An inquiry system and device based on artificial intelligence technology

Abstract

The application provides an inquiry system and equipment based on artificial intelligence technology, comprising: a data analysis module for receiving physiological information data of a patient collected by a smart wearable device worn on the patient; and performing big data analysis on the physiological information data to preliminarily determine the physical data of the patient; a strategy determination module for determining an inquiry question strategy and an inquiry answer strategy for the patient based on the physical data in combination with a medical knowledge graph and machine learning; an intelligent question and answer module for recognizing the inquiry and answer voice of the patient, combining the inquiry question strategy and the inquiry answer strategy, realizing a voice inquiry process with the patient, and obtaining inquiry information; and a result analysis module for determining the condition of the patient and providing a personalized treatment plan by using machine learning based on the inquiry information, so as to solve the problems of congestion and shortage of medical resources in the medical service system.

Description

Technical Field

[0001] This invention relates to the field of medical technology, and in particular to a consultation system and device based on artificial intelligence technology. Background Technology

[0002] With population growth and an aging population, the demand for medical services is increasing. Furthermore, modern people have increasingly diverse needs for medical services, requiring more detailed and personalized care beyond basic diagnosis and treatment. Traditional medical services require patients to go to the hospital for diagnosis and treatment, which involves long queues during peak periods, causing inconvenience. Moreover, due to varying levels of doctor expertise, consultations may be incomplete or inaccurate, failing to accurately grasp the patient's condition and ultimately failing to provide satisfactory medical services. At the same time, the significant demand for medical services has led to congestion and strain on the healthcare system. Summary of the Invention

[0003] This invention provides a consultation system and device based on artificial intelligence technology to solve the problems of congestion and shortage of medical resources in the medical service system.

[0004] A diagnostic system based on artificial intelligence technology includes:

[0005] The data analysis module is used to receive physiological information data collected by the patient from the smart wearable device worn on the patient's body; and to perform big data analysis on the physiological information data to preliminarily determine the patient's physical data;

[0006] The strategy determination module is used to determine the questioning and response strategies for patients based on body data, combined with medical knowledge graphs and machine learning.

[0007] The intelligent question-and-answer module is used to recognize the patient's voice questions and answers, and combine the question-and-answer strategies to realize the voice consultation process with the patient and obtain consultation information.

[0008] The results analysis module is used to determine the patient's condition and provide personalized treatment plans based on the consultation information and machine learning.

[0009] Preferably, the data analysis module includes:

[0010] The receiving unit is used to receive physiological information data of the patient collected by a smart wearable device worn on the patient's body;

[0011] The integration unit is used to integrate physiological information data of the same physiological attribute according to the collection time, and to establish a physiological data change curve under that attribute based on the integration result.

[0012] The recording unit is used to determine the recording order of physiological data change curves based on the correlation between physiological attributes, and to integrate the physiological data change curves based on the recording order to obtain target physiological information data.

[0013] The analysis unit is used to perform big data analysis on physiological information data based on the correlation between physiological attributes, determine the evaluation parameters of the target physiological information data under the body evaluation indicators, and preliminarily determine the patient's body data based on the evaluation parameters.

[0014] Preferably, the strategy determination module includes:

[0015] The data acquisition unit is used to acquire first body data in a normal state and second body data in an abnormal state from body data, and to acquire first disease information related to the first body data and second disease information related to the second body data based on the medical knowledge graph.

[0016] The association unit is used to obtain the first body keywords of the first body data, obtain the first disease keywords of the first disease information, associate the first body keywords and the first disease keywords to obtain the first association result, obtain the second body keywords of the second body data, obtain the second disease keywords of the second disease information, associate the second body keywords and the second disease keywords to obtain the second association result.

[0017] The expansion unit is used to expand the first association result based on the keyword-related words to obtain the first comprehensive association result, and to expand the second association result to obtain the second comprehensive association result;

[0018] The script generation unit is used to collect doctors' historical consultation records, analyze and process the historical consultation records to obtain consultation script templates, input the consultation script templates and the first comprehensive correlation results into a pre-designed disease-free script generation model to obtain the first script set, and input the consultation script templates and the second comprehensive correlation results into a pre-designed script generation model to obtain the second script set.

