Health consultation processing method and device, equipment, medium and product

Abstract

The invention provides a health consultation processing method and device, equipment, a medium and a product, and the method comprises the steps: responding to a health consultation request of a current user, and determining an actual consultation object corresponding to the health consultation request; a pre-stored health archive library of the actual consultation object is obtained, the health archive library comprises personal portraits and health portraits of the actual consultation object, the personal portraits are obtained through portraying according to personal information of the actual consultation object, and the health portraits are obtained through portraying according to health information of the actual consultation object in all age groups; and generating reply content for the health consultation request according to the health archive library of the actual consultation object. The accuracy and quality of health consultation suggestions can be improved.

Description

Technical Field

[0001] The present application relates to the field of health data processing, and in particular to a health consultation processing method, device, equipment, medium and product. Background Art

[0002] In the field of modern medical health, health consultation has become an important service mode. Through online platforms or mobile applications, users can easily obtain professional medical advice and health guidance.

[0003] However, traditional health consultation methods have limitations in information acquisition and processing, resulting in low accuracy of health advice given to users. Summary of the invention

[0004] Based on the above technical status, the present application provides a health consultation processing method, device, equipment, medium and product, which can improve the accuracy of health advice and thus improve the quality of health advice.

[0005] In order to achieve the above technical objectives, this application specifically proposes the following technical solutions:

[0006] According to the first aspect of an embodiment of the present application, a health consultation processing method is provided, including: in response to a health consultation request of a current user, determining an actual consultation object corresponding to the health consultation request; obtaining a pre-stored health record library of the actual consultation object, the health record library including a personal portrait and a health portrait of the actual consultation object, the personal portrait being obtained by drawing a portrait based on the personal information of the actual consultation object, and the health portrait being obtained by drawing a portrait based on the health information of the actual consultation object at different age groups; and generating reply content for the health consultation request based on the health record library of the actual consultation object.

[0007] In some implementations, the response to the health consultation request of the current user to determine the actual consultation object corresponding to the health consultation request includes: determining the actual consultation object corresponding to the health consultation request by identifying the consultation object role in the health consultation request, the consultation object role being the current user or a bound user corresponding to the current user; when it is identified that the actual consultation object is the bound user of the current user, outputting a first prompt message to switch the current user to the bound user.

[0008] In some implementations, determining the actual consulting object corresponding to the health consultation request by identifying the consulting object role in the health consultation request includes: obtaining multiple modal data about the consulting object in the health consultation request; extracting multiple modal features from the multiple modal data respectively through a multimodal model, and fusing the multiple modal features to obtain fused features; inputting the fused features as independent variables into a logic and regression model, and determining the consulting object role corresponding to the independent variable as the actual consulting object through the logic and regression model according to a pre-learned mapping relationship between the independent variable and the dependent variable, wherein the dependent variable includes multiple preset consulting object roles.

[0009] In some implementations, the generating of reply content to the health consultation request based on the health record library of the actual consultation object includes: when it is determined that further questions need to be asked in response to the health consultation request, determining whether there are answers corresponding to the further questions in the health record library of the actual consultation object; when it is determined that there are answers corresponding to the further questions in the health record library of the actual consultation object, generating reply content to the health consultation request based on the answers corresponding to the further questions stored in the health record library of the actual consultation object.

[0010] In some implementations, the health consultation request includes a condition consultation request. After generating a reply content for the health consultation request based on the health record library of the actual consultation object, the method further includes: generating medical record information for this consultation based on the condition consultation request and the corresponding reply content; and adding the medical record information for this consultation to the health record library of the actual consultation object.

[0011] In some implementations, the health consultation request includes a medication consultation request, and the health record library includes at least one of a current medication plan and a health history; wherein, generating a reply content to the health consultation request based on the health record library of the actual consultation object includes: generating medication suggestion reply content for the current consultation medication in the medication consultation request based on the current medication plan and at least one of the health history of the actual consultation object, and the medication suggestion reply content includes suggestions on whether the current consultation medication needs to avoid medication risks.

[0012] In some implementations, the method further includes: in a case where it is detected that a current test report of the actual consulting object has been uploaded, determining whether there is a historical test report in the health record library of the actual consulting object; if there is a historical test report in the health record library of the actual consulting object, outputting a second prompt information on whether a comparative interpretation is required; in a case where feedback information indicating that a comparative interpretation is required is obtained, generating a comparative interpretation result for the current test report and the historical test report.

[0013] In some implementations, when feedback information requiring comparative interpretation is obtained, a comparative interpretation result is generated for the current test report and the historical test report, including: when feedback information requiring comparative interpretation is obtained, a comparative interpretation result is generated for the current test report and the historical test report based on the health record library of the actual consultation object.

[0014] According to the second aspect of an embodiment of the present application, a health consultation processing device is provided, comprising: a determination unit, for determining, in response to a health consultation request of a current user, an actual consultation object corresponding to the health consultation request; an acquisition unit, for acquiring a pre-stored health record library of the actual consultation object, the health record library including a personal portrait and a health portrait of the actual consultation object, the personal portrait being obtained by drawing a portrait based on the personal information of the actual consultation object, and the health portrait being obtained by drawing a portrait based on the health information of the actual consultation object at different age groups; and a generation unit, for generating reply content for the health consultation request based on the health record library of the actual consultation object.

[0015] According to a third aspect of an embodiment of the present application, an electronic device is provided, comprising a memory and a processor; the memory is connected to the processor and is used to store programs; the processor is used to implement the health consultation processing method as described in the first aspect by running the program in the memory.

[0016] According to a fourth aspect of an embodiment of the present application, a storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the health consultation processing method as described in the first aspect is implemented.

[0017] According to a fifth aspect of an embodiment of the present application, a computer program product is provided, comprising computer program instructions, which, when executed by a processor, cause the processor to execute: the health consultation processing method as described in the first aspect.

