnosis and treatment. However, the deployment of large-scale models requires substantial computing resources, increasing costs and hindering the rapid deployment and easy application of large-scale models in intelligent TCM robots. Summary of the Invention
[0004] In view of the above, this application provides the following technical solution:
[0005] A TCM-assisted diagnosis and treatment system based on a large model, the system comprising:
[0006] Client-side, server-side, and TCM robot; among them,
[0007] The client is used to obtain target symptom information of the target user based on the diagnosis and treatment needs, and send the target symptom information to the server.
[0008] The server is used to respond to the diagnosis and treatment request, call the target large model, so that TCM diagnosis and treatment information matching the target symptom information can be obtained based on the target large model, and send the TCM diagnosis and treatment information to the client.
[0009] The TCM robot is used to execute target diagnostic and treatment operations based on target instructions, wherein the target instructions are execution instructions generated by the client for the TCM robot based on the TCM diagnostic and treatment information.
[0010] Optionally, the system further includes a large model fine-tuning module and a large model calling module, wherein,
[0011] The large model fine-tuning module is used to fine-tune the pre-trained model based on the fine-tuning dataset of the target scene to obtain the target large model;
[0012] The large model invocation module is used by the server to respond to the client's invocation request, invoke the target large model to process the target symptom information, and obtain TCM diagnosis and treatment information.
[0013] Optionally, the server includes a communication interface, a calling interface, and a first data processing module, wherein,
[0014] The communication interface is used to connect with the client and receive the target symptom information from the client;
[0015] The calling interface is used to connect with the large model calling module and send the target symptom information to the large model calling module;
[0016] The first data processing module is used to store the call requests for the target large model corresponding to the target symptom information sent by each of the clients into a request thread pool, so that the target large model can be called based on the thread corresponding to the request thread pool.
[0017] Optionally, the server side further includes a second data processing module, which is used to process the call requests in the request thread pool to obtain a sub-thread corresponding to each call request, wherein the server side calls the target large model based on the sub-thread.
[0018] Optionally, the client includes: an information parsing module;
[0019] The information parsing module is used to parse the TCM diagnosis and treatment information and generate target instructions based on the type of diagnosis and treatment information in the parsing results.
[0020] Optionally, the client further includes: an information collection module and a server-side invocation module;
[0021] The information collection module is used to collect symptom collection information corresponding to the user based on the initial symptom information input by the user, and to generate target symptom information based on the initial symptom information and the symptom collection information.
[0022] The server-side calling module is used to call the communication interface of the server to transmit the target symptom information to the server and receive the TCM diagnosis and treatment information fed back by the server.
[0023] Optionally, the client further includes: a data interaction module;
[0024] The data interaction module is used to transmit the interaction information between the user and the TCM robot to the TCM robot, and the interaction information is used to control the current target diagnosis and treatment operation of the TCM robot.
[0025] Optionally, the TCM robot includes: an information recording module;
[0026] The information recording module is used to record the operation information of the TCM robot when performing the target diagnosis and treatment operation.
[0027] Optionally, the TCM robot further includes: an information output module;
[0028] The information output module is used to output target information corresponding to the target diagnosis and treatment operation, and the target information is at least used to enable the user to obtain the operation progress corresponding to the target diagnosis and treatment operation.
[0029] Optionally, the server may further include: a task monitoring module;
[0030] The task monitoring module is used to monitor the call requests in the request thread pool, obtain monitoring results, and send the monitoring results to the second data processing module, so that the second data processing module generates a sub-thread corresponding to each call request based on the monitoring results.
[0031] As described above, this application discloses a TCM-assisted diagnosis and treatment system based on a large model. The system includes a client, a server, and a TCM robot. The client obtains target symptom information of a target user based on diagnostic needs and sends this information to the server. The server responds to diagnostic needs by invoking a target large model to obtain TCM diagnostic and treatment information matching the target symptom information, and then sends this information to the client. The TCM robot executes target diagnostic and treatment operations based on target instructions, where the target instructions are execution instructions generated by the client for the TCM robot based on the TCM diagnostic and treatment information. This application allows the server to invoke the target large model used to generate diagnostic and treatment information, eliminating the need to deploy the large model on the client or the TCM robot. This effectively reduces the computational resources required for the TCM robot to use the large model, expanding the application scope and data processing efficiency of the large model in TCM robot research. Attached Figure Description
[0032] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only embodiments of this application. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.
