Multi-modal data fusion traditional Chinese medicine inquiry and acupoint compatibility method, device and equipment

Abstract

The invention relates to the technical field of traditional Chinese medicine diagnosis and treatment, and discloses a multi-modal data fused traditional Chinese medicine inquiry and acupoint compatibility method, device and equipment, and the method comprises the steps: obtaining the multi-modal data of a user, and carrying out the preliminary diagnosis based on the multi-modal data, and obtaining a preliminary diagnosis result; based on the preliminary diagnosis result, determining missing symptoms in combination with a standard operation process of traditional Chinese medicine diagnosis, and obtaining the missing symptoms through inquiry interaction to form a complete symptom set; reasoning and analyzing the complete symptom set by using a traditional Chinese medicine syndrome knowledge graph to obtain a final diagnosis result; and generating an acupoint compatibility diagnosis and treatment scheme according to the final diagnosis result. By integrating the multi-modal interaction technology and the intelligent analysis algorithm, an intelligent and standardized traditional Chinese medicine inquiry process is established, and the personalized diagnosis and treatment scheme is generated by using the multi-modal data, so that the accuracy and service efficiency of traditional Chinese medicine syndrome differentiation are improved, the time and economic cost of the user is reduced, and the user experience is improved. And a technical framework which can be copied and popularized is provided for online traditional Chinese medicine health services.

Description

Technical Field

[0001] This invention relates to the field of traditional Chinese medicine diagnosis and treatment technology, specifically to a method, device, and equipment for TCM consultation and acupoint matching based on multimodal data fusion. Background Technology

[0002] Traditional Chinese medicine (TCM) diagnosis and treatment, as a core practice of traditional medicine, is characterized by safety and minimal side effects, earning it widespread patient trust. However, TCM diagnosis and treatment require a high level of expertise, making it difficult for non-professionals to grasp the complex diagnostic logic, let alone accurately formulate treatment principles, methods, and acupoint combinations. Traditional TCM transmission relies heavily on apprenticeship or long-term clinical experience accumulation, resulting in ordinary patients and grassroots medical staff lacking access to systematic and standardized diagnostic and treatment guidance.

[0003] While existing online consultation platforms and health advice tools can provide basic information queries, they generally lack specialized support tailored to the characteristics of Traditional Chinese Medicine (TCM) diagnosis and treatment, particularly in scenarios requiring comprehensive information from observation, auscultation, inquiry, and palpation, and dynamic adjustments to treatment principles. Current mainstream intelligent question-answering models, although capable of generating natural language responses, still suffer from inaccurate diagnosis and insufficient guidance when faced with vague symptom descriptions, incomplete information, or user questions straying from the main theme. Furthermore, the failure to effectively integrate multimodal clinical data (such as auscultation sounds, tongue appearance, and facial color) significantly limits the accuracy of diagnostic methods and the personalization of treatment plans. These issues not only affect the convenience for patients to obtain reliable TCM diagnosis and treatment guidance but also hinder the modernization and clinical application of TCM techniques. Summary of the Invention

[0004] This invention provides a method, apparatus, and equipment for TCM consultation and acupoint matching based on multimodal data fusion, in order to solve the problems of inaccurate understanding of user needs, low efficiency of consultation process, and insufficient ability to generate personalized treatment plans.

[0005] In a first aspect, the present invention provides a method for TCM consultation and acupoint combination based on multimodal data fusion, the method comprising: Acquire the user's multimodal data and perform a preliminary diagnosis based on the multimodal data to obtain preliminary diagnostic results; Based on the preliminary diagnostic results, the missing symptoms were identified in accordance with the standard operating procedures of traditional Chinese medicine diagnosis, and the missing symptoms were obtained through interactive consultation to form a complete set of symptoms. The TCM syndrome knowledge graph is used to perform reasoning analysis on the complete symptom set to obtain the final diagnosis result; Acupoint combination treatment plan is generated based on the final diagnosis results.

[0006] The multimodal data fusion method for TCM consultation and acupoint matching provided by this invention integrates multimodal interaction technology and intelligent analysis algorithms to establish an intelligent and standardized TCM consultation process. It uses multimodal data to generate personalized treatment plans, which significantly improves the accuracy and efficiency of TCM diagnosis and service. It not only reduces the time and economic costs for users, but also provides a replicable and scalable technical framework for online TCM health services.

[0007] In one optional implementation, the multimodal data includes: text data, voice data, and image data. A preliminary diagnosis is performed based on the multimodal data to obtain preliminary diagnostic results, including: Based on multimodal data, an intent recognition model is used to identify user intent; If the user's intention is to consult about acupoint selection, the text data, voice data, and image data are processed separately to obtain the diagnostic symptoms and severity level as preliminary diagnostic results.

[0008] In one alternative implementation, if the user's intent is to seek advice on traditional Chinese medicine knowledge or psychological counseling, a search-enhanced question-and-answer mode is initiated to generate and output a response.

[0009] The multimodal data fusion method for TCM consultation and acupoint matching provided by this invention accurately identifies user intent, effectively processes text, voice, and image data to extract diagnostic symptoms and severity levels for acupoint matching consultation needs, and ensures the accuracy of preliminary diagnosis; for TCM common sense consultation or psychological counseling needs, it relies on the search-enhanced question-and-answer generation mode to quickly generate responses and improve the efficiency of handling non-diagnostic needs.

[0010] In one optional implementation, based on the preliminary diagnostic results and in conjunction with the standard operating procedures of traditional Chinese medicine diagnosis, missing symptoms are identified, and these missing symptoms are obtained through interactive consultation to form a complete symptom set, including: Based on the preliminary diagnosis results, multiple preliminary symptoms were identified, and the priority of these preliminary symptoms was ranked according to the first-level standard operating procedure of traditional Chinese medicine diagnosis to obtain a priority sequence of each preliminary symptom. Based on the priority sequence, a pre-set number of high-priority symptoms are determined, and it is detected whether key symptoms are missing from the high-priority symptoms. If key symptoms are missing, follow-up questions will be asked based on the standard operating procedure for secondary diagnosis in Traditional Chinese Medicine to obtain the missing symptoms. Based on the initial symptoms and the missing symptoms obtained, a complete set of symptoms is formed according to the first-level standard operating procedure of traditional Chinese medicine diagnosis.