[0018] The embodiments of the present application provide a health consultation processing method, device, equipment, medium and product. When receiving the health consultation request of the current user, the method first determines the actual consultation object in the health consultation request, and can accurately identify the specific consultation object in the health consultation request, thereby ensuring the pertinence and accuracy of the subsequent reply processing. After that, the reply content is customized and generated in combination with the health archive of the actual consultation object, which can avoid repeated inquiries for information already in the health archive, thereby improving the interaction efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0020] Figure 1 A flowchart of a health consultation processing method provided in an embodiment of the present application;

[0021] Figure 2 A flowchart of a specific implementation method for determining the actual consulting object provided in an embodiment of the present application;

[0022] Figure 3 A schematic diagram of the principle of determining the actual consultation object based on multimodal data provided in an embodiment of the present application;

[0023] Figure 4 A schematic diagram of building a health archive library provided in an embodiment of the present application;

[0024] Figure 5 A flowchart of updating the health archive provided in an embodiment of the present application;

[0025] Figure 6 A structural schematic diagram of a health consultation processing device provided in an embodiment of the present application;

[0026] Figure 7 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0027] The technical solution provided in the embodiments of the present application can be exemplarily applied to hardware devices such as processors, electronic devices, servers (including cloud servers), or packaged into software programs to be run. When the hardware device executes the processing of the technical solution of the embodiments of the present application, or the above-mentioned software program is run, the automatic splitting of the target task and the automatic calling of the application program interface required for the task can be achieved to complete the purpose of the target task. The embodiments of the present application only exemplarily introduce the specific processing of the technical solution of the present application, and do not limit the specific implementation form of the technical solution of the present application. Any technical implementation form that can execute the processing of the technical solution of the present application can be adopted by the embodiments of the present application.

[0028] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0029] Before introducing this application solution, the relevant technologies are first introduced:

[0030] In the healthcare sector, some medical institutions and health platforms have begun using electronic health record systems to record users' medical histories, test results, and treatment records, but this requires users to actively provide and update information. Users may neglect, forget, or be unwilling to share personal information, resulting in incomplete or inaccurate health records. In addition, users' health data may be scattered across multiple medical institutions and health applications, lacking unified management, making it difficult to quickly obtain comprehensive patient information, affecting the accuracy and efficiency of health consultations.

[0031] In addition, the information provided by users during the initial consultation is often not comprehensive enough, resulting in the need to ask multiple questions to obtain more details. During multiple consultations, the same questions may be asked repeatedly, especially when the user's historical health records are incomplete or unavailable, which will prolong the consultation time and lead to inefficient interaction.

[0032] In summary, the current health consultation methods have many shortcomings in information acquisition, management and processing. There is an urgent need for a new method that can effectively utilize user health records, reduce repeated questions, and improve consultation efficiency and user experience.

[0033] In view of this, the embodiments of the present application are dedicated to providing a health consultation processing method, device, equipment, medium and product, which can effectively utilize the user's health records by establishing a health archive of the user's entire life cycle, reduce repeated questions, and improve consultation efficiency and user experience. The following embodiments will be described in detail one by one.

[0034] Exemplary Methods

[0035] Figure 1 A flowchart of a health consultation processing method provided in an embodiment of the present application. Figure 1 As shown, the health consultation processing method provided in this embodiment includes steps S101-S103:

[0036] S101. In response to a health consultation request of a current user, determine an actual consultation object corresponding to the health consultation request.

[0037] In this embodiment, the current user refers to the user corresponding to the current login account. A health consultation request refers to a request submitted to the medical and health service platform through a mobile application, a web platform, or a telephone hotline, etc., for obtaining professional medical advice or health guidance.

[0038] The current user may correspond to one or more bound users. These bound users are associated with the current user. The bound user may be a family member or friend of the current user. Whether it is the current user himself or his bound user, a health consultation request initiated through the current user's login account is deemed to be a health consultation request initiated by the current user. Since the current user may not be the actual consultant, it is necessary to further determine the actual consultation object in the health consultation request.

[0039] The actual consultation object refers to the user who actually makes the health consultation request. It can be obtained by the current user or bound user selecting a character (current user or bound user) by themselves, or by identifying the character in the health consultation request. The following is a detailed introduction to the implementation method of determining the actual consultation object by identifying the character in the health consultation request of the current user:

[0040] Figure 2 This is a flowchart of a specific implementation method for determining the actual consulting object provided in the embodiment of the present application. Figure 2 As shown, in response to the health consultation request of the current user, determining the actual consultation object corresponding to the health consultation request includes the following steps S201 and S202:

[0041] S201. Determine the actual consulting object corresponding to the health consultation request by identifying the consulting object role in the health consultation request.

[0042] Among them, the consulting object role refers to the character role corresponding to the object of actual health consultation. This role can be the current user or a bound user associated with the current user. Bound users refer to other users who have an association with the current login account. The current user not only has the authority to initiate health consultation services for himself, but also can help other bound users to conduct health consultations. In addition, bound users can also use the current user's login account to initiate health consultations for themselves or others. In order to determine the actual consulting object in the health consultation request, step a1 can be implemented in the following ways:

[0043] In some implementations, the role of the consulting object in the health consultation request can be identified by detecting the role keywords contained in the health consultation request. The role keywords can be keywords such as "I", "my father", "my child", etc. If the role keyword in the health consultation request is "my father", it can be determined that the actual consulting object in the health consultation request is the father of the current user.

[0044] In some implementations, if no role keyword is detected in the health consultation request, the role of the consultation object can be predicted through historical consultation records. For example, if the historical consultation records of the current user indicate that the current user often consults about children's health issues, the role of the consultation object this time can be assumed to be a child.

[0045] In some implementations, by identifying the consulting object role in the health consultation request, determining the actual consulting object corresponding to the health consultation request includes steps a1 to a3:

[0046] Step a1: Acquire multiple modal data about the consulting object role in the health consultation request.

[0047] Figure 3 This is a schematic diagram of the principle of determining the actual consultation object based on multimodal data provided in the embodiment of the present application. Figure 3 As shown, data in multiple modalities may include data in modalities such as photos, faces, voice, and text of the consulting object character.

[0048] For example, when the current user initiates a health consultation request, he or she may provide information such as a photo, voice and text description of the actual consultation object.

[0049] Step a2: extract features from data of multiple modalities respectively through a multimodal model, and fuse the extracted features of multiple modalities to obtain fused features.

[0050] Continue reading Figure 3, the multimodal model can extract useful feature information from different types of modal data. For example, facial features are extracted from photos, voiceprint features are extracted from speech, and semantic features are extracted from text. Then, the extracted features of multiple modalities are fused to generate fused features. For example, facial features, voiceprint features, and semantic features are spliced ​​together to obtain a comprehensive fused feature. Feature fusion can be achieved in many ways, such as splicing, weighted averaging, or deep learning methods.

[0051] Step a3: input the fused features as independent variables into the logistic and regression model, so that the logistic and regression model determines the consulting object role corresponding to the independent variable as the actual consulting object based on the learned mapping relationship between the independent variable and the dependent variable (multiple preset consulting object roles).

[0052] Among them, the logic and regression model determines the consulting object role as the current user himself and the current user's bound user by analyzing the fusion features.