[0019] The strategy determination unit is used to extract the questioning script set from the first script set and the second script set as the questioning strategy, and to extract the response script set from the first script set and the second script set as the response strategy.

[0020] Preferably, the strategy determination module further includes:

[0021] The model building unit is used to train the initial machine learning model based on the consultation script template to generate a script generation model, and to further train the script generation model based on negative statement keywords to obtain a disease-free script generation model, and to further train the script generation model based on positive statement keywords to obtain a disease-free script generation model.

[0022] Preferred intelligent question-answering module includes:

[0023] The speech recognition unit is used to receive the patient's question and answer speech, and to recognize the question and answer speech to obtain the speech recognition result;

[0024] The intent determination unit is used to perform language understanding on the speech recognition results, determine the user's intent, and determine the speech response template based on the user's intent;

[0025] The response confirmation unit is used to fill in the voice response template based on the consultation questioning strategy and consultation response strategy to achieve intelligent voice response;

[0026] The information determination unit is used to obtain consultation information based on the patient's response.

[0027] Preferably, the intent determination unit includes:

[0028] The semantic analysis unit is used to classify the speech recognition results into the speech recognition model, obtain semantic features, extract semantic keywords from the speech features, and predict the user's first intent based on the semantic keywords.

[0029] The above analysis unit is used to determine whether the speech recognition result contains the above content in the speech question and answer process;

[0030] If so, determine the preceding intent of the preceding content, perform a fluency analysis between the preceding intent and the first intent to determine the fluency value, perform a tightness analysis between the preceding intent and the first intent to determine the tightness value, and adjust the first intent based on the preceding intent, combining the fluency value and the tightness value to obtain the second intent;

[0031] Otherwise, the user's first intent will be taken as the second intent;

[0032] The human body analysis unit is used to determine the user's human medical map based on the user's body data, mark abnormal data in the human medical map as key points, decompose the second intent according to medical characteristics, determine the intent distribution characteristics, mark the intent distribution characteristics on the human medical map, and determine the matching degree between the intent distribution characteristics and key points based on the marking results.

[0033] The intent acquisition unit is used to determine whether the matching degree is greater than the preset matching degree;

[0034] If so, the second intent should be taken as the user's final intent;

[0035] Otherwise, based on the deviation between the intent distribution characteristics and the key points, the second intent is corrected to obtain the user's final intent.

[0036] Preferably, the response determination unit includes:

[0037] The extraction unit is used to determine response keywords based on the user's intent and extract consultation strategies related to the response keywords from the consultation questioning strategy and consultation response strategy;

[0038] The filling unit is used to determine the filling attributes in the voice response template, extract the corresponding consultation keywords from the consultation strategy based on the filling attributes, and fill the consultation keywords into the voice response template to obtain the initial voice response content;

[0039] The adjustment unit inputs the initial voice response content into the text fluency adjustment model for fluency adjustment to obtain the target voice response content;

[0040] The voice determination unit is used to match the corresponding response tone based on the tone of the user's question and answer speech, and generate voice text from the target voice response content using the response tone to achieve intelligent voice response.

[0041] Preferably, the results analysis module includes:

[0042] The disease diagnosis unit is used to train a machine learning model based on the patient's historical medical history and disease information to obtain a disease diagnosis model. The medical history information is input into the disease diagnosis model to determine the patient's disease.

[0043] The treatment plan acquisition unit is used to acquire historical treatment plans of patients with the same condition as the patient and to determine the recovery status of the historical patients under the historical treatment plans.

[0044] The treatment plan adjustment unit is used to make personalized adjustments to historical treatment plans based on the historical patients and their corresponding disease recovery status, so as to obtain personalized treatment plans for the patients.

[0045] Preferably, the scheme adjustment unit includes:

[0046] The difference comparison unit is used to compare the patient's physical data with the patient's physical data based on the patient's historical physical data to determine physical differences, and to determine the recovery characteristics of abnormal recovery based on the patient's historical recovery status.

[0047] The adjustment and replacement unit is used to extract the program features corresponding to the physical differences and recovery characteristics from historical treatment programs, extract the target program features that correspond to the patient's physical data from other historical treatment programs, and replace the program features in the historical treatment programs with the target program features to obtain the patient's personalized treatment program.