[0011] The multimodal data fusion method for TCM consultation and acupoint matching provided by this invention combines standard TCM diagnostic procedures, prioritizes preliminary symptoms, specifically detects missing key symptoms, and then follows up with secondary procedures to obtain missing symptoms, ultimately forming a complete symptom set. This ensures the completeness and accuracy of symptom collection and avoids diagnostic bias due to missing symptoms. Simultaneously, it adheres to standardized TCM diagnostic procedures, improving the standardization and professionalism of diagnosis. Furthermore, it optimizes consultation interaction efficiency, accurately focuses on key symptoms, and effectively enhances the overall quality and user experience of intelligent TCM consultation.

[0012] In one optional implementation, a TCM syndrome knowledge graph is used to perform reasoning analysis on the complete symptom set to obtain a final diagnostic result, including: A TCM syndrome knowledge graph is constructed, which includes: symptom nodes and their clinical feature data, syndrome nodes and their diagnostic elements, treatment nodes and their indications, and multidimensional relationships between nodes. Based on the TCM syndrome knowledge graph, a directed weighted network with symptoms and syndromes as nodes is established, and the directed weighted network is used to perform reasoning analysis on the complete symptom set to obtain the final diagnosis result.

[0013] In one optional implementation, a directed weighted network is established based on a Traditional Chinese Medicine (TCM) syndrome knowledge graph, with symptoms and syndromes as nodes, including: Quantitative assessments were conducted based on multiple clinical characteristics of each symptom to determine the number of times each symptom had an effective impact on each syndrome. Based on the number of times each symptom has an effective influence on each syndrome, a random walk model is used for iterative calculation to obtain the convergence weight value of each syndrome node, and a directed weighted network is established based on the convergence weight value.

[0014] The multimodal data fusion method for TCM diagnosis and acupoint matching provided by this invention constructs a knowledge graph covering symptoms, syndromes, treatment nodes, and multidimensional associations. By combining a directed weighted network and a random walk model, it accurately mines the potential associations between symptoms and syndromes, significantly improving the accuracy of the final diagnosis. It realizes the systematization and intelligentization of the TCM syndrome differentiation process, presenting the complex syndrome differentiation logic of TCM in a quantitative and networked manner, and enhancing the standardization and interpretability of diagnosis.

[0015] In one optional implementation, a treatment plan based on acupoint combination is generated according to the final diagnostic result, including: Based on the final diagnosis, an acupoint combination treatment plan is generated using the acupoint rule library. The acupoint combination treatment plan will be fed back to the user through at least one of the following methods: text, voice, image, or video.

[0016] The multimodal data fusion method for TCM consultation and acupoint matching provided by this invention generates a scheme based on the final diagnosis result using an acupoint matching rule library, ensuring the scientific and targeted nature of acupoint matching, conforming to TCM diagnosis and treatment standards, and providing feedback to users through multimodal methods such as text, voice, images, and video, thereby improving the comprehensibility and ease of use of the scheme and meeting the information acquisition habits of different users.

[0017] Secondly, this invention provides a multimodal data fusion device for traditional Chinese medicine consultation and acupoint matching, the device comprising: The preliminary diagnosis module is used to acquire the user's multimodal data and perform a preliminary diagnosis based on the multimodal data to obtain the preliminary diagnosis results; The symptom completion module is used to identify missing symptoms based on the preliminary diagnosis results and in accordance with the standard operating procedures of traditional Chinese medicine diagnosis. It also obtains the missing symptoms through consultation interaction to form a complete symptom set. The reasoning and analysis module is used to perform reasoning and analysis on the complete set of symptoms using the TCM syndrome knowledge graph to obtain the final diagnosis result. The treatment plan generation module is used to generate acupoint combination treatment plans based on the final diagnosis results.

[0018] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the method described in the first aspect or any corresponding embodiment thereof.

[0019] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to perform the method described in the first aspect or any corresponding embodiment thereof. Attached Figure Description

[0020] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0021] Figure 1 This is a schematic diagram of an application scenario according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the first step of the multimodal data fusion method for TCM consultation and acupoint matching according to an embodiment of the present invention. Figure 3This is a schematic diagram of the second process of the multimodal data fusion method for TCM consultation and acupoint matching according to an embodiment of the present invention; Figure 4 This is a schematic diagram of a TCM syndrome knowledge graph in a TCM consultation and acupoint matching method based on multimodal data fusion according to an embodiment of the present invention. Figure 5 This is a schematic diagram of a directed weighted network in the method of TCM consultation and acupoint matching based on multimodal data fusion according to an embodiment of the present invention; Figure 6 This is a complete flowchart of a multimodal data fusion method for TCM consultation and acupoint matching according to an embodiment of the present invention; Figure 7 This is a structural block diagram of a TCM consultation and acupoint matching device based on multimodal data fusion according to an embodiment of the present invention; Figure 8 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0024] As an optional application scenario of this invention, such as Figure 1 As shown, this multimodal data fusion TCM consultation and acupoint matching system may include at least one terminal device and at least one server. Figure 1 The system is illustrated in the example, which includes a computer 101, a mobile terminal 102, and a server 103, and the terminal devices such as the computer 101 and the mobile terminal 102 are connected to the server 103 through a network 110.

[0025] Specifically, the terminal device can be a smartphone, tablet, laptop, PDA, desktop computer, game console, smart TV, smart wearable device, in-vehicle terminal, VR (Virtual Reality) device, AR (Augmented Reality) device, etc. Server 103 can be a standalone physical server, a server cluster, a distributed system, or a cloud server providing cloud services. Network 110 can be a wired or wireless network, examples of which include, but are not limited to, the Internet, corporate intranet, local area network, wide area network, mobile communication network, and combinations thereof.