[0033] Figure 1 is a schematic diagram of the structure of a TCM auxiliary diagnosis and treatment system based on a large model provided in an embodiment of this application;
[0034] Figure 2 is a schematic diagram of a traditional Chinese medicine diagnosis and treatment system corresponding to an application scenario provided in an embodiment of this application;
[0035] Figure 3 is a flowchart of a multi-threaded call process corresponding to an application scenario provided in an embodiment of this application. Detailed Implementation
[0036] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.
[0037] The terms "first" and "second," etc., used in this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units may include steps or units not listed, but may include steps or units not listed.
[0038] This application provides a large-scale model-based TCM-assisted diagnosis and treatment system. This system can be applied to scenarios where a TCM robot performs corresponding diagnostic and treatment operations, such as assisting patients in rehabilitation or enabling users to use a TCM robot for TCM health maintenance. Referring to Figure 1, which is a schematic diagram of the structure of a large-scale model-based TCM-assisted diagnosis and treatment system provided in this application embodiment, the system may include a client 101, a server 102, and a TCM robot 103.
[0039] The client 101 can be a user terminal for patients requiring TCM diagnosis and treatment, such as a mobile phone, laptop, or other terminal device. Users can upload corresponding symptom information or generate corresponding treatment needs through the client. Specifically, in scenarios where the client interacts with the TCM robot, the client can load an application corresponding to the TCM robot. This application can then transmit relevant information, instructions, or other parameters to the TCM robot, enabling the robot to combine the user's information and perform corresponding diagnostic and treatment operations.
[0040] In this embodiment, the client 101 can communicate with both the server 102 and the TCM robot 103, enabling the client to send relevant information to either the server or the TCM robot. Specifically, the client obtains the target user's target symptom information and sends it to the server. In one embodiment, the user can directly input relevant symptom information into the client, thus identifying the user-inputted symptom information as the target symptom information. Correspondingly, in another embodiment, the client further includes an information collection module. This module collects symptom collection information corresponding to the user based on the initial symptom information input by the user, and generates target symptom information based on the initial symptom information and the symptom collection information. In this embodiment, the user can use their medical needs as the initial symptom information, or they can use partial symptom information as the initial symptom information, where partial symptom information may be information determined by the user based on their own state. Then, the client's information collection module can collect relevant user information based on the initial symptom information to obtain symptom collection information. For example, the information collection module can collect images of the user's tongue and eyes through the client's camera, using these images as symptom collection information. Correspondingly, the information collection module can also collect relevant data such as the user's pulse and heart rate through the client's corresponding human bio-information recording unit. This allows the initial symptom information input by the user to be processed with the collected symptom collection information to obtain target symptom information, making the target symptom information more accurately represent the user's current physical state. Correspondingly, the client also includes a server-side calling module, which is used to call the server's communication interface to transmit the target symptom information to the server. After processing by the server, it forms TCM diagnosis and treatment information, which is then fed back to the server-side calling module. The server-side calling module inputs the TCM diagnosis and treatment information to the client, which can further input it to the client's information parsing module, enabling the client to generate relevant instructions for the TCM robot.
[0041] When server 102 receives the target symptom information sent by the user, it can respond to the user's diagnostic needs, i.e., the user's need for diagnosis and treatment through a TCM robot, by invoking the target large-scale model corresponding to the current application scenario. This allows the server to obtain TCM diagnostic and treatment information matching the target symptom information based on the target large-scale model. The target large-scale model can be a large model obtained by fine-tuning a pre-trained model based on TCM diagnostic and treatment plan datasets from different treatment scenarios, capable of processing symptom information and obtaining corresponding diagnostic and treatment plans. The target large-scale model can be deployed on the server that interacts with the client. This can be done on server 102, on a dedicated server for storing or recording large-scale model information, or on specialized equipment for easy maintenance. Specifically, a calling interface can be deployed on the server side, which calls the target large-scale model to process the symptom information and obtain TCM diagnostic and treatment information. This information is then sent to the client via the server.