[0048] A diagnostic device for implementing an artificial intelligence-based diagnostic system includes:

[0049] Smart wearable devices are worn by patients to collect their physiological information data;

[0050] Intelligent question-and-answer devices are used to store and execute one or more program instructions.

[0051] Compared with the prior art, the present invention has achieved the following beneficial effects:

[0052] By receiving physiological information data collected from patients through smart wearable devices, and performing big data analysis on this data, the system initially determines the patient's physical condition. This data analysis provides accurate information for designing intelligent consultation systems. Based on this data, and combined with medical knowledge graphs and machine learning, the system determines questioning and response strategies, providing a solid data foundation for intelligent Q&A. Then, it recognizes the patient's spoken responses and, combined with the questioning and response strategies, implements a voice-based consultation process, ensuring smooth and targeted intelligent consultation. Based on this information, machine learning is used to determine the patient's condition and provide personalized treatment plans. Ultimately, the intelligent consultation system can provide fast and efficient services to patients via the internet or mobile devices, reducing waiting time and medical costs. Through natural language processing and machine learning, the system accurately identifies and analyzes patient conditions, providing more precise diagnoses and treatment recommendations. Furthermore, the system provides doctors with a wealth of medical knowledge and treatment options, helping them better understand patients' conditions and develop more appropriate treatment plans. Intelligent consultation systems can help hospitals optimize the allocation of medical resources, improve the utilization rate of medical resources, and alleviate hospital congestion.

[0053] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings.

[0054] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0055] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0056] Figure 1 This is a structural diagram of a medical consultation system based on artificial intelligence technology in an embodiment of the present invention;

[0057] Figure 2 This is a structural diagram of the data analysis module in an embodiment of the present invention;

[0058] Figure 3 This is a structural diagram of a medical consultation device based on artificial intelligence technology in an embodiment of the present invention. Detailed Implementation

[0059] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0060] Example 1:

[0061] This invention provides a diagnostic system based on artificial intelligence technology, such as... Figure 1 As shown, it includes:

[0062] The data analysis module is used to receive physiological information data collected by the patient from the smart wearable device worn on the patient's body; and to perform big data analysis on the physiological information data to preliminarily determine the patient's physical data;

[0063] The strategy determination module is used to determine the questioning and response strategies for patients based on body data, combined with medical knowledge graphs and machine learning.

[0064] The intelligent question-and-answer module is used to recognize the patient's voice questions and answers, and combine the question-and-answer strategies to realize the voice consultation process with the patient and obtain consultation information.

[0065] The results analysis module is used to determine the patient's condition and provide personalized treatment plans based on the consultation information and machine learning.

[0066] The working principle of the above design scheme is as follows: It receives physiological information data from patients, including pulse, respiration, heart rate, and other body data, collected by a smart wearable device worn on the patient's body. This data is then analyzed using big data to preliminarily determine the patient's physical condition and identify changes in the patient's body over a period of time. Based on this physical data, and combining medical knowledge graphs and machine learning (the medical knowledge graph is constructed based on medical knowledge, and the machine learning is artificial intelligence learning based on big data), it determines the questioning and response strategies for the patient. The questioning strategy includes key points for asking the patient, and the response information includes key points for answering the patient, providing an accurate data foundation for intelligent question-and-answer communication. Then, it identifies the patient's voice and, combined with the questioning and response strategies, implements a voice-based consultation process to obtain consultation information. Based on this information, machine learning is used to determine the patient's condition and provide a personalized treatment plan.

[0067] The beneficial effects of the above design scheme are as follows: By receiving physiological information data collected from patients through smart wearable devices, big data analysis is performed on this data to preliminarily determine the patient's physical data. The analysis and processing of the received data provides accurate physical information for the design of intelligent consultation. Based on this physical data, combined with medical knowledge graphs and machine learning, questioning and response strategies are determined, providing an accurate data foundation for intelligent Q&A with patients. Then, the system recognizes the patient's voice and, combined with the questioning and response strategies, implements a voice-based consultation process, obtaining consultation information and ensuring the smoothness and relevance of the intelligent consultation. Based on this consultation information, machine learning is used to determine the patient's condition and provide personalized treatment plans, offering intelligent treatment options. Ultimately, the intelligent consultation system can provide patients with fast and efficient services through the internet or mobile devices, reducing patient waiting time and medical costs. Through natural language processing and machine learning technologies, the intelligent consultation system can accurately identify and analyze patients' conditions, thereby providing more precise diagnostic and treatment suggestions. Intelligent consultation systems can also provide doctors with a wealth of medical knowledge and treatment plans, helping them better understand patients' conditions and develop more appropriate treatment plans. Intelligent consultation systems can also help hospitals optimize the allocation of medical resources, improve resource utilization, and alleviate hospital congestion.