[0026] This invention provides a method for TCM consultation and acupoint matching based on multimodal data fusion. By integrating multimodal interaction technology and intelligent analysis algorithms, it aims to accurately understand user needs, improve the efficiency of the consultation process, and enhance the ability to generate personalized treatment plans.

[0027] According to an embodiment of the present invention, a method for TCM consultation and acupoint matching based on multimodal data fusion is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0028] This embodiment provides a multimodal data fusion method for TCM consultation and acupoint matching, which can be used in the aforementioned computer system. Figure 2 This is a flowchart of a method for TCM consultation and acupoint matching based on multimodal data fusion according to an embodiment of the present invention, such as... Figure 2 As shown, the process includes the following steps: Step S201: Obtain the user's multimodal data and perform a preliminary diagnosis based on the multimodal data to obtain a preliminary diagnosis result.

[0029] Specifically, users can initiate a consultation request through mobile terminals (including but not limited to WeChat mini programs, standalone applications (APPs) or web-based interactive interfaces) and input data in any or a combination of the following forms: (1) Text input: such as "My child is 4 years old, a girl. She started having a fever and cough yesterday. Today her temperature is 36.3 degrees Celsius, but she is still coughing and has some shortness of breath. What should I do?"; (2) Voice description: record audio of the main symptoms; (3) Image upload: take photos of the child's tongue, complexion, skin, etc.

[0030] An intelligent agent receives multimodal data submitted by users and transmits it to a cloud server for preprocessing. After data cleaning and format standardization, the validity of each modality's data is verified before it is used by the subsequent symptom analysis module. This process is a mature existing technology and will not be elaborated further here.

[0031] A preliminary diagnosis is made based on the information contained in the multimodal data, which determines the symptoms and symptom labels to obtain a preliminary diagnosis result. For example, if the symptoms are fever and cough, the symptom labels include: started yesterday, mild, etc. This is just an example and is not a limitation.

[0032] Step S202: Based on the preliminary diagnosis results, combined with the standard operating procedures of traditional Chinese medicine diagnosis, the missing symptoms are determined, and the missing symptoms are obtained through consultation and interaction to form a complete set of symptoms.

[0033] Specifically, based on the acquired symptoms and their corresponding symptom tag set, and combined with the standard operating procedures for TCM diagnosis, it is determined whether the current multimodal data lacks key symptoms. If so, an active inquiry process is triggered, and a three-stage mechanism is used to complete and optimize the symptoms, thereby dynamically constructing a complete symptom set.

[0034] Step S203: Use the TCM syndrome knowledge graph to perform reasoning analysis on the complete set of symptoms to obtain the final diagnosis result.

[0035] Specifically, a TCM syndrome knowledge graph is constructed using TCM knowledge, including the multidimensional relationships between symptoms, syndromes and treatment principles. Given a standardized set of symptoms, an improved Google Search (PageRank) algorithm is used to calculate node importance and perform reasoning analysis to output diagnostic conclusions that conform to the TCM syndrome differentiation logic.

[0036] Step S204: Generate acupoint combination treatment plan based on the final diagnosis results.

[0037] Specifically, based on the final diagnosis results, the system automatically matches data from the TCM acupoint matching rule base and case base to construct a structured acupoint matching prompt template. This template is then used to generate personalized acupoint matching schemes through a pre-trained TCM big data model. Simultaneously, it integrates multimodal output interfaces for text, video, and voice to provide technical support for subsequent interactive displays.

[0038] The multimodal data fusion method for TCM consultation and acupoint matching provided in this embodiment integrates multimodal interaction technology and intelligent analysis algorithms to establish an intelligent and standardized TCM consultation process. It uses multimodal data to generate personalized treatment plans, which significantly improves the accuracy and efficiency of TCM diagnosis and service. This not only reduces the time and economic costs for users, but also provides a replicable and scalable technical framework for online TCM health services.

[0039] This embodiment provides a multimodal data fusion method for TCM consultation and acupoint matching, which can be used in the aforementioned computer system. Figure 3 This is a flowchart of a method for TCM consultation and acupoint matching based on multimodal data fusion according to an embodiment of the present invention, such as... Figure 3 As shown, the process includes the following steps: Step S301: Obtain the user's multimodal data and perform a preliminary diagnosis based on the multimodal data to obtain a preliminary diagnosis result.

[0040] Specifically, the multimodal data includes: text data, voice data, and image data. A preliminary diagnosis is performed based on the multimodal data to obtain a preliminary diagnostic result. Step S301 includes: Step S3011: Based on at least one type of multimodal data, identify the user's intent using an intent recognition model.

[0041] Specifically, the process begins with user intent recognition, followed by the use of a multimodal large language model to achieve: entity recognition and relation extraction of textual symptoms, semantic parsing and sentiment analysis of speech signals, and image feature extraction and pathological identification. The intent recognition stage employs a machine learning-based intent recognition model (including a classification model and a large language model) to identify and classify user intents, which include inquiries about traditional Chinese medicine knowledge, responses to psychological counseling needs, and consultations based on syndrome differentiation and acupoint selection.

[0042] In step S3012, if the user's intention is to consult on acupoint selection, the at least one type of multimodal data that has been acquired is processed to obtain the diagnostic symptoms and severity level as a preliminary diagnostic result.

[0043] Specifically, the corresponding data processing flow is matched according to the user's intent. When the intent is for syndrome differentiation and acupoint selection consultation, the multimodal data processing flow is activated: symptom entities are extracted from the text through an entity recognition model, and entity normalization is performed according to the "Classification and Code of Traditional Chinese Medicine Diseases". The complex symptom description is decomposed into independent symptom entities, and timestamps and severity levels are marked simultaneously; the speech signal is converted into text through a speech recognition model, non-symptom redundant words are filtered based on natural language processing technology, and standardized terminology mapping for symptom description is achieved with reference to the standard library of TCM terminology, while timestamps and severity levels are marked simultaneously; the texture, color, and other feature parameters of the tongue and complexion are extracted through computer vision algorithms, and a text description conforming to TCM diagnostic terminology is output. The feature parameters are quantified and scored, and timestamps and feature quantification levels are marked simultaneously.