[0053] For example, if the current user takes a photo of his father, the multimodal model and the logic and regression model identify the consulting object in the photo as the father of the user asking the question, and then the user interface or voice prompt will ask whether to switch to the father role. If the current user's father inputs his own voice, the multimodal model and the logic and regression model identify the voice as the voice of the current user's father, and then the user interface or voice prompt will ask whether to switch to the father role.

[0054] In the above example, the father role is the dependent variable, and the father role's multimodal data such as voice, face, photo, text, etc. are independent variables. During the training process, the logic and regression model determines the causal relationship between the dependent variable and the independent variable by specifying the dependent variable and the independent variable. For example, by collecting a large amount of user data as training samples, including the father role's multimodal data such as voice, face, photo and text, and the corresponding consulting object role label (such as father role). The logic and regression model is trained with the training samples to learn the corresponding relationship between the multimodal data and the consulting object role. And solve the various parameters of the model based on the training samples, and evaluate whether the fitting effect of the logic and regression model for the training samples reaches the expected value. If the fitting effect reaches the expected value, it can be applied to role prediction.

[0055] In the above embodiment, when the current user has built a health archive for bound users, the question can be identified to remind whether the question is for his bound user, such as his father, mother or child. If yes, perform the above steps. If not, it can be prompted to help others build a health archive. For example, when it is detected that the current user is helping other users other than the bound user for health consultation, it can be prompted to build a health archive for other users. For example, if the current user asks: What should I do if a 65-year-old suddenly has chest pain? When it is recognized that the current user is 35 years old, it will remind: Are you asking for someone else? Then it can be reminded to build a health archive for other family members. Or, if the current user asks: What should I do if a 6-year-old child itches? Similarly, if it does not match the age of the current user, it can also be reminded to create a health archive for other family members.

[0056] In this embodiment, the actual consultation object can be switched or the user role can be selected according to the specific situation, or it can be switched according to the characteristics of the user recognized, such as voice recognition, face recognition, or according to the user's conversation. When switching according to the characteristics of the user recognized, such as voice recognition, face recognition, or according to the user's conversation, the accuracy of identifying the consultation object in the health consultation request can be improved by utilizing the association information between the data of multiple modes, so as to provide users with more personalized and accurate health consultation services.

[0057] In addition, compared with using a multimodal model alone for role recognition, the efficiency of role recognition can be improved by combining the multimodal model with the logic and regression model.

[0058] Continue reading Figure 2 , after step S201, step S202 may also be included.

[0059] S202: When it is identified that the consulting object is a bound user of the current user, output first prompt information for switching the current user to the bound user.

[0060] This step is to determine the type of the consulting object. If the consulting object is the current user himself, then it is determined that the actual consulting object corresponding to the health consulting request is the current user.

[0061] If the consulting object is a bound user of the current user (such as a family member, relative, or friend), the actual consulting object corresponding to the health consulting request is determined to be a bound user, and then the first prompt information for switching the current user to the bound user is output to prompt the actual consulting object to be switched to the bound user. For example, the health consulting request is "My father moved heavy objects yesterday and his back hurts a little. I don't know if it's a muscle strain." When the actual consulting object is identified as the father of the user asking the question, the first prompt information can be output, such as displaying a "switch" option to ask whether to switch roles.

[0062] Continue reading Figure 1 , after step S101, the health consultation processing method of this embodiment may further include step S102.

[0063] S102: Acquire the pre-stored health record library of the actual consulting object.

[0064] In this embodiment, the health archive includes the personal portrait, health portrait and folder of the actual consultation object.

[0065] Among them, the personal portrait can be obtained based on the personal information of the actual consulting object. Personal information includes name, age, gender, contact information, etc.

[0066] Health profiles can be obtained based on the health information of the actual consulting object at various ages. Health information includes health history, medication plan, review plan, diet advice, exercise advice, etc. The specific content of each of the health history, medication plan, review plan, diet advice and exercise advice can be found in the following Table 1.

[0067] The folder is used to store various health information of the user, such as medical records, report sheets, medicines, physical examination reports, etc., so that it can be easily checked and updated at any time.

[0068] The health record library can be represented as follows: Table 1:

[0069] Table 1 Health records database

[0070]

[0071]

[0072] In order to ensure the accuracy and personalization of health consultation, the construction of a health record library requires the integration of information from multiple sources.

[0073] Figure 4 This is a schematic diagram of building a health archive library provided in the embodiment of this application. Figure 4As shown, by obtaining the historical health data, historical conversation data and self-uploaded health records of the actual consulting object, and then extracting information, a health record library can be constructed. The following are the specific construction steps, including steps b1 to b4:

[0074] Step b1: Obtain the historical health data of the actual consulting object.

[0075] Historical health data can be obtained by collecting historical physical examination reports, historical discharge summaries, historical outpatient medical records, historical examination reports and historical test reports of actual consulting subjects.

[0076] Step b2: Obtain historical conversation data of the actual consulting object.

[0077] Historical conversation data includes online consultation records, telephone consultation records, and face-to-face consultation records.

[0078] Step b3: Obtain the health records of the actual consulting subjects uploaded voluntarily.

[0079] The health record includes the user's personal health log, diet record, exercise record, medication use record and other health information (such as health questionnaires, health assessment reports, etc.).

[0080] Step b4: extract the personal information and health information of the actual consulting object from the historical health data, historical conversation process and self-uploaded health records of the actual consulting object to build a health record library.

[0081] Through the above steps, the health record library of the actual consultation object can be built comprehensively and accurately, providing users with personalized health management and consulting services.

[0082] Among them, the health consultation request includes the identifier of the actual consultation object. The current user corresponds to a health database. The health database includes the health archive of the current user and the health archive of the bound user corresponding to the current user. When the actual consultation object is the current user himself, the health archive of the current user himself is obtained. When the actual consultation object is the bound user of the current user, the health archive of the bound user is obtained.

[0083] S103. Generate a reply to the health consultation request based on the health record database of the actual consultation object.

[0084] After obtaining the health record library of the actual consultation object, the health record library can be combined to generate a response to the health consultation request. The actual consultation scenario may include multiple rounds of questions. In multiple rounds of questions, the health record library can be combined to generate a response to the health consultation request to achieve multiple rounds of interaction with the health record library. The specific implementation process is as follows:

[0085] In some embodiments, step S103 includes step c1 and step c2:

[0086] Step c1: when it is determined that further questions are required for the health consultation request, it is determined whether there are answers corresponding to the further questions in the health record library of the actual consultation object.