[0068] Example 2:

[0069] Based on Embodiment 1, this embodiment of the invention provides a diagnostic system based on artificial intelligence technology, such as... Figure 2 As shown, the data analysis module includes:

[0070] The receiving unit is used to receive physiological information data of the patient collected by a smart wearable device worn on the patient's body;

[0071] The integration unit is used to integrate physiological information data of the same physiological attribute according to the collection time, and to establish a physiological data change curve under that attribute based on the integration result.

[0072] The recording unit is used to determine the recording order of physiological data change curves based on the correlation between physiological attributes, and to integrate the physiological data change curves based on the recording order to obtain target physiological information data.

[0073] The analysis unit is used to perform big data analysis on physiological information data based on the correlation between physiological attributes, determine the evaluation parameters of the target physiological information data under the body evaluation indicators, and preliminarily determine the patient's body data based on the evaluation parameters.

[0074] In this embodiment, the correlation between physiological attributes includes, for example, the correlation between heart rate and pulse, and the correlation between blood pressure and blood sugar.

[0075] In this embodiment, the related data in the target physiological information data are ordered adjacently, which facilitates the analysis and viewing of the data.

[0076] In this embodiment, the patient's physical data approximation includes physiological test data as well as evaluation data of the physiological test data.

[0077] The beneficial effects of the above design scheme are: by integrating physiological information data according to the correlation of physiological attributes, it facilitates the viewing and analysis of the data. Furthermore, by adding evaluation data to the physiological information data based on body evaluation indicators, the cleanliness and high value of the obtained body data are ensured, providing an effective data foundation for intelligent consultation.

[0078] Example 3:

[0079] Based on Embodiment 1, this embodiment of the invention provides a diagnostic system based on artificial intelligence technology, including a strategy determination module, comprising:

[0080] The data acquisition unit is used to acquire first body data in a normal state and second body data in an abnormal state from body data, and to acquire first disease information related to the first body data and second disease information related to the second body data based on the medical knowledge graph.

[0081] The association unit is used to obtain the first body keywords of the first body data, obtain the first disease keywords of the first disease information, associate the first body keywords and the first disease keywords to obtain the first association result, obtain the second body keywords of the second body data, obtain the second disease keywords of the second disease information, associate the second body keywords and the second disease keywords to obtain the second association result.

[0082] The expansion unit is used to expand the first association result based on the keyword-related words to obtain the first comprehensive association result, and to expand the second association result to obtain the second comprehensive association result;

[0083] The script generation unit is used to collect doctors' historical consultation records, analyze and process the historical consultation records to obtain consultation script templates, input the consultation script templates and the first comprehensive correlation results into a pre-designed disease-free script generation model to obtain the first script set, and input the consultation script templates and the second comprehensive correlation results into a pre-designed script generation model to obtain the second script set.

[0084] The strategy determination unit is used to extract the questioning script set from the first script set and the second script set as the questioning strategy, and to extract the response script set from the first script set and the second script set as the response strategy.

[0085] In this embodiment, expanding the first association result and the second association result specifically involves obtaining the same keyword from the keywords in the first association result and the second association result, enriching the association result, and making the keywords in the obtained comprehensive association result more abundant.

[0086] In this embodiment, the consultation script template is a collection of commonly used consultation scripts, such as --- the pain is caused by ----, and the specific situation needs to be determined by --- examination. The specific key information to be filled in is determined based on the first comprehensive association result and the second comprehensive association result.