[0044] For example, if a user inputs: "My child is 4 years old, a girl. She started having a fever and cough yesterday. Today her temperature is 36.3 degrees Celsius, but she still coughs and is a little short of breath. What should I do?", and the intent is determined to be a need for acupoint selection based on syndrome differentiation, which falls under the category of diagnostic intent, then the main diagnostic process begins. During the main diagnostic process, at least one modality of data can be used for diagnosis. The input modality data is then structured and parsed. The specific process includes: (1) Text processing: Based on the input text "My child is 4 years old, a girl. She started having a fever and cough yesterday. Today her temperature is 36.3 degrees Celsius, but she is still coughing and has some shortness of breath. What should I do?", we use a multimodal large language model (MLLM) or named entity recognition model to extract symptom entities (such as "fever", "cough", "shortness of breath"). According to the "Classification and Code of Diseases in Traditional Chinese Medicine", "shortness of breath" is mapped to the standardized term "asthma". The compound description is also decomposed (such as "fever and cough" is converted into independent entities "fever" and "cough"). At the same time, the time of symptom occurrence is marked (such as "started yesterday" as the onset time of symptoms, and "today" as the duration of symptoms), and the severity level is quantified according to the description (such as "some shortness of breath" corresponds to mild asthma).

[0045] (2) Speech processing: Based on the input speech signal “the cough sound is very heavy”, after generating text using MLLM or speech recognition (ASR) model, first remove non-symptom words (such as “very”, “have”), then standardize the symptom description: “the cough sound is very heavy” is converted to “the cough sound is heavy and dull”, and at the same time, mark the timestamp of the current time and assess the severity (such as “heavy and dull” corresponds to moderate symptoms).

[0046] (3) Image processing: Based on the input image "tongue image", features are extracted using MLLM or computer vision algorithms (such as color features showing a pale red tongue and a slightly thick white coating). These features are mapped to a standardized description that conforms to TCM diagnosis (such as "pale red tongue with a thick white coating"). The current time is marked with a timestamp (current time). Finally, a quantitative score is given: a pale red tongue is rated as level 1 (normal), and a thick white coating is rated as level 2 (mildly abnormal).

[0047] The agent based on the MLLM architecture will call a dedicated large-scale tongue diagnosis model to process the uploaded tongue photos. For example, the Qwen2.5-VL large-scale model can be fine-tuned and trained using the low-rank adaptation (LoRA) method. The training data covers at least several hundred strictly labeled clinical tongue images and selected public datasets, which can identify key features such as the thickness and color of the tongue coating.

[0048] In some optional implementations, if the user's intention is to seek advice on traditional Chinese medicine knowledge or psychological counseling, a search-enhanced question-and-answer mode is initiated to generate and output the response results.

[0049] Specifically, when the intent is for consultation on basic TCM knowledge or psychological counseling, a sequential question-and-answer mode based on Retrieval Enhancement Generation (RAG) is initiated. This directly generates a response and terminates the subsequent processing flow. For example, if a user enters, "What is the principle behind pediatric massage for treating cough?", intent recognition determines that the user's intent is for TCM knowledge inquiry, not for diagnosis or treatment. In this case, a single-turn RAG question-and-answer mode is activated, skipping the main consultation process. Combining the current dialogue and historical context, a pre-built vector database (containing structured medical knowledge, such as expert experience, clinical cases, and authoritative literature) is retrieved. RAG technology is then used to fuse the most relevant fragments to construct an enhanced prompt, ensuring that the generated content accurately matches the question and outputs a professional answer.

[0050] The multimodal data fusion method for TCM consultation and acupoint matching provided in this embodiment accurately identifies user intent. For acupoint matching consultation needs, it effectively processes text, voice, and image data to extract diagnostic symptoms and severity levels, ensuring the accuracy of preliminary diagnosis. For TCM common sense consultation or psychological counseling needs, it relies on the search-enhanced question-and-answer generation mode to quickly generate responses, improving the efficiency of handling non-diagnostic needs.

[0051] Step S302: Based on the preliminary diagnosis results, combined with the standard operating procedures of traditional Chinese medicine diagnosis, the missing symptoms are determined, and the missing symptoms are obtained through consultation and interaction to form a complete set of symptoms.

[0052] Specifically, step S302 includes: Step S3021: Identify multiple preliminary symptoms based on the preliminary diagnosis results, and sort the priority of the multiple preliminary symptoms according to the first-level standard operating procedure of traditional Chinese medicine diagnosis to obtain the priority sequence of each preliminary symptom.

[0053] Specifically, based on the Standard Operating Procedure (SOP) for TCM diagnosis, when a node matching a user-provided patient symptom is encountered, the identified symptoms are automatically prioritized. For the identified symptoms, dynamic grading and ranking are performed based on their occurrence time and clinical value. The shorter the time interval between the occurrence time and the current time, the higher the symptom priority, meaning the current symptom has a higher priority than historical symptoms. The higher the severity level, the higher the symptom priority, meaning acute symptoms have a higher priority than chronic disease-related symptoms, chronic disease-related symptoms have a higher priority than constitution-regulating symptoms, and acute symptoms are graded according to the degree of impact on vital signs.

[0054] For example, when the node "User has provided symptoms of the child" is matched, the following tagged symptom set is intelligently sorted: {[Fever, started yesterday], [Cough, started yesterday], [Wheezing, today, mild], [Heavy cough, today, moderate], [Pale red tongue, today, normal], [Thick white tongue coating, today, mild]}. Based on the above priority sorting rules, the following sorting results and clinical basis are obtained: (1) Wheezing (today, mild) → acute respiratory condition, priority assessment; (2) Heavy, hoarse cough (today, moderate) → Cough worsens, intervention required; (3) Thick white tongue coating (today, mild) → a secondary basis for auxiliary diagnosis; (4) Pale red tongue (today, normal) → Basic physiological state; (5) Fever (started yesterday, —) → If there is no fever today, the fever can be temporarily relieved; (6) Cough (started yesterday, —) → has been covered by today's symptoms.