[0087] After receiving the health consultation request, it is necessary to evaluate whether further questions are needed for the current health consultation request. If it is determined that further questions are needed, the health record library of the actual consultation object is searched to see whether there are answers corresponding to the questions to be asked.

[0088] For example, user A asks about dizziness. After preliminary analysis, it is considered that more background information is needed, such as whether there is any other medical history, living habits, etc. The health archive of user A will be searched first to see whether there is any information such as medical history and living habits.

[0089] Step c2: when it is determined that the answers to the follow-up questions exist in the health record library of the actual consultation object, a reply content is generated for the health consultation request based on the answers to the follow-up questions stored in the health record library of the actual consultation object.

[0090] If the answers corresponding to the questions to be asked are found in the health archive, the answers are directly extracted and the reply content is generated based on the extracted answers. The reply content will no longer contain the questions to be asked, but directly provide relevant answers to improve the efficiency and accuracy of the consultation.

[0091] For example, if a record of user A having a history of hypertension is found in his health record library, the response content can be: "According to your health record, you have a history of hypertension. Dizziness may be caused by hypertension. It is recommended that you check your blood pressure regularly and follow the doctor's treatment recommendations."

[0092] By combining the above steps c1 and c2 with the user's health record library to generate reply content, a personalized reply can be given each time a user's question is answered, avoiding repeated inquiries for information already in the user's health record library, thereby improving the efficiency of health consultation.

[0093] Generally, health consultation requests may include self-examination of symptoms before seeing a doctor, drug query during medication, and interpretation of examination reports after examination. These different types of requests can be responded to in conjunction with the health archive. The following is a detailed introduction to these different types of requests:

[0094] In some embodiments, the health consultation request may include a condition consultation request. The condition consultation request is a self-examination of symptoms by the user before seeing a doctor, such as entering some symptoms for multiple rounds of consultation, and finally giving the condition analysis results, treatment suggestions and precautions. Figure 5 As shown, the health consultation processing method of this embodiment may also include the following steps S104 and S105:

[0095] S104. Generate medical record information for this consultation based on the medical consultation request and the corresponding reply content.

[0096] Through multiple rounds of interaction according to the medical consultation request, the medical condition analysis results, treatment suggestions and precautions can be obtained. At the same time, the user interface or voice prompts can be used to ask whether to generate a medical record. If feedback to generate a medical record is received, the medical record information of this health consultation is generated.

[0097] For example, when user A consults about health, he mentions: "I have often felt dizzy since last year." The system can automatically identify the symptom of "dizziness" and the time of onset of "since last year." Then, the system can ask user A: "You mentioned that you often felt dizzy since last year. Do you agree to add this information to your health record?"

[0098] S105. Add the medical record information of this consultation to the health record library of the actual consultation object.

[0099] Adding the generated medical record information to the actual consultation object's health record library can ensure that the information in the health record library is updated in real time.

[0100] For example, after user A agrees to add, the symptom of "dizziness" and the onset time of "since last year" will be added to his health record library for use in subsequent health consultations.

[0101] Through the above steps, when the actual consulting object's medical history, onset time and other health information are automatically identified during the conversation, the system will automatically ask whether to add the health information to the health archive. After receiving feedback agreeing to add, the automatically identified health information will be added to the health archive of the actual consulting object, realizing real-time update of the health archive for use in subsequent health consultations.

[0102] In some embodiments, step S104 can be implemented by a disease condition consultation model. The disease condition consultation model can be trained by the following steps:

[0103] First, prompt instructions (prompt word instructions) are generated based on medical knowledge, and the prompt word instructions are input into the Generative Pre-Trained Transformer (GPT) model, Kimi or Qwen-max model to generate multiple rounds of interaction data and condition analysis data about medical consultation requests.

[0104] For example, Q represents questions and A represents replies. The multi-round interaction can be as follows:

[0105] Q: My father was lifting some heavy objects yesterday and his back hurt a bit. I wonder if he had a muscle strain?

[0106] A: What part is it exactly? (Show the back picture)

[0107] Q: Click on the left scapular area.

[0108] A: How severe is your father's back pain? Is it intermittent or constant? Is there any pain when pressing on the area, or any redness, swelling, or bruises?

[0109] Q: It doesn't hurt when I press it, and it's not red or swollen, but it hurts more when I move.

[0110] A: Okay, besides back pain, are there any other symptoms of discomfort? Do you have any other diseases in the past?

[0111] Q: I don’t feel any other discomfort, but I have had high blood pressure for 10 years.

[0112] Disease analysis data:

[0113] Your father developed back pain after lifting heavy objects, which worsened during activities. Considering his age and history of hypertension, the possibility of muscle strain is small, but an atypical attack of angina pectoris cannot be ruled out.

[0114] Common symptoms of angina pectoris are chest pain or chest tightness, and some people may experience radiating pain in the back. At the same time, angina pectoris may be an early sign of **acute myocardial infarction**. Timely diagnosis and treatment are crucial to preventing adverse consequences.

[0115] Disposal suggestions:

[0116] See a cardiologist promptly, complete electrocardiogram and other examinations, avoid fatigue, and eliminate tension and anxiety.

[0117] Note:

[0118] Strictly control blood pressure levels and take medication as directed by your doctor after diagnosis.

[0119] Next, the generated multi-round interaction data and disease conclusion data are manually annotated or rewritten to obtain data for supervised fine-tuning (SFT). These data contain user questions and corresponding correct answers, which are used to guide the large language model (hereinafter referred to as the large model) to learn how to generate more accurate and relevant answers. Specifically, all Qs are used as training samples, and all A, disease analysis data, treatment suggestions and precautions are used as labels to train the large model.

[0120] Afterwards, the large model is trained through supervised learning to obtain the disease consultation model. The trained disease consultation model can then be applied to disease consultation.

[0121] In order to ensure the diversity and personalization of the interaction process and the reply content, the constructed training samples can also be diverse, such as including reply content that combines user portraits, and also including reply content that does not combine user portraits. At the same time, it is necessary to ensure the positive and negative examples of the training samples. This allows the model to not only correctly reply to user questions, but also provide a comprehensive reply for some unconventional questions. In addition, it is also necessary to construct instruction data so that it can reply to various instruction questions of users.

[0122] It is worth noting that during multiple rounds of consultations, in addition to building diverse data and rounds, users can also choose the number of interaction rounds by themselves, giving the user the right to choose and then giving a condition analysis (diagnosis recommendations, treatment recommendations, precautions, etc.) based on the content of the interaction.