[0087] The beneficial effects of the above design scheme are as follows: After classifying physical data according to abnormal conditions, related diseases are analyzed. The association between physical data and diseases is determined based on the keywords in the analysis, and the keywords are expanded to facilitate the richness and accuracy of the association results. Then, a consultation script model is obtained based on the doctor's consultation information. This model is combined with the key results obtained from the analysis to analyze both diseased and disease-free cases, resulting in corresponding script sets. Finally, a consultation response strategy is obtained. This strategy first ensures the accuracy of the information provided to the patient by considering both diseased and disease-free cases, and secondly, it ensures the fluency of communication with the patient through the consultation script model, providing a foundation for the content of questions and answers in intelligent consultation.

[0088] Example 4:

[0089] Based on Embodiment 3, this embodiment of the invention provides a diagnostic system based on artificial intelligence technology, including a strategy determination module, which further includes:

[0090] The model building unit is used to train the initial machine learning model based on the consultation script template to generate a script generation model, and to further train the script generation model based on negative statement keywords to obtain a disease-free script generation model, and to further train the script generation model based on positive statement keywords to obtain a disease-free script generation model.

[0091] The beneficial effects of the above design scheme are: by training the disease-free dialogue generation model and the disease-related dialogue generation model separately, the normal and abnormal physical data of patients can be analyzed separately to obtain the corresponding consultation dialogue, ensuring the accuracy of the obtained consultation dialogue.

[0092] Example 5:

[0093] Based on Embodiment 1, this embodiment of the invention provides a medical consultation system based on artificial intelligence technology, including an intelligent question-answering module, comprising:

[0094] The speech recognition unit is used to receive the patient's question and answer speech, and to recognize the question and answer speech to obtain the speech recognition result;

[0095] The intent determination unit is used to perform language understanding on the speech recognition results, determine the user's intent, and determine the speech response template based on the user's intent;

[0096] The response confirmation unit is used to fill in the voice response template based on the consultation questioning strategy and consultation response strategy to achieve intelligent voice response;

[0097] The information determination unit is used to obtain consultation information based on the patient's response.

[0098] In this embodiment, the user's intent is, for example, to determine the cause of the disease or to determine a treatment plan.

[0099] The beneficial effects of the above design scheme are as follows: by recognizing and determining the patient's voice and further determining the patient's intention, and determining the voice response template based on the user's intention, the voice response template is filled in based on the questioning and response strategies, realizing intelligent voice response, verifying the correctness of the voice response, and ultimately obtaining consultation information based on the patient's response content, providing a data foundation for providing personalized treatment plans for patients, and finally realizing intelligent consultation with patients, which is convenient for patients and reduces the pressure on hospitals.

[0100] Example 6:

[0101] Based on Embodiment 5, this embodiment of the invention provides a diagnostic system based on artificial intelligence technology, including an intent determination unit, comprising:

[0102] The semantic analysis unit is used to classify the speech recognition results into the speech recognition model, obtain semantic features, extract semantic keywords from the speech features, and predict the user's first intent based on the semantic keywords.

[0103] The above analysis unit is used to determine whether the speech recognition result contains the above content in the speech question and answer process;

[0104] If so, determine the preceding intent of the preceding content, perform a fluency analysis between the preceding intent and the first intent to determine the fluency value, perform a tightness analysis between the preceding intent and the first intent to determine the tightness value, and adjust the first intent based on the preceding intent, combining the fluency value and the tightness value to obtain the second intent;

[0105] Otherwise, the user's first intent will be taken as the second intent;

[0106] The human body analysis unit is used to determine the user's human medical map based on the user's body data, mark abnormal data in the human medical map as key points, decompose the second intent according to medical characteristics, determine the intent distribution characteristics, mark the intent distribution characteristics on the human medical map, and determine the matching degree between the intent distribution characteristics and key points based on the marking results.

[0107] The intent acquisition unit is used to determine whether the matching degree is greater than the preset matching degree;

[0108] If so, the second intent should be taken as the user's final intent;

[0109] Otherwise, based on the deviation between the intent distribution characteristics and the key points, the second intent is corrected to obtain the user's final intent.

[0110] In this embodiment, the determined second intent must satisfy the preconditions regarding the fluency and tightness values ​​of the aforementioned intent.

[0111] In this embodiment, tightness is used to determine the relevance of the context content, ensuring the relevance of the consultation response to the preceding text.

[0112] In this embodiment, fluency is used to determine the smoothness of the context content, ensuring that the consultation response flows smoothly with the preceding text.