[0042] The client can then parse the TCM diagnosis and treatment information to determine the target instructions for controlling the TCM robot to perform the desired diagnostic and treatment operations. The client can break down the diagnosis and treatment information according to its type to obtain the corresponding target instructions. The TCM robot then executes these target instructions, for example, by performing acupuncture on the patient. Specifically, the TCM robot can automatically complete acupoint selection and treatment processes, standardizing and reproducing moxibustion techniques such as meridian-based, cyclical, and pecking methods. It can also have functions such as automatic acupoint location, intelligent acupoint matching, needle insertion, and simulation of human manipulation techniques.
[0043] This allows for the direct deployment of the target large model on the server side, eliminating the need to deploy it on the client or the TCM robot. This effectively reduces the computational resources required when the TCM robot calls the large model, promotes the application of large model technology in TCM robots, and enables the rapid deployment of large models on TCM intelligent robots.
[0044] In one embodiment of this application, the large-model-based TCM auxiliary diagnosis and treatment system further includes a large-model fine-tuning module and a large-model invocation module. The large-model fine-tuning module is used to fine-tune the pre-trained model based on the target scenario's fine-tuning dataset to obtain the target large model. The large-model invocation module is used to respond to client invocation requests on the server side, invoking the target large model to process the target symptom information and obtain TCM diagnosis and treatment information.
[0045] The large model fine-tuning module and the large model invocation module can be configured on the server side of the large model-based TCM auxiliary diagnosis and treatment system, or on other servers, such as cloud servers. This allows the large model invocation module to be called via the server-side interface, enabling the processing of target symptom information to obtain TCM diagnosis and treatment information. Specifically, the large model fine-tuning module is mainly used to fine-tune the pre-trained model. When fine-tuning the pre-trained model, the current TCM dataset from the TCM diagnosis and treatment system can be used as the fine-tuning dataset. This dataset allows the pre-trained model to learn relevant information from the TCM dataset, enabling it to automatically analyze and process TCM-related data. Specifically, the fine-tuning dataset can be books, medical records, literature, and other materials related to TCM external treatment methods, creating a question-and-answer TCM treatment plan dataset from symptom information. Correspondingly, more specific datasets can be obtained as fine-tuning datasets based on the classification of application scenarios. For example, if a TCM robot is mainly used in acupuncture scenarios, the fine-tuning dataset could be acupuncture treatment records from TCM diagnostic data; or, if a TCM robot is mainly used in thermotherapy scenarios, the fine-tuning dataset could mainly be thermotherapy treatment data. During the fine-tuning of the pre-trained model, the model parameters and weights can be adjusted based on the fine-tuning dataset to ensure that the target model meets the needs of the current scenario.
[0046] It should be noted that when the large model is fine-tuned using the large model fine-tuning module, this is done before the server calls the large model. The target large model generated after the large model fine-tuning module fine-tunes the large model can be called by the server. No further model fine-tuning is performed during the use of the target large model; it can be called directly in subsequent calls.
[0047] The large model invocation module is primarily used to respond to client invocation requests. It invokes the target large model to process the target symptom information sent by the client, thereby obtaining TCM diagnostic and treatment information. This TCM diagnostic and treatment information mainly consists of treatment plans corresponding to the target symptom information, such as acupuncture points and treatment duration. In this embodiment, the pre-trained model is fine-tuned using a fine-tuned dataset. This allows for the automatic processing of symptom information using the fine-tuned target large model. The fine-tuned target large model can be directly invoked by the server without further fine-tuning, thus improving the efficiency of model application.
[0048] In one embodiment of this application, the server includes a communication interface, a calling interface, and a first data processing module. The communication interface connects to a client and receives target symptom information from the client; the calling interface connects to a large model calling module and sends the target symptom information to the large model calling module; the first data processing module stores the calling requests for the target large model corresponding to the target symptom information sent by each client into a request thread pool, so that the target large model is called based on the thread corresponding to the request thread pool.