[0123] In some embodiments, the health consultation request also includes a medication consultation request, and the health archive includes at least one of the current medication plan and health history. Then step S103 includes: generating medication suggestion reply content for the current consultation medication in the medication consultation request according to at least one of the current medication plan and health history of the actual consultation object, and the medication suggestion reply content includes suggestions on whether the current consultation medication needs to avoid medication risks.

[0124] Among them, medication consultation requests refer to drug inquiries when taking medications, such as giving preliminary responses to some input drugs, guiding users to interact, and finally giving medication guidance. When the user asking the question asks about medication, the current medication plan of the actual consulting object can be used to identify the medication being taken, and the disease history in the health archive can be queried to avoid medication risks and give targeted auxiliary alternative suggestions. That is, when responding to the health consultation request of the user asking the question, information such as the medication plan and health archive can be traced back.

[0125] For example, user A is asking for his mother: "My mother just had hip surgery last week and the wound is a bit painful now. Can she take ibuprofen?" By querying the health archive of user A's mother, it can be found that user A is currently taking other medications and has a history of gastric ulcer. Therefore, a reply suggestion of "It is not recommended that you take ibuprofen for pain relief at present" can be generated, and combined with the traced medication plan and disease history for analysis, and the analysis results (such as: ibuprofen and other drugs have drug interactions, which will increase the risk of bleeding; since your mother has a gastric ulcer, taking ibuprofen may aggravate the symptoms of the gastric ulcer, etc.) can be presented to user A.

[0126] Among them, medication consultation can include asking about the drug's indications, contraindications, drug interactions, drug-food interactions, etc. At the same time, combining the drug with the user's personalized information (such as gender, age, medical history, medication history, allergy history, etc.) can give more personalized medication recommendations that are more in line with the user's actual situation.

[0127] In addition to being implemented through models, medication consultation can also call on internal and external knowledge, such as drug contraindications, indications, interactions, etc., to recommend medications based on the disease or respond based on the drug instructions to provide suggestions on which diseases the user inquires about the drugs for or which diseases they are prohibited from using.

[0128] Medication consultation can be achieved by combining training data + calling plug-ins (i.e. calling authoritative knowledge such as instructions + website knowledge).

[0129] In some embodiments, the health consultation request can also be an inspection consultation request, such as giving a preliminary interpretation based on the input text questions or report sheets, guiding the user to conduct multiple rounds of interaction, obtaining a comprehensive interpretation and giving treatment suggestions and life guidance. Inspection consultation request refers to the inspection and interpretation of the actual consultation object after the examination. The specific implementation method of step S103 includes steps d1 to d3:

[0130] Step d1: when it is detected that the current test report of the actual consulting object is uploaded, it is determined whether there is a historical test report in the health record library of the actual consulting object.

[0131] If the user who asks the question uploads the test report of the actual consulting object, the system will start to detect whether there is a historical test report in the health record library of the actual consulting object.

[0132] In some embodiments, when a comparison instruction for the currently uploaded test report is detected, such as "Please help me compare and analyze it", it will be checked whether there is a historical test report in the health record library of the actual consulting object.

[0133] Step d2: If there is a historical test report in the health record library of the actual consulting object, a second prompt message is output to indicate whether a comparison and interpretation is required.

[0134] If there are historical test reports in the actual consultation object's health record library, the person will be asked whether a comparative interpretation is required.

[0135] In some embodiments, whether comparative interpretation is required may also be determined by identifying a comparative interpretation instruction.

[0136] Step d3: when feedback information requiring comparative interpretation is obtained, a comparative interpretation result is generated for the current test report and the historical test report.

[0137] There are many ways to implement step d3, which can be specifically as follows:

[0138] In some implementations, when multiple test reports or a single test report of an actual consulting object are received, the health history in the health archive of the actual consulting object is identified and interpreted in combination with the health history (such as disease history, medication history, etc.). Specifically, step d3 includes: when feedback information that needs to be compared and interpreted is obtained, a comparative interpretation result is generated for the current test report and the historical test report according to the health archive of the actual consulting object.

[0139] In some implementations, when multiple test reports or a single test report are detected and a question is also detected, the question can be answered in combination with the report. Specifically, step d3 includes: when feedback information that needs to be compared and interpreted is obtained, a comparison and interpretation result is generated based on the question of the test report, the current test report and the historical test report.

[0140] The joint interpretation of multiple test reports can be performed through the received joint interpretation instruction or the selected joint interpretation option.

[0141] In some embodiments, the examination consultation request can be implemented through the examination consultation model. The training process of the examination consultation model is similar to the training process of the disease consultation model, and can be obtained through the following steps:

[0142] First, prompt instructions (prompt word instructions) are generated based on medical knowledge, and the prompt word instructions are input into the Generative Pre-Trained Transformer (GPT) model, Kimi or Qwen-max model to generate report classification, multiple report forms combined with portraits, single report form combined with portraits, multiple report forms combined with user questions, single report form combined with user questions, multiple report form time series comparison, multiple report form joint interpretation, single report form interpretation and other data.

[0143] Next, the generated data is manually annotated or rewritten to obtain data for supervised fine-tuning (SFT). These data contain test reports and corresponding interpretation results, which are used to guide the large language model (hereinafter referred to as the large model) to learn how to generate more accurate and relevant interpretation results.

[0144] Afterwards, the large model is trained through supervised learning to obtain the inspection and consultation model. The trained inspection and consultation model can then be applied to inspection and consultation.

[0145] At the same time, in order to ensure the diversity and personalization of the interaction process and the reply content, the constructed training samples can also be diverse, such as including reply content that combines user portraits, and also including reply content that does not combine user portraits. At the same time, it is necessary to ensure the positive and negative examples of the training samples. This allows the model to not only correctly reply to user questions, but also provide a comprehensive reply for some unconventional questions. In addition, it is also necessary to construct instruction data so that it can reply to various instruction questions of users.

[0146] In addition, the health consultation processing method of this embodiment can also provide functions such as TCM syndrome differentiation, finding a hospital, finding a department, and dietary advice. Among them, TCM syndrome differentiation can perform a preliminary analysis based on the received input symptoms, signs, etc. and give corresponding health advice plans. The hospital search function can recommend a suitable hospital according to user needs. The department search function can recommend a suitable department according to user needs. The dietary advice function can give dietary advice based on the information input by the user, including recommended diets and taboo diets.