[0113] In this embodiment, the human medical atlas is used to graphically display body data for easy and intuitive viewing.

[0114] In this embodiment, the second intention is corrected based on the deviation between the intention distribution features and the key points. Specifically, the intention distribution features with deviations are moved closer to the key points to correct the intention distribution features.

[0115] The beneficial effects of the above design scheme are: by analyzing and correcting the patient's intent step by step from the human medical atlas determined by semantics, contextual content, fluency and tightness, and user body data, the accuracy of the user's final intent is guaranteed, providing a foundation for achieving accurate intelligent consultation responses.

[0116] Example 7:

[0117] Based on Embodiment 5, this embodiment of the invention provides a consultation system based on artificial intelligence technology, including a response determination unit, comprising:

[0118] The extraction unit is used to determine response keywords based on the user's intent and extract consultation strategies related to the response keywords from the consultation questioning strategy and consultation response strategy;

[0119] The filling unit is used to determine the filling attributes in the voice response template, extract the corresponding consultation keywords from the consultation strategy based on the filling attributes, and fill the consultation keywords into the voice response template to obtain the initial voice response content;

[0120] The adjustment unit inputs the initial voice response content into the text fluency adjustment model for fluency adjustment to obtain the target voice response content;

[0121] The voice determination unit is used to match the corresponding response tone based on the tone of the user's question and answer speech, and generate voice text from the target voice response content using the response tone to achieve intelligent voice response.

[0122] The beneficial effects of the above design scheme are as follows: By determining the response keywords based on the user's intent, extracting consultation strategies related to the response keywords from the consultation questioning strategy and consultation response strategy, the content of the consultation response is determined. Combined with the determined filling attributes in the voice response template, the corresponding consultation keywords are extracted from the consultation strategy based on the filling attributes, and the consultation keywords are filled into the voice response template to obtain the initial voice response content. This ensures the correctness of the sentence form of the obtained voice response content, and finally realizes intelligent consultation voice response, realizing intelligent consultation with patients, which is convenient for patients and reduces the pressure on hospitals.

[0123] Example 8:

[0124] Based on Embodiment 1, this embodiment of the invention provides a diagnostic system based on artificial intelligence technology, including a result analysis module:

[0125] The disease diagnosis unit is used to train a machine learning model based on the patient's historical medical history and disease information to obtain a disease diagnosis model. The medical history information is input into the disease diagnosis model to determine the patient's disease.

[0126] The treatment plan acquisition unit is used to acquire historical treatment plans of patients with the same condition as the patient and to determine the recovery status of the historical patients under the historical treatment plans.

[0127] The treatment plan adjustment unit is used to make personalized adjustments to historical treatment plans based on the historical patients and their corresponding disease recovery status, so as to obtain personalized treatment plans for the patients.

[0128] The beneficial effects of the above design scheme are as follows: By training a disease diagnosis model, the patient's condition is determined by inputting the consultation information into the model, ensuring the efficiency and accuracy of the diagnosis. Then, historical treatment plans for patients with the same condition are obtained, and the recovery status of these patients under the historical treatment plans is determined. Based on the historical patients and their corresponding recovery status, the historical treatment plans are adjusted to obtain personalized treatment plans for the patient. This ensures both the efficiency of personalized treatment plan development and the suitability of the personalized treatment plan for the patient. In addition, the consultation system can also provide doctors with a wealth of medical knowledge and treatment plans, helping them to better understand the patient's condition and develop more reasonable treatment plans.

[0129] Example 9:

[0130] Based on Embodiment 8, this embodiment of the invention provides a diagnostic system based on artificial intelligence technology, including a scheme adjustment unit, comprising:

[0131] The difference comparison unit is used to compare the patient's physical data with the patient's physical data based on the patient's historical physical data to determine physical differences, and to determine the recovery characteristics of abnormal recovery based on the patient's historical recovery status.

[0132] The adjustment and replacement unit is used to extract the program features corresponding to the physical differences and recovery characteristics from historical treatment programs, extract the target program features that correspond to the patient's physical data from other historical treatment programs, and replace the program features in the historical treatment programs with the target program features to obtain the patient's personalized treatment program.