[0055] Step S3022: Determine a preset number of high-priority symptoms based on the priority sequence, and detect whether the high-priority symptoms are missing key symptoms.

[0056] Specifically, the top three high-priority symptoms are determined according to the priority sequence (e.g., wheezing, heavy cough, and thick white tongue coating in step S3021), and the TCM knowledge graph is used to determine whether the top three high-priority symptoms are missing key symptoms. For example, wheezing may lack specific time and degree. This is just an example, but not a limitation.

[0057] Step S3023: If key symptoms are missing, follow up with further questions based on the secondary standard operating procedure of TCM diagnosis to obtain the missing symptoms.

[0058] Specifically, based on the priority ranking results, when a high-priority symptom set is found to lack key correlation information that meets preset criteria, a secondary SOP follow-up questioning process is automatically triggered. The follow-up questions include, but are not limited to, symptom duration, aggravating / relieving factors, diurnal rhythm characteristics, and accompanying symptoms. Taking cough as an example, based on the secondary SOP adjacency list, when the system determines that the cough time feature has not been collected, the first round generates a natural language query through a large model (such as "Does your child cough when sleeping at night?"). Based on the user feedback and the secondary SOP logic tree, it dynamically determines whether to initiate progressive follow-up questions. For example, when the user confirms that the cough is not obvious at night, a secondary query of "Does the cough during the day have sputum?" is automatically triggered, and after receiving an affirmative answer, further follow-up questions are asked about features such as "sputum color". This process will continue until the necessary diagnostic dimensions for the symptom are collected.

[0059] Step S3024: Based on the initial symptoms and the acquired missing symptoms, a complete set of symptoms is formed according to the first-level standard operating procedure of traditional Chinese medicine diagnosis.

[0060] Specifically, the system calls the first-level SOP to complete the necessary basic symptoms, dynamically integrates the initial symptoms and the derived symptoms obtained through follow-up questions, generates a complete set of symptoms that conforms to the TCM syndrome differentiation, and marks each symptom with the collection timestamp, severity, and priority.

[0061] Based on the completeness assessment of the current symptom set, the system automatically invokes the first-level SOP process to complete the basic symptom collection, including but not limited to palm temperature, tongue characteristics (tongue body / tongue coating), and bowel and bladder status. The system dynamically adjusts the collected items based on information already provided by the user—for example, if a tongue coating photo has already been uploaded, tongue appearance will not be requested again. Finally, a standardized symptom set conforming to the requirements of Traditional Chinese Medicine (TCM) diagnosis is generated. Each symptom contains the following structured fields: [symptom name, time description, severity, priority]. For example, a complete symptom set might be: {[Wheezing, today, mild, highest priority], [Heavy cough + evening cough, today, moderate, high priority], [Thick white tongue coating, today, mild, medium priority], [Pale red tongue, today, normal, medium priority], [Fever, started yesterday, —, low priority]}. This is just an example and is not a limitation.

[0062] The multimodal data fusion method for TCM consultation and acupoint matching provided in this embodiment combines standard TCM diagnostic procedures, prioritizes preliminary symptoms, and specifically detects missing key symptoms. It then follows a secondary process to retrieve missing symptoms, ultimately forming a complete symptom set. This ensures the completeness and accuracy of symptom collection, avoiding diagnostic bias due to missing symptoms. Furthermore, it adheres to standardized TCM diagnostic procedures, enhancing the standardization and professionalism of diagnosis. It also optimizes consultation interaction efficiency, accurately focuses on key symptoms, and effectively improves the overall quality and user experience of intelligent TCM consultation.

[0063] Step S303: Use the TCM syndrome knowledge graph to perform reasoning analysis on the complete set of symptoms to obtain the final diagnosis result.

[0064] Specifically, step S303 includes: Step S3031: Construct a TCM syndrome knowledge graph. The TCM syndrome knowledge graph includes: symptom nodes and their clinical feature data, syndrome nodes and their diagnostic elements, treatment nodes and their indications, and multidimensional relationships between nodes.

[0065] Specifically, based on multi-dimensional medical knowledge entities, a knowledge graph of traditional Chinese medicine syndromes is constructed, such as... Figure 4The diagram shown is a schematic of a Traditional Chinese Medicine (TCM) syndrome knowledge graph, comprising three core nodes: symptoms, syndromes, and treatment principles. Specifically, it includes: symptom nodes and their clinical characteristic data; syndrome nodes and their diagnostic elements; treatment nodes and their indications; and multidimensional relationships between nodes. Symptom nodes refine clinical attributes such as onset time, severity, and priority; syndrome nodes integrate diagnostic features such as cold, heat, and dampness syndromes; and treatment nodes include treatment principles, methods, and acupoint combinations. Using semantic association modeling technology, diagnostic associations between symptoms and syndromes are established, as well as intervention decision mappings between syndromes and treatments. Furthermore, the potential associations between symptoms and treatments are explored, forming a multi-layered, reasonable TCM syndrome knowledge network, providing structured knowledge support for intelligent clinical decision support. For example... Figure 5 In the text, the symptom "thick white tongue coating" is classified as "cold syndrome"; the complex symptom "heavy and turbid cough with nighttime cough" is associated with "cold syndrome" with a confidence level of 0.8; the treatment principle corresponding to "cold syndrome" is "warming yang", which points to the following acupoint sequence: tonifying spleen → tonifying kidney yang → clearing Tianhe (a point on the back of the body). This is just an example and is not a limitation.

[0066] Step S3032: Based on the TCM syndrome knowledge graph, a directed weighted network with symptoms and syndromes as nodes is established, and the directed weighted network is used to perform reasoning analysis on the complete symptom set to obtain the final diagnosis result.