[0147] In addition, in some embodiments, health assessment results can also be generated based on medical record information. For example, the diagnostic information and medical advice information of the outpatient medical records are extracted from the medical record information such as the outpatient medical record and the discharge summary, and trend graphs are generated for some of the indicators. For example, for the outpatient medical record with the chief complaint of "left foot joint pain for 3 days", the diagnostic information and medical advice information therein can be automatically extracted as "preliminary diagnosis: gouty arthritis" and "medication plan: drug A, drug B". And a trend graph is generated based on the specific values ​​of the same indicator (such as uric acid, alanine aminotransferase, γ-glutamyl transpeptidase) in the examination results of different time periods for visualization.

[0148] It should be noted that during the training process of the above models, different tasks (such as medical consultation and medication consultation) can be implemented through the same model, or different models can be trained separately. Regardless of the method, it is necessary to ensure the uniformity of the answer format and style. For example, in answering questions, it is usually necessary to answer the user's questions directly first, then analyze and answer them, and finally summarize and cover them. The more uniform the format and style of the training data, the more it can maximize the upper limit of the model's effectiveness in specific tasks. On the basis of the uniform answer format and style, instructions can also be used to make the response paradigm more diverse, and to provide a more practical format and style for different questions.

[0149] In addition, the training dataset can contain both high-difficulty training samples and low-difficulty training samples to ensure that the model can handle training samples of various difficulty levels.

[0150] Also pay attention to the diversity and label balance of different tasks if the same model is used to implement different tasks. If two tasks are of similar difficulty, but the data share of task 1 is much larger than that of task 2, the fine-tuned model may perform poorly on task 2.

[0151] In addition, avoid introducing knowledge that the model has not been exposed to during the pre-training phase to reduce the risk of the model hallucinating.

[0152] Among them, the joint interpretation and comparative interpretation of the report sheets can be performed by jointly interpreting or comparatively interpreting two report sheets uploaded at different times, or by jointly interpreting or comparatively interpreting the report sheets stored in the health records.

[0153] If there are reports stored in the health archive, the same type of reports in the user's health archive can be identified to remind the user to conduct a joint interpretation or a comparative interpretation, and the user can choose whether to conduct a joint or comparative interpretation.

[0154] The above functions can be realized through models. For example, by making the model learn the classification of different reports, which reports (including examination or test reports) belong to the same category. When multiple reports are identified, such as small items of the test such as white blood cells in a blood routine, the user can be reminded to compare and interpret. When the examination identifies a chest CT or a head CT, or the description results of the report are the same part and result, the user can also be reminded to compare and interpret. Through comparative interpretation, users can know the changing trend of their examination or test results.

[0155] The embodiment of the present application generates respective health archives based on all the information (reports, medical records, etc.) of the current user and its bound users, so that each time a health consultation request is received, valid information can be called from the health archives for condition analysis, medication guidance, and examination and test interpretation, etc. The health archive is updated according to the current health consultation result, so that each reply to the user's question will be personalized in combination with the information in the health archives, and the user will not be repeatedly asked for the existing information in the health archives, making the interaction more efficient.

[0156] Specifically, for users, the information stored in the health archive can be stored, and personalized responses can be provided for their subsequent consultations. For example, if the user has a history of allergy to a certain drug in his health archive, the consultation will not be asked again, and a more appropriate response can be given (no more allergic drugs will be recommended to him). In addition, if the user has already done certain examinations, the consultation process will not ask about the examinations he has done, and a suitable diagnosis and suggestions will be given in combination with the examinations. In addition, if the user has a history of certain diseases, examinations and drugs that are contraindicated for these diseases will not be recommended. Therefore, the value of the health space lies in reducing repeated inquiries during the interaction process, providing users with more personalized response suggestions, and improving consultation efficiency.

[0157] Exemplary Devices

[0158] Corresponding to the above-mentioned health consultation processing method, the embodiment of the present application also provides a health consultation processing device. Figure 6 Schematic diagram of a health consultation processing device provided in an embodiment of the present application. Figure 6 As shown, the health consultation processing device provided by the embodiment of the present application includes: a determination unit 61, an acquisition unit 62 and a generation unit 63; the determination unit 61 is used to respond to the health consultation request of the current user and determine the actual consultation object corresponding to the health consultation request; the acquisition unit 62 is used to obtain a pre-stored health record library of the actual consultation object, the health record library includes a personal portrait and a health portrait of the actual consultation object, the personal portrait is obtained based on the personal information of the actual consultation object, and the health portrait is obtained based on the health information of the actual consultation object at each age group; the generation unit 63 is used to generate a reply content for the health consultation request based on the health record library of the actual consultation object.

[0159] In some embodiments, the determination unit 61 determines the actual consultation object corresponding to the health consultation request in response to the health consultation request of the current user, specifically including: determining the actual consultation object corresponding to the health consultation request by identifying the consultation object role in the health consultation request, wherein the consultation object role is the current user or the bound user corresponding to the current user; when it is identified that the actual consultation object is the bound user of the current user, outputting a first prompt information to switch the current user to the bound user.

[0160] In some embodiments, the determination unit 61 determines the actual consultation object corresponding to the health consultation request by identifying the consultation object role in the health consultation request, specifically including: obtaining multiple modal data about the consultation object in the health consultation request; extracting multiple modal features from the multiple modal data respectively through a multimodal model, and fusing the multiple modal features to obtain fused features; inputting the fused features as independent variables into a logic and regression model, and determining the consultation object role corresponding to the independent variable as the actual consultation object through the logic and regression model according to the pre-learned mapping relationship between the independent variable and the dependent variable, wherein the dependent variable includes multiple preset consultation object roles.

[0161] In some embodiments, the generation unit 63 generates reply content to the health consultation request based on the health record library of the actual consultation object, including: when it is determined that further questions need to be asked in response to the health consultation request, determining whether there are answers corresponding to the questions to be asked in the health record library of the actual consultation object; when it is determined that there are answers corresponding to the questions to be asked in the health record library of the actual consultation object, generating reply content to the health consultation request based on the answers corresponding to the questions to be asked stored in the health record library of the actual consultation object.

[0162] In some embodiments, the health consultation request includes a medical condition consultation request, and the generation unit 63 is also used to perform the following steps: generate medical record information for this consultation based on the medical condition consultation request and the corresponding reply content; and add the medical record information for this consultation to the health record library of the actual consultation object.

[0163] In some embodiments, the health consultation request includes a medication consultation request, and the health record library includes at least one of a current medication plan and a health history; wherein, the generation unit 63 generates a reply content for the health consultation request based on the health record library of the actual consultation object, including: generating medication suggestion reply content for the current consultation medication in the medication consultation request based on the current medication plan and at least one of the health history of the actual consultation object, and the medication suggestion reply content includes suggestions on whether the current consultation medication needs to avoid medication risks.