[0049] The server-side of the large-model-based TCM auxiliary diagnosis and treatment system in this embodiment includes a communication interface that can connect to the client to achieve data interaction with the client. The calling interface in the server-side is used to connect to the large-model calling module to achieve data interaction with the target large model. Correspondingly, the server-side also includes a first data processing module, which stores the calling requests to the target large model in a request thread pool to avoid request blocking by establishing a thread pool.
[0050] Furthermore, the server-side also includes a second data processing module. This module processes the call requests in the request thread pool, obtaining a sub-thread corresponding to each request. The server then calls the target large model based on these sub-threads. An asynchronous coroutine is established on the server side. Each time a client makes a call, the server-side module sequentially adds each client's call request to the request queue, generating an event loop object to load all client request tasks. A thread pool is established to prevent blocking, and sub-threads are allocated to the request tasks in the event loop object within the thread pool. The target large model is then called within these sub-threads, achieving concurrent operation through the thread pool. This improves the efficiency of calling the target large model.
[0051] Furthermore, in this embodiment, the server side also includes a task monitoring module. This module monitors the call requests in the request thread pool, obtains monitoring results, and sends the monitoring results to the second data processing module. The second data processing module then generates a sub-thread corresponding to each call request based on the monitoring results. This allows the call thread to be determined based on the monitoring results, satisfying the client's call requests while avoiding problems such as low call efficiency.
[0052] In one embodiment of this application, the client includes an information parsing module. This module parses TCM diagnostic information and generates target instructions based on the type of diagnostic information in the parsing results. The client receives TCM diagnostic information returned by the server, breaks it down and parses it using the information parsing module, and generates target instructions to control the TCM robot based on the type of diagnostic information. Specifically, the types of diagnostic information can be categorized as treatment method, treatment location, treatment technique, treatment duration, etc., thereby generating corresponding target instructions. This allows for precise control of the TCM robot to perform corresponding diagnostic operations, meeting practical application needs and achieving good therapeutic effects.
[0053] Furthermore, in this embodiment, the client also includes a data interaction module. This module transmits the interaction information between the user and the TCM robot to the TCM robot. This interaction information is used to control the current target treatment operation of the TCM robot. The data interaction module can be a voice interaction module, a character interaction module, etc. Specifically, it can be through the voice interaction module in the client, where users can input voice interaction commands to the TCM robot. For example, when the TCM robot is treating a user, the user can input corresponding pause commands or adjustments to the manipulation intensity through the voice interaction module, allowing the TCM robot to respond to the interaction information and adjust the current target treatment operation. This data interaction module can improve the user experience.
[0054] In one embodiment of this application, the TCM robot further includes an information recording module. This module records operational information when the TCM robot performs a target diagnostic or treatment operation. The recorded operational information includes operational information determined according to the target instruction, and may also include user adjustment information, such as the intensity of the user's adjustments, during the treatment operation. Recording this information in the TCM robot allows it to be retrieved the next time the same treatment plan is executed, ensuring rapid execution of the corresponding treatment operation without requiring the input of symptom information each time, thus improving the efficiency of the TCM robot's operation.
[0055] Furthermore, the TCM robot also includes an information output module. This module outputs target information corresponding to the target diagnostic and treatment operation. This target information is at least sufficient to allow the user to obtain the progress of the operation corresponding to the target diagnostic and treatment operation. This information output module can be a voice output module or a display output module. By outputting the corresponding target information, such as the currently performed operation, specifically the selected treatment area and technique, the user can cooperate to complete the current treatment based on these target operations. This allows the user to obtain the progress of the TCM robot's diagnostic and treatment operation, improving the user experience.
[0056] The following describes the TCM-assisted diagnosis and treatment system based on a large model in this application embodiment, using a specific application scenario as an example. Referring to Figure 2, in the application scenario shown in Figure 2, the TCM-assisted diagnosis and treatment system includes a large model fine-tuning module, a large model calling module, a server, a client, and a TCM robot.