[0133] In this embodiment, the historical patient data corresponding to other historical treatment plans are the same as the patient's physical data corresponding to the physical differences from the historical treatment plans.

[0134] The beneficial effects of the above design scheme are: by analyzing the historical treatment plans of historical patients based on their physical data and recovery status, and integrating multiple historical treatment plans that meet the patient's physical data, a personalized treatment plan for the patient is obtained. This ensures both the efficiency of developing a personalized treatment plan and the suitability of the personalized treatment plan for the patient.

[0135] Example 10:

[0136] This invention provides a diagnostic device for implementing the artificial intelligence-based diagnostic system described in any one of embodiments 1-9, such as... Figure 3 As shown, it includes:

[0137] Smart wearable devices are worn by patients to collect their physiological information data;

[0138] A smart question-answering device for storing and executing one or more program instructions to perform a system as described in any one of Examples 1-9.

[0139] The beneficial effects of the above design scheme are as follows: Intelligent consultation devices can provide patients with fast and efficient services through channels such as the internet or mobile devices, reducing patient waiting time and medical costs. Through technologies such as natural language processing and machine learning, the intelligent consultation system can accurately identify and analyze patients' conditions, thereby providing more precise diagnostic and treatment suggestions. The intelligent consultation system can also provide doctors with a wealth of medical knowledge and treatment plans, helping them better understand patients' conditions and develop more appropriate treatment plans. The intelligent consultation system can help hospitals optimize the allocation of medical resources, improve the utilization rate of medical resources, and alleviate hospital congestion.

[0140] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A diagnostic system based on artificial intelligence technology, characterized in that, include: The data analysis module is used to receive physiological information data of patients collected by smart wearable devices worn on the patient's body; And big data analysis of physiological information data is conducted to preliminarily determine the patient's physical data; The strategy determination module, based on body data and combined with medical knowledge graphs and machine learning, determines the questioning and response strategies for patient consultations, including: The data acquisition unit is used to acquire first body data in a normal state and second body data in an abnormal state from body data, and to acquire first disease information related to the first body data and second disease information related to the second body data based on the medical knowledge graph. The association unit is used to obtain the first body keywords of the first body data, obtain the first disease keywords of the first disease information, associate the first body keywords and the first disease keywords to obtain the first association result, obtain the second body keywords of the second body data, obtain the second disease keywords of the second disease information, associate the second body keywords and the second disease keywords to obtain the second association result. The expansion unit is used to expand the first association result based on the keyword-related words to obtain the first comprehensive association result, and to expand the second association result to obtain the second comprehensive association result; The script generation unit is used to collect doctors' historical consultation records, analyze and process the historical consultation records to obtain consultation script templates, input the consultation script templates and the first comprehensive association result into a pre-designed disease script generation model to obtain the first script set, and input the consultation script templates and the second comprehensive association result into the pre-designed disease script generation model to obtain the second script set. The strategy determination unit is used to extract the questioning script set from the first script set and the second script set as the questioning strategy, and to extract the response script set from the first script set and the second script set as the response strategy. The intelligent question-and-answer module is used to recognize the patient's voice questions and answers, and combine the question-and-answer strategies to realize the voice consultation process with the patient and obtain consultation information. The results analysis module is used to determine the patient's condition and provide personalized treatment plans based on the consultation information and machine learning.

2. The online consultation system based on artificial intelligence technology according to claim 1, characterized in that, The data analysis module includes: The receiving unit is used to receive physiological information data of the patient collected by a smart wearable device worn on the patient's body; The integration unit is used to integrate physiological information data of the same physiological attribute according to the collection time, and to establish a physiological data change curve under that attribute based on the integration result. The recording unit is used to determine the recording order of physiological data change curves based on the correlation between physiological attributes, and to integrate the physiological data change curves based on the recording order to obtain target physiological information data. The analysis unit is used to perform big data analysis on physiological information data based on the correlation between physiological attributes, determine the evaluation parameters of the target physiological information data under the body evaluation indicators, and preliminarily determine the patient's body data based on the evaluation parameters.

3. The online consultation system based on artificial intelligence technology according to claim 1, characterized in that, The strategy determination module also includes: The model building unit is used to train the initial machine learning model based on the consultation script template to generate a script generation model, and to further train the script generation model based on negative statement keywords to obtain a disease script generation model, and to further train the script generation model based on affirmative statement keywords to obtain a disease script generation model.