[0164] In some embodiments, the generation unit 63 is also used to perform the following steps: when it is detected that the current test report of the actual consulting object is uploaded, determining whether there is a historical test report in the health record library of the actual consulting object; if there is a historical test report in the health record library of the actual consulting object, outputting a second prompt information on whether a comparative interpretation is required; when feedback information that a comparative interpretation is required is obtained, generating a comparative interpretation result for the current test report and the historical test report.

[0165] In some embodiments, when the generation unit 63 obtains feedback information that requires comparative interpretation, it generates comparative interpretation results for the current test report and the historical test report, specifically including: when the feedback information that requires comparative interpretation is obtained, based on the health record library of the actual consulting object, a comparative interpretation result is generated for the current test report and the historical test report.

[0166] The health consultation processing device provided in this embodiment belongs to the same application concept as the health consultation processing method provided in the above embodiments of this application, and can execute the health consultation processing method provided in any of the above embodiments of this application, and has the corresponding functional modules and beneficial effects of executing the health consultation processing method. For technical details not fully described in this embodiment, please refer to the specific processing content of the health consultation processing method provided in the above embodiments of this application, which will not be repeated here.

[0167] The functions implemented by the above determination unit 61, acquisition unit 62 and generation unit 63 may be implemented by the same or different processors respectively, which is not limited in the embodiment of the present application.

[0168] It should be understood that the units in the above devices can be implemented in the form of a processor calling software. For example, the device includes a processor, the processor is connected to a memory, and instructions are stored in the memory. The processor calls the instructions stored in the memory to implement any of the above methods or realize the functions of each unit of the device, wherein the processor can be a general-purpose processor, such as a CPU or a microprocessor, etc., and the memory can be a memory in the device or a memory outside the device. Alternatively, the units in the device can be implemented in the form of hardware circuits, and the functions of some or all units can be realized by designing the hardware circuits. The hardware circuit can be understood as one or more processors; for example, in one implementation, the hardware circuit is an ASIC, and the functions of some or all of the above units are realized by designing the logical relationship of the components in the circuit; for another example, in another implementation, the hardware circuit can be implemented by PLD, taking FPGA as an example, which can include a large number of logic gate circuits, and the connection relationship between the logic gate circuits is configured by the configuration file, so as to realize the functions of some or all of the above units. All units of the above devices can be implemented in the form of a processor calling software, or in the form of hardware circuits, or in part by a processor calling software, and the remaining part is implemented in the form of hardware circuits.

[0169] In an embodiment of the present application, a processor is a circuit with the ability to process signals. In one implementation, the processor may be a circuit with the ability to read and run instructions, such as a CPU, a microprocessor, a GPU, or a DSP; in another implementation, the processor may implement certain functions through the logical relationship of a hardware circuit, and the logical relationship of the hardware circuit is fixed or reconfigurable, such as a hardware circuit implemented by an ASIC or PLD, such as an FPGA. In a reconfigurable hardware circuit, the process of the processor loading a configuration document to implement the hardware circuit configuration can be understood as the process of the processor loading instructions to implement the functions of some or all of the above units. In addition, it can also be a hardware circuit designed for artificial intelligence, which can be understood as an ASIC, such as an NPU, TPU, DPU, etc.

[0170] It can be seen that each unit in the above device can be one or more processors (or processing circuits) configured to implement the above method, such as: CPU, GPU, NPU, TPU, DPU, microprocessor, DSP, ASIC, FPGA, or a combination of at least two of these processor forms.

[0171] In addition, all or part of the units in the above device can be integrated together, or can be implemented independently. In one implementation, these units are integrated together and implemented in the form of a SOC. The SOC may include at least one processor for implementing any of the above methods or implementing the functions of each unit of the device. The type of the at least one processor may be different, for example, including a CPU and an FPGA, a CPU and an artificial intelligence processor, a CPU and a GPU, etc.

[0172] Exemplary Electronic Devices

[0173] The present application embodiment provides an electronic device, see Figure 7 As shown, the electronic device includes:

[0174] Memory 200 and processor 210;

[0175] The memory 200 is connected to the processor 210 and is used to store programs;

[0176] The processor 210 is used to implement the health consultation processing method disclosed in any of the above embodiments by running the program stored in the memory 200.

[0177] Specifically, the electronic device may further include: a bus, a communication interface 220 , an input device 230 and an output device 240 .

[0178] The processor 210, the memory 200, the communication interface 220, the input device 230 and the output device 240 are connected to each other via a bus.

[0179] A bus may include a pathway that transfers information between components of a computer system.

[0180] The processor 210 may be a general-purpose processor, such as a general-purpose central processing unit (CPU), a microprocessor, etc., or an application-specific integrated circuit (ASIC), or one or more integrated circuits for controlling the execution of the program of the scheme of the present invention. It may also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0181] The processor 210 may include a main processor, and may also include a baseband chip, a modem, and the like.

[0182] The memory 200 stores a program for executing the technical solution of the present invention, and may also store an operating system and other key services. Specifically, the program may include a program code, and the program code includes a computer operation instruction. More specifically, the memory 200 may include a read-only memory (ROM), other types of static storage devices that can store static information and instructions, a random access memory (RAM), other types of dynamic storage devices that can store information and instructions, a disk storage, a flash, and the like.

[0183] The input device 230 may include a device for receiving data and information input by a user, such as a keyboard, a mouse, a camera, a scanner, a light pen, a voice input device, a touch screen, a pedometer, or a gravity sensor.

[0184] Output device 240 may include devices that allow information to be output to a user, such as a display screen, printer, speaker, etc.

[0185] The communication interface 220 may include any transceiver or the like to communicate with other devices or communication networks, such as Ethernet, Radio Access Network (RAN), Wireless Local Area Network (WLAN), etc.

[0186] The processor 210 executes the program stored in the memory 200 and calls other devices, which can be used to implement each step of any health consultation processing method provided in the above embodiments of the present application.

[0187] An embodiment of the present application also proposes a chip, which includes a processor and a data interface. The processor reads and runs a program stored in a memory through the data interface to execute the health consultation processing method introduced in any of the above embodiments. The specific processing process and its beneficial effects can be found in the embodiment introduction of the above-mentioned health consultation processing method.

[0188] Exemplary computer program products and storage media

[0189] In addition to the above-mentioned methods and devices, an embodiment of the present application may also be a computer program product, which includes computer program instructions, which, when executed by a processor, enable the processor to execute the steps of the health consultation processing method according to various embodiments of the present application described in any of the above embodiments of this specification.