[0067] Specifically, based on the aforementioned TCM syndrome knowledge graph, a reasoning engine performs multi-dimensional correlation analysis on the symptom set, outputting structured diagnostic conclusions, including the nature of the disease, its location, pathological factors, and information on primary and secondary symptoms, and generating natural language diagnostic descriptions that conform to TCM theoretical expressions. A directed weighted network with symptoms and syndromes as nodes is extracted from the TCM syndrome knowledge graph, such as... Figure 5 The diagram illustrates a directed weighted network. It optimizes diagnostic decisions by quantifying dialectical relationships and automatically generates natural language descriptions of syndrome differentiation that conform to Traditional Chinese Medicine (TCM) theory by inputting structured prompts into a large language model. For example, the final diagnosis might be: "Based on symptom priority, the current primary symptom is asthma (acute exacerbation, deficiency-cold syndrome) accompanied by nighttime cough (cold syndrome), which should be treated by warming Yang; the tongue appearance shows a thick white coating, indicating internal obstruction of phlegm and dampness, further supporting the pathogenesis of cold syndrome. The overall syndrome differentiation indicates that cold syndrome is the primary condition, and the treatment should focus on warming Yang, supplemented by strengthening the spleen and eliminating dampness." This is just an example and is not a limitation.

[0068] In some optional implementations, step S3032 above includes: Step a1: Quantitatively assess the multidimensional clinical characteristics of each symptom to determine the number of times each symptom has an effective impact on each syndrome.

[0069] Specifically, a comprehensive quantitative assessment is conducted based on the multidimensional clinical characteristics of symptoms (including onset time, severity, priority, etc.). First, each characteristic is quantified and scored separately, and then integrated using a weighted model to obtain an initial comprehensive score for each symptom. This score is then normalized to determine the final weight of each symptom. Subsequently, this weight is multiplied by the symptom's confidence level in relation to the syndrome to obtain the number of effective influences of each symptom on a specific syndrome. For example, taking "asthma" as an example, due to its acute onset (today), mild severity, and highest priority, it is calculated to have a weight of 0.8. If its confidence level in relation to "cold syndrome" is 1, then the number of effective influences is 8. In contrast, the symptom "fever," due to its earlier onset (yesterday) and lower priority, only receives a weight of 0.1. If the confidence level of "fever" in relation to "heat syndrome" is also 1, then the number of effective influences is only 1. The number of effective influences directly determines the symptom's influence on the related syndrome, allowing high-weighted symptoms (such as "asthma") to dominate in syndrome differentiation.

[0070] Step a2: Based on the number of times each symptom has an effective influence on each syndrome, iterative calculation is performed using a random walk model to obtain the convergence weight value of each syndrome node, and a directed weighted network is established based on the convergence weight value.

[0071] Specifically, the importance of a syndrome node is positively correlated with the number of effective influences of its incoming edges. Based on the number of effective influences of each symptom on each syndrome, a random walk model is used for iterative calculation to obtain the convergence weight value of each syndrome node, which significantly improves the clinical reasoning ability of the directed weighted network. The convergence weight value is used to reflect the importance of the syndrome node.

[0072] The calculation process of the convergence weight value of each syndrome node includes: assigning the same initial weight to all nodes, and then using the following formula to perform iterative calculation until the result converges, so as to obtain the stable convergence weight value of each syndrome.

[0073]

[0074]

[0075] in, , These represent the iterative weight values ​​for the syndrome node and the symptom node, respectively; It is the damping factor (0 < ≤1), which is the probability that any node will be randomly pointed to; The total number of all nodes in the graph; Represents all pointing symptoms A collection of symptoms; Indicates symptoms For symptoms The number of effective influences, Indicates symptoms The total number of effective effects across all syndromes.

[0076] The multimodal data fusion method for TCM consultation and acupoint matching provided in this embodiment constructs a knowledge graph covering symptoms, syndromes, treatment nodes, and multidimensional associations. By combining a directed weighted network and a random walk model, it accurately mines the potential associations between symptoms and syndromes, significantly improving the accuracy of the final diagnosis. It realizes the systematization and intelligentization of the TCM syndrome differentiation process, presenting the complex syndrome differentiation logic of TCM in a quantitative and networked manner, and enhancing the standardization and interpretability of diagnosis.

[0077] Step S304: Generate acupoint combination treatment plan based on the final diagnosis result.

[0078] Specifically, step S304 includes: Step S3041: Based on the final diagnosis result, generate an acupoint combination treatment plan using the acupoint rule library.

[0079] Specifically, based on the aforementioned diagnostic description and combined with the TCM acupoint combination rule base, a standardized acupoint combination scheme is generated using a pre-trained TCM large model. This scheme includes diagnostic basis, acupoint combination principles, personalized acupoint combinations, operational guidelines, and precautions. It not only integrates the key points of acupoint combination for different disease categories, commonly used acupoint combinations, and expert experience cases from the TCM acupoint combination knowledge base, but also dynamically optimizes the combination weights based on the patient's constitution, age, and other characteristics to ensure the scientific nature and individual suitability of the scheme.

[0080] Step S3042: The acupoint combination treatment plan is fed back to the user through at least one of the following methods: text, voice, image, and video.

[0081] Specifically, the treatment plan is dynamically converted into at least one interactive feedback method, including text descriptions, voice feedback, acupoint location diagrams, and massage operation videos. Multi-dimensional resources from the knowledge base are utilized, including acupoint location diagrams, standardized operation videos (such as a demonstration of the "Pushing Kan Gong" technique), real-time voice guidance, and post-massage care suggestions (such as "Avoid cold and raw foods for 2 hours after massage"). The feedback content of the treatment plan is presented through the user terminal in the form of intelligent voice broadcasts, interactive graphic guidance, and step-by-step short video demonstrations. Key precautions (such as the unsuitability of using warm and pungent Huoxiang Zhengqi Water for summer heat fever) are highlighted in red or displayed in pop-up windows to ensure that users clearly understand the key points of operation and risk warnings. The entire process, while ensuring the rigor of traditional Chinese medicine theory, reduces the difficulty of execution for users through multi-modal methods, improving the usability and safety of the plan.