4. The online consultation system based on artificial intelligence technology according to claim 1, characterized in that, The intelligent question-answering module includes: The speech recognition unit is used to receive the patient's question and answer speech, and to recognize the question and answer speech to obtain the speech recognition result; The intent determination unit is used to perform language understanding on the speech recognition results, determine the user's intent, and determine the speech response template based on the user's intent; The response confirmation unit is used to fill in the voice response template based on the consultation questioning strategy and consultation response strategy to achieve intelligent voice response; The information determination unit is used to obtain consultation information based on the patient's response.

5. A diagnostic system based on artificial intelligence technology according to claim 4, characterized in that, The intent determination unit includes: The semantic analysis unit is used to classify the speech recognition results into the speech recognition model, obtain semantic features, extract semantic keywords from the speech features, and predict the user's first intent based on the semantic keywords. The above analysis unit is used to determine whether the speech recognition result contains the above content in the speech question and answer process; If so, determine the preceding intent of the preceding content, perform a fluency analysis between the preceding intent and the first intent to determine the fluency value, perform a tightness analysis between the preceding intent and the first intent to determine the tightness value, and adjust the first intent based on the preceding intent, combining the fluency value and the tightness value to obtain the second intent; Otherwise, the user's first intent will be taken as the second intent; The human body analysis unit is used to determine the user's human medical map based on the user's body data, mark abnormal data in the human medical map as key points, decompose the second intent according to medical characteristics, determine the intent distribution characteristics, mark the intent distribution characteristics on the human medical map, and determine the matching degree between the intent distribution characteristics and key points based on the marking results. The intent acquisition unit is used to determine whether the matching degree is greater than the preset matching degree; If so, the second intent should be taken as the user's final intent; Otherwise, based on the deviation between the intent distribution characteristics and the key points, the second intent is corrected to obtain the user's final intent.

6. A diagnostic system based on artificial intelligence technology according to claim 4, characterized in that, The response confirmation unit includes: The extraction unit is used to determine response keywords based on the user's intent and extract consultation strategies related to the response keywords from the consultation questioning strategy and consultation response strategy; The filling unit is used to determine the filling attributes in the voice response template, extract the corresponding consultation keywords from the consultation strategy based on the filling attributes, and fill the consultation keywords into the voice response template to obtain the initial voice response content; The adjustment unit inputs the initial voice response content into the text fluency adjustment model for fluency adjustment to obtain the target voice response content; The voice determination unit is used to match the corresponding response tone based on the tone of the user's question and answer speech, and generate voice text from the target voice response content using the response tone to achieve intelligent voice response.

7. A diagnostic system based on artificial intelligence technology according to claim 1, characterized in that, The results analysis module includes: The disease diagnosis unit is used to train a machine learning model based on the patient's historical medical history and disease information to obtain a disease diagnosis model. The medical history information is input into the disease diagnosis model to determine the patient's disease. The treatment plan acquisition unit is used to acquire historical treatment plans of patients with the same condition as the patient and to determine the recovery status of the historical patients under the historical treatment plans. The treatment plan adjustment unit is used to make personalized adjustments to historical treatment plans based on the historical patients and their corresponding disease recovery status, so as to obtain personalized treatment plans for the patients.

8. A diagnostic system based on artificial intelligence technology according to claim 7, characterized in that, The scheme adjustment unit includes: The difference comparison unit is used to compare the patient's physical data with the patient's physical data based on the patient's historical physical data to determine physical differences, and to determine the recovery characteristics of abnormal recovery based on the patient's historical recovery status. The adjustment and replacement unit is used to extract the program features corresponding to the physical differences and recovery characteristics from historical treatment programs, extract the target program features that correspond to the patient's physical data from other historical treatment programs, and replace the program features in the historical treatment programs with the target program features to obtain the patient's personalized treatment program.

9. A consultation device for implementing the consultation system based on artificial intelligence technology according to any one of claims 1-8, characterized in that, include: Smart wearable devices are worn by patients to collect their physiological information data; A smart question-answering device for storing and executing one or more program instructions to perform the system as described in any one of claims 1-8.

Images