[0190] The computer program product may be written in any combination of one or more programming languages ​​to write program codes for performing the operations of the embodiments of the present application, including object-oriented programming languages, such as Java, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as an independent software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0191] In addition, an embodiment of the present application may also be a storage medium on which a computer program is stored. The computer program is executed by a processor to execute the steps of the health consultation processing method according to various embodiments of the present application described in any of the above embodiments of this specification, and specifically the steps of the above health consultation processing method can be implemented.

[0192] For the aforementioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the order of the actions described, because according to the present application, some steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0193] It should be noted that each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same or similar parts between the embodiments can be referred to each other. For the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0194] The steps in the methods of each embodiment of the present application can be adjusted in order, combined and deleted according to actual needs, and the technical features recorded in each embodiment can be replaced or combined.

[0195] The modules and sub-modules in the devices and terminals of the various embodiments of the present application can be combined, divided and deleted according to actual needs.

[0196] In the several embodiments provided in the present application, it should be understood that the disclosed terminals, devices and methods can be implemented in other ways. For example, the terminal embodiments described above are only schematic, for example, the division of modules or submodules is only a logical function division, and there may be other division methods in actual implementation, for example, multiple submodules or modules can be combined or integrated into another module, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or modules, which can be electrical, mechanical or other forms.

[0197] The modules or submodules described as separate components may or may not be physically separated, and the components of the modules or submodules may or may not be physical modules or submodules, that is, they may be located in one place, or they may be distributed on multiple network modules or submodules. Some or all of the modules or submodules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0198] In addition, each functional module or submodule in each embodiment of the present application may be integrated into one processing module, or each module or submodule may exist physically separately, or two or more modules or submodules may be integrated into one module. The above-mentioned integrated modules or submodules may be implemented in the form of hardware or in the form of software functional modules or submodules.

[0199] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0200] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly by hardware, software units executed by a processor, or a combination of the two. The software units may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0201] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0202] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be 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 the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A health consultation processing method, characterized in that: include: In response to a health consultation request of a current user, determining an actual consultation object corresponding to the health consultation request; Obtaining a pre-stored health archive of the actual consulting object, wherein the health archive includes a personal portrait and a health portrait of the actual consulting object, wherein the personal portrait is obtained by profiling the actual consulting object based on the personal information of the actual consulting object, and the health portrait is obtained by profiling the actual consulting object based on health information of the actual consulting object at various age groups; Generate a reply to the health consultation request based on the health record library of the actual consultation object.

2. The method according to claim 1, characterized in that The step of responding to the health consultation request of the current user and determining the actual consultation object corresponding to the health consultation request includes: Determine the actual consulting object corresponding to the health consultation request by identifying the consulting object role in the health consultation request, where the consulting object role is the current user or a bound user corresponding to the current user; In the case where it is identified that the actual consultation object is a bound user of the current user, first prompt information for switching the current user to the bound user is output.

3. The method according to claim 2, characterized in that The step of determining the actual consulting object corresponding to the health consulting request by identifying the consulting object role in the health consulting request includes: Acquire data of multiple modalities about the consultation object in the health consultation request; Extracting features of multiple modalities from the data of the multiple modalities respectively through a multimodal model, and fusing the features of the multiple modalities to obtain fused features; The fusion feature is input as an independent variable into the logic and regression model, and the logic and regression model determines the consulting object role corresponding to the independent variable as the actual consulting object according to the pre-learned mapping relationship between the independent variable and the dependent variable, and the dependent variable includes multiple preset consulting object roles.

4. The method according to claim 1, characterized in that: Generating a reply to the health consultation request according to the health record library of the actual consultation object includes: In the case where it is determined that further questions are required for the health consultation request, determining whether there are answers corresponding to the further questions in the health record library of the actual consultation object; When it is determined that the answers to the continued questions exist in the health record library of the actual consultation object, a reply content is generated for the health consultation request based on the answers to the continued questions stored in the health record library of the actual consultation object.

5. The method according to any one of claims 1 to 4, characterized in that The health consultation request includes a condition consultation request. After generating a reply content for the health consultation request according to the health archive of the actual consultation object, the method further includes: Generate medical record information of this consultation according to the medical consultation request and the corresponding reply content; The medical record information of the consultation is added to the health record library of the actual consultation object.

6. The method according to any one of claims 1 to 4, characterized in that The health consultation request includes a medication consultation request, and the health archive includes at least one of a current medication plan and a health history; Wherein, generating a reply content to the health consultation request according to the health record library of the actual consultation object includes: Based on at least one of the current medication plan and health history of the actual consulting subject, a medication suggestion reply content is generated for the current consulting medication in the medication consultation request, and the medication suggestion reply content includes suggestions on whether the current consulting medication needs to avoid medication risks.

7. The method according to any one of claims 1 to 4, characterized in that The method further comprises: In the case of detecting that a current test report of the actual consulting object has been uploaded, determining whether there is a historical test report in the health record library of the actual consulting object; If there is a historical test report in the health archive of the actual consulting object, output a second prompt message indicating whether a comparison and interpretation is required; When feedback information requiring comparative interpretation is obtained, a comparative interpretation result is generated for the current test report and the historical test report.

8. The method according to claim 7, characterized in that When feedback information requiring comparative interpretation is obtained, a comparative interpretation result is generated for the current test report and the historical test report, including: When feedback information requiring comparative interpretation is obtained, a comparative interpretation result is generated for the current test report and the historical test report according to the health record library of the actual consulting object.

9. A health consultation processing device, characterized in that: include: a determination unit, configured to respond to a health consultation request of a current user and determine an actual consultation object corresponding to the health consultation request; an acquisition unit, used for acquiring a pre-stored health archive of the actual consulting object, wherein the health archive includes a personal portrait and a health portrait of the actual consulting object, wherein the personal portrait is obtained by profiling the actual consulting object based on the personal information of the actual consulting object, and the health portrait is obtained by profiling the actual consulting object based on the health information of the actual consulting object at various age groups; A generating unit is used to generate reply content to the health consultation request according to the health record library of the actual consultation object.

10. An electronic device, characterized in that: including memory and processor; The memory is connected to the processor and is used to store programs; The processor is configured to implement the method according to any one of claims 1 to 8 by running the program in the memory.

11. A storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by the processor, the method according to any one of claims 1 to 8 is implemented.

12. A computer program product, characterized in that The method comprises computer program instructions, which, when executed by a processor, cause the processor to implement the method as claimed in any one of claims 1 to 8.

Images