[0057] Specifically, the large model fine-tuning module is used to generate a large TCM-assisted decision-making model. It first establishes a TCM treatment plan dataset, collecting TCM treatment information related to massage, acupuncture, and moxibustion. An open-source generative large model is selected as a pre-trained model, and fine-tuned to obtain the TCM-assisted decision-making large model, i.e., the target large model, and its model and weight information are saved. The large model invocation module loads the saved target large model and its weight information. The target large model infers based on the symptom information input by the user through the client, outputting TCM diagnosis and treatment information corresponding to the symptom information, specifically a TCM treatment plan, and returns this plan to the client's server-side invocation module.
[0058] The server first establishes a connection between the client and server using the FastAPI module, providing the client with the server's IP address and open port URL. Simultaneously, an asynchronous coroutine is created on the server, adding each client's request to a request queue and generating an event loop object to load client request tasks. A thread pool is established, allocating threads within the thread pool to the request tasks in the event loop object, and invoking the large model invocation module to achieve multi-threaded concurrent operation. The server inputs the patient's symptom information received from the client into the large model invocation module, which performs inference using the target large model. The large model invocation module outputs the corresponding TCM treatment plan and transmits the plan back to the server. The server then feeds back the TCM treatment plan to the corresponding client. The client receives the patient's symptom information from the user, inputs it to the server via a fixed URL, and waits for the server to return a TCM treatment plan. The client breaks down the information in the returned TCM treatment plan, converts various types of information into machine instructions, and invokes a TCM robot to perform TCM treatment on the patient.
[0059] Referring to Figure 3, this is a flowchart of a multi-threaded call process for an application scenario provided in this embodiment. The server-side builds an API based on a web framework, providing the client with the server's IP address and port for accessing server resources. For simultaneous calls from different clients, an asynchronous coroutine is established on the server side. All client calls are added to a request queue, generating an event loop object to load the request tasks. A thread pool is established, allocating threads to the request tasks in the event loop object within the thread pool. The target model is then invoked, and it uses this model to infer the input patient's symptom information, obtaining a corresponding TCM treatment plan. This plan is then passed to the server, which provides feedback to the client. The client provides the server with the patient's symptom information and receives the TCM treatment plan returned by the server. The client then decomposes and parses the TCM treatment plan information, converting it into machine instructions to control a TCM intelligent robot to perform TCM treatment on the patient.
[0060] For example, client 1 receives symptom information 1 from user 1: lower back pain and weakness, relieved by massage; leg and knee weakness, pain worsens after exertion and improves with rest, often recurring. Client 1 transmits symptom information 1 to the server. Upon receiving symptom information 1, the server adds the request body to the request queue and allocates a thread from the thread pool. The thread calls the large model call module, inputting symptom information 1. The large TCM-assisted decision-making model then infers the corresponding TCM treatment plan 1: Moxibustion: 1. Acupoints: Zhi Shi, San Yin Jiao, Ge Shu, Qi Hai Yu, Tai Chong; 2. Operation: All acupoints are treated with gentle moxibustion, suspended for 7 minutes per acupoint, once daily. The TCM treatment plan 1 is then output to the server, which transmits it to the corresponding client 1. Client 1 receives the returned TCM treatment plan 1, breaks down the information in the plan, and obtains: treatment method: "moxibustion", treatment acupoints: "Zhishi", "Sanyinjiao", "Geshu", "Qihaiyu", "Taichong", treatment technique: "gentle moxibustion", treatment duration: "7 minutes per acupoint", and other treatment information. Based on the treatment information, the TCM intelligent robot is controlled to perform TCM treatment on the patient.
[0061] At a random time during the entire process of client 1, client 2 receives the symptom information input by user 2: joint pain, leg pain, back pain, the location of the pain is not fixed, accompanied by limb swelling and heaviness.