[0082] The multimodal data fusion method for TCM consultation and acupoint matching provided in this embodiment generates a scheme based on the final diagnosis result using an acupoint matching rule library, ensuring the scientific and targeted nature of the acupoint matching, conforming to TCM diagnosis and treatment standards, and providing feedback to users through multimodal methods such as text, voice, images, and video, thereby improving the comprehensibility and ease of use of the scheme and meeting the information acquisition habits of different users.

[0083] like Figure 6 The diagram illustrates the complete workflow of TCM consultation and acupoint combination methods. Starting with a user initiating a consultation, when a user has health consultation or treatment needs, a request is sent through the interactive interface. Subsequently, an Agent receives user input. The Agent acts as an intelligent interactive proxy, responsible for collecting various types of input information from the user, including text, voice, and images. After collecting user input, the process enters the intelligent semantic analysis stage. Using technologies such as natural language processing and multimodal recognition, the information input by the user is broken down and understood, extracting key information such as symptom descriptions and question types. After semantic analysis, user intent is categorized. This step is a crucial branch point in the process, dividing user needs into two main types: knowledge-based needs and treatment-based needs. For knowledge-based needs, which are non-treatment-based, the RAG (Enhanced Search) generation process is triggered. RAG (Retrieval Augmented Generation) technology retrieves relevant information from a massive professional medical knowledge base, combines it with generative AI to organize and optimize the search results, and finally outputs professional answers to respond to users' knowledge-related questions in an easy-to-understand or professionally rigorous manner. For diagnosis and treatment needs, it enters a multimodal diagnosis and treatment process: first, it integrates multiple types of user input and extracts symptom features through multimodal symptom intelligent recognition, then narrows the scope of diseases through intelligent symptom screening, then clarifies the syndrome type or pathological mechanism through dialectical reasoning and outputs intelligent syndrome differentiation results, then generates a multimodal diagnosis and treatment plan covering multiple dimensions, and if it involves acupoints in traditional Chinese medicine, it makes intelligent recommendations for acupoint combinations, and finally presents the details of the plan to users in a multimodal interactive form.

[0084] This embodiment also provides a multimodal data fusion device for traditional Chinese medicine consultation and acupoint matching. This device is used to implement the above embodiments and preferred embodiments, and details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0085] This embodiment provides a multimodal data fusion device for traditional Chinese medicine consultation and acupoint matching, such as... Figure 7 As shown, it includes: The preliminary diagnosis module 701 is used to acquire the user's multimodal data and perform a preliminary diagnosis based on the multimodal data to obtain a preliminary diagnosis result.

[0086] The symptom completion module 702 is used to determine missing symptoms based on the preliminary diagnosis results and the standard operating procedures of traditional Chinese medicine diagnosis, and to obtain the missing symptoms through consultation interaction to form a complete symptom set.

[0087] The reasoning and analysis module 703 is used to perform reasoning and analysis on the complete set of symptoms using the TCM syndrome knowledge graph to obtain the final diagnosis result.

[0088] The treatment plan generation module 704 is used to generate acupoint combination treatment plans based on the final diagnosis results.

[0089] In some alternative implementations, the preliminary diagnostic module 701 includes: An intent recognition unit is used to recognize user intent based on at least one type of multimodal data using an intent recognition model.

[0090] The acupoint matching response unit is used to process at least one type of multimodal data that has been acquired if the user intends to consult on acupoint matching, and to obtain the diagnostic symptoms and severity level as a preliminary diagnostic result.

[0091] The consultation response unit is used to initiate a search-enhanced question-and-answer mode if the user's intention is to consult about traditional Chinese medicine knowledge or seek psychological counseling, and to generate and output the response results.

[0092] In some alternative implementations, the symptom completion module 702 includes: The preliminary sorting unit is used to identify multiple preliminary symptoms based on the preliminary diagnosis results, and sort the priority of the multiple preliminary symptoms according to the first-level standard operating procedure of traditional Chinese medicine diagnosis to obtain the priority sequence of each preliminary symptom.

[0093] The symptom absence detection unit is used to determine a preset number of high-priority symptoms based on a priority sequence, and to detect whether the high-priority symptoms are missing key symptoms.

[0094] The missing symptom completion unit is used to ask follow-up questions based on the secondary standard operating procedure of traditional Chinese medicine diagnosis to obtain the missing symptoms if key symptoms are missing.

[0095] The complete symptom identification unit is used to form a complete symptom set based on the initial symptoms and the acquired missing symptoms, following the first-level standard operating procedure of traditional Chinese medicine diagnosis.

[0096] In some alternative implementations, the inference analysis module 703 includes: The knowledge graph construction unit is used to construct a TCM syndrome knowledge graph, which includes: symptom nodes and their clinical feature data, syndrome nodes and their diagnostic elements, treatment nodes and their indications, and multidimensional relationships between nodes.

[0097] The network reasoning unit is used to establish a directed weighted network with symptoms and syndromes as nodes based on the TCM syndrome knowledge graph, and to use the directed weighted network to reason and analyze the complete symptom set to obtain the final diagnosis result.

[0098] In some alternative implementations, the network inference unit includes: The quantitative assessment subunit is used to quantitatively assess the multiple clinical characteristics of each symptom and determine the number of times each symptom has an effective impact on each syndrome.

[0099] The iterative calculation subunit is used to perform iterative calculations using a random walk model based on the number of times each symptom has an effective influence on each syndrome, to obtain the convergence weight value of each syndrome node, and to establish a directed weighted network based on the convergence weight value.

[0100] In some optional implementations, the treatment plan generation module 704 includes: The treatment plan generation unit is used to generate acupoint combination and treatment plans based on the final diagnosis results and using the acupoint combination rule library.