[0062] Client 2 transmits the symptom information to the server. The server receives symptom information 2, adds the request body to the request queue, and allocates a thread. Within the thread, the large model call module is invoked, and symptom information 2 is input into the TCM auxiliary decision-making large model for model inference, resulting in TCM treatment plan 2: Massage: 1. Acupoints: Fenglong, Zusanli, Pishu; 2. Operation: 3 minutes of finger pressure per acupoint. Treatment plan 2 is input to the server, which then transmits it to the corresponding client 2. Client 2 receives the returned treatment plan 2, breaks down the information in the plan, and obtains: Treatment method: "massage", Acupoints: "Fenglong", "Zusanli", "Pishu", Treatment technique: "finger pressure", Treatment duration: "3 minutes per acupoint", etc. Based on the obtained treatment information, the client controls the TCM intelligent robot to perform the corresponding TCM treatment on the patient.
[0063] This application provides a method for reducing computational resource consumption in the application of large models in TCM intelligent robots, which is conducive to the promotion and application of large model technology in TCM intelligent robots and promotes the development of TCM intelligent robots in the direction of intelligence.
[0064] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.
[0065] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.
[0066] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.
[0067] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A TCM-assisted diagnosis and treatment system based on a large model, characterized in that, The system includes: Client-side, server-side, and TCM robot; among them, The client is used to obtain target symptom information of the target user based on the diagnosis and treatment needs, and send the target symptom information to the server. The server is used to respond to the diagnosis and treatment request, call the target large model, so that TCM diagnosis and treatment information matching the target symptom information can be obtained based on the target large model, and send the TCM diagnosis and treatment information to the client. The TCM robot is used to execute target diagnostic and treatment operations based on target instructions, wherein the target instructions are execution instructions generated by the client for the TCM robot based on the TCM diagnostic and treatment information.
2. The system according to claim 1, characterized in that, The system also includes a large model fine-tuning module and a large model calling module, wherein, The large model fine-tuning module is used to fine-tune the pre-trained model based on the fine-tuning dataset of the target scene to obtain the target large model; The large model invocation module is used by the server to respond to the client's invocation request, invoke the target large model to process the target symptom information, and obtain TCM diagnosis and treatment information.
3. The system according to claim 2, characterized in that, The server includes a communication interface, a calling interface, and a first data processing module, wherein... The communication interface is used to connect with the client and receive the target symptom information from the client; The calling interface is used to connect with the large model calling module and send the target symptom information to the large model calling module; The first data processing module is used to store the call requests for the target large model corresponding to the target symptom information sent by each of the clients into a request thread pool, so that the target large model can be called based on the thread corresponding to the request thread pool.
4. The system according to claim 3, characterized in that, The server also includes a second data processing module, which is used to process the call requests in the request thread pool and obtain a sub-thread corresponding to each call request, wherein the server calls the target large model based on the sub-thread.
5. The system according to claim 1, characterized in that, The client includes: an information parsing module; The information parsing module is used to parse the TCM diagnosis and treatment information and generate target instructions based on the type of diagnosis and treatment information in the parsing results.
6. The system according to claim 3, characterized in that, The client also includes: an information collection module and a server-side invocation module; The information collection module is used to collect symptom collection information corresponding to the user based on the initial symptom information input by the user, and to generate target symptom information based on the initial symptom information and the symptom collection information. The server-side calling module is used to call the communication interface of the server to transmit the target symptom information to the server and receive the TCM diagnosis and treatment information fed back by the server.
7. The system according to claim 1, characterized in that, The client also includes: a data interaction module; The data interaction module is used to transmit the interaction information between the user and the TCM robot to the TCM robot, and the interaction information is used to control the current target diagnosis and treatment operation of the TCM robot.
8. The system according to claim 1, characterized in that, The TCM robot includes: an information recording module; The information recording module is used to record the operation information of the TCM robot when performing the target diagnosis and treatment operation.
9. The system according to claim 1, characterized in that, The TCM robot also includes: an information output module; The information output module is used to output target information corresponding to the target diagnosis and treatment operation, and the target information is at least used to enable the user to obtain the operation progress corresponding to the target diagnosis and treatment operation.
10. The system according to claim 4, characterized in that, The server also includes: a task monitoring module; The task monitoring module is used to monitor the call requests in the request thread pool, obtain monitoring results, and send the monitoring results to the second data processing module, so that the second data processing module generates a sub-thread corresponding to each call request based on the monitoring results.