[0101] The solution display unit is used to provide feedback to users on acupoint combination treatment plans through at least one of the following methods: text, voice, image, and video.

[0102] The multimodal data fusion TCM consultation and acupoint matching device provided in this embodiment of the invention can execute the multimodal data fusion TCM consultation and acupoint matching method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the above modules and units are the same as in the corresponding embodiments described above, and will not be repeated here.

[0103] Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0104] The following is a detailed reference. Figure 8 This diagram illustrates a suitable structural schematic for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 801, which can perform various appropriate actions and processes based on a program stored in read-only memory (ROM) 802 or a program loaded from memory 808 into random access memory (RAM) 803. The RAM 803 also stores various programs and data required for the operation of the electronic device. The processor 801, ROM 802, and RAM 803 are interconnected via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0105] Typically, the following devices can be connected to I / O interface 805: input devices 806 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 807 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 808 including, for example, magnetic tapes, hard disks, etc.; and communication devices 809. Communication device 809 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 8 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0106] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 809, or installed from a memory 808, or installed from a ROM 802. When the computer program is executed by the processor 801, it performs the functions defined in the multimodal data fusion method for TCM consultation and acupoint matching according to embodiments of the present invention.

[0107] Figure 8 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0108] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as recordable on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the multimodal data fusion method for TCM diagnosis and acupoint matching shown in the above embodiments is implemented.

[0109] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A multi-modal data fusion traditional Chinese medicine consultation and acupoint compatibility method, characterized in that, The method comprises: Obtaining multi-modal data of a user, and performing preliminary diagnosis based on the multi-modal data to obtain a preliminary diagnosis result; Based on the preliminary diagnosis result, missing symptoms are determined in combination with a standard operation process of traditional Chinese medicine diagnosis, and the missing symptoms are obtained through interactive interrogation to form a complete symptom set; Using a traditional Chinese medicine syndrome knowledge graph to perform reasoning analysis on the complete symptom set to obtain a final diagnosis result; Generating an acupoint compatibility diagnosis and treatment scheme according to the final diagnosis result.

2. The method of claim 1, wherein, The multi-modal data includes text data, voice data, and image data. Based on the multi-modal data, a preliminary diagnosis is performed to obtain a preliminary diagnosis result, which includes: According to at least one multi-modal data, a user's intention is identified using an intention recognition model; If the user's intention is acupoint compatibility consultation, the obtained at least one multi-modal data is processed to obtain diagnosis symptoms and severity levels as the preliminary diagnosis result.

3. The method of claim 2, wherein, If the user's intention is traditional Chinese medicine common sense consultation or psychological counseling needs, a question and answer mode based on retrieval enhancement is started to generate a response result and output.

4. The method of claim 1, wherein, Based on the preliminary diagnosis result, missing symptoms are determined in combination with a standard operation process of traditional Chinese medicine diagnosis, and the missing symptoms are obtained through interactive interrogation to form a complete symptom set, which includes: According to the preliminary diagnosis result, a plurality of preliminary symptoms are identified, and the priority of the plurality of preliminary symptoms is sorted according to a first-level standard operation process of traditional Chinese medicine diagnosis to obtain a priority sequence of each preliminary symptom; Based on the priority sequence, a preset number of high-priority symptoms are determined, and it is detected whether the high-priority symptoms are missing key symptoms; If the key symptoms are missing, follow-up questions are asked based on a second-level standard operation process of traditional Chinese medicine diagnosis to obtain missing symptoms; Based on the initial symptoms and the obtained missing symptoms, a complete symptom set is formed according to the first-level standard operation process of traditional Chinese medicine diagnosis.

5. The method of claim 1, wherein, Using a traditional Chinese medicine syndrome knowledge graph to perform reasoning analysis on the complete symptom set to obtain a final diagnosis result, which includes: A traditional Chinese medicine syndrome knowledge graph is constructed, which includes symptom nodes and their clinical feature data, syndrome nodes and their syndrome elements, treatment nodes and their indications, and multi-dimensional association relationships between nodes; Based on the traditional Chinese medicine syndrome knowledge graph, a directed and weighted network with symptom and syndrome nodes is established, and the complete symptom set is analyzed using the directed and weighted network to obtain a final diagnosis result.

6. The method of claim 5, wherein, Based on the traditional Chinese medicine syndrome knowledge graph, a directed and weighted network with symptom and syndrome nodes is established, which includes: According to the multi-clinical features of each symptom, the effective influence times of each symptom on each syndrome are determined; Based on the effective influence times of each symptom on each syndrome, a random walk model is used for iterative calculation to obtain the convergence weight values of each syndrome node, and a directed and weighted network is established based on the convergence weight values.

7. The method of claim 1, wherein, According to the final diagnosis result, an acupoint compatibility diagnosis and treatment scheme is generated using an acupoint compatibility rule library; The acupoint compatibility diagnosis and treatment scheme is fed back to the user in at least one of the following modes: text, voice, image, and video. The device comprises:

8. A traditional Chinese medicine inquiry and acupoint compatibility device for multi-modal data fusion, characterized in that, ​ A preliminary diagnosis module is configured to acquire multi-modal data of a user and perform preliminary diagnosis based on the multi-modal data to obtain a preliminary diagnosis result. A symptom completion module is configured to determine missing symptoms based on the preliminary diagnosis result in combination with a standard operation process of TCM diagnosis, acquire the missing symptoms through an inquiry interaction, and form a complete symptom set. An inference analysis module is configured to perform inference analysis on the complete symptom set by using a TCM syndrome knowledge graph to obtain a final diagnosis result. A diagnosis and treatment scheme generation module is configured to generate an acupoint compatibility diagnosis and treatment scheme according to the final diagnosis result.

9. An electronic device, comprising: The computer readable storage medium stores computer instructions, and the computer instructions are used to make the computer execute the method in any one of claims 1 to 7. The computer readable storage medium stores computer instructions, and the computer instructions are used to make the computer execute the method in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, ​

Images