AI-based virtual digital patient system
Through the virtual digital patient system based on AI, real disease scenarios are simulated, the problem of insufficient clinical teaching resources in medical schools is solved, the consultation ability and teaching resource utilization efficiency of medical students are improved, and personalized teaching is realized.
Patent Information
- Application Number
- CN202411428820.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-14
- Publication Date
- 2024-11-26
AI Technical Summary
Medical schools are facing insufficient clinical teaching resources, including the lack of real patients' resources, high teaching risks, difficulty in realizing personalized teaching, limited teaching resources and insufficient opportunities for clinical skills practice.
Using an AI-based virtual digital patient system, by obtaining and generating patient portraits, simulating the patient visit process, supporting medical students to interact with virtual patients in natural language, assessing communication skills and condition assessment capabilities, and providing personalized learning feedback.
It improves medical students' consultation ability and communication ability, reduces their dependence on traditional physical patients and simulated teaching aids, improves the efficiency of teaching resources, and supports personalized teaching and teachers' accurate grasp of students' learning situation.
Smart Images

Figure CN119028199A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of teaching technology, and in particular to an AI-based virtual digital patient system. Background Art
[0002] Currently, many medical schools are facing insufficient clinical teaching resources. 1. Lack of real patient resources: Medical schools are often limited by the number and types of real patient resources when conducting clinical teaching. Patients with different diseases and different stages of their condition are difficult to obtain at any time, and real patients are often unwilling to participate in teaching activities frequently. 2. Reduced teaching risks: There is a high risk of teaching in a real clinical environment, such as misdiagnosis, mistreatment, or unnecessary harm to patients. 3. Personalized teaching: Different students have different learning abilities and progress, and traditional teaching methods are difficult to meet the needs of all students. 4. Limited teaching resources: Medical schools have limited teaching resources, including teachers, teaching equipment, and venues. 5. Little clinical skills practice: Clinical skills are one of the important abilities that medical students must master. In a real clinical environment, students often find it difficult to get enough practice opportunities. Summary of the invention
[0003] The present invention aims to solve at least one of the technical problems in the above-mentioned technologies to a certain extent. To this end, the purpose of the present invention is to propose an AI-based virtual digital patient system, which simulates the process of a patient's consultation and doctor's consultation through complex and changeable symptoms, enhances the medical students' consultation and communication abilities when facing real patients, supports medical students to interact with virtual patients in natural language, simulates the real consultation process, evaluates the medical students' communication skills and disease assessment ability, and facilitates the teacher to accurately grasp the students' learning situation.
[0004] To achieve the above objectives, an embodiment of the present invention proposes an AI-based virtual digital patient system, comprising:
[0005] Acquisition module, used to obtain patient portraits in different scenarios;
[0006] The first generation module is used to generate a large patient model based on patient portraits in different scenarios; wherein the large patient model includes several disease cases;
[0007] The second generation module is used to generate a virtual digital patient based on the disease case;
[0008] The collection module is used to collect the interaction information between the student end and the virtual digital patient, and send the interaction information to the teacher end.
[0009] According to some embodiments of the present invention, the acquisition module is further used to acquire diagnostic information generated by the student terminal based on the interactive information and send the diagnostic information to the teacher terminal.
[0010] According to some embodiments of the present invention, the acquisition module includes:
[0011] The first building block is used to:
[0012] Obtain basic information about the patient;
[0013] The basic information is parsed to determine the scene information and content information, and the scene information is divided into a number of scene nodes based on the preset scene categories; the scene information includes the patient in different scenes; the content information includes symptom description, medical records and behavioral observations;
[0014] According to the logical relationship between the scene categories, a node level is assigned to each scene node, and the connection relationship between the scene nodes is determined according to the node level; the connection relationship includes a time sequence and a spatial position relationship;
[0015] Filling the content information into the corresponding scene node to obtain a filled scene node; the filled scene node represents the state of the patient at different times, places or situations;
[0016] Constructing a scene tree according to the filled scene nodes and the connection relationship;
[0017] The first determination module is used to:
[0018] Based on data mining and machine learning, feature extraction is performed on the content data in each scene node in the scene tree to determine the patient labels in different scenes;
[0019] Build patient portraits for different scenarios based on patient labels in different scenarios.
[0020] According to some embodiments of the present invention, the method further includes: an updating module, which is used to update the basic information of the patient and update the patient portrait based on the updated basic information of the patient.
[0021] According to some embodiments of the present invention, the first generating module includes:
[0022] The second building block is used to build a convolutional neural network model;
[0023] A training module is used for the patient portraits including static portraits and dynamic portraits; a neural network model is trained based on the static portraits and the dynamic portraits, and during the training process, the neural network model is evaluated based on a cross-validation method, and when the evaluation is qualified, a large patient model is generated.
[0024] According to some embodiments of the present invention, it also includes: an analysis module for analyzing disease cases, evaluating disease progression, and providing diagnosis and treatment recommendations.
[0025] According to some embodiments of the present invention, the present invention further includes: a second determining module, configured to:
[0026] Compare the diagnostic information with the treatment recommendations and determine the student's deficiencies based on the comparison results;
[0027] Based on the deficiencies, inquiries are made in the patient model, and the corresponding recommended cases are determined and sent to the student end.
[0028] According to some embodiments of the present invention, it also includes: a case management module, which is used to provide a case library management function, and the case library management function includes adding, editing and sharing cases.
[0029] According to some embodiments of the present invention, it further includes: a preprocessing module, which is used to preprocess the basic information before the first building module performs information parsing on the basic information;
[0030] The preprocessing module comprises:
[0031] Speech preprocessing module for:
[0032] Inputting the speech information included in the basic information into a pre-trained speech model to determine a time-frequency spectrogram corresponding to the speech information;
[0033] Determine the spectrum envelope of the speech information according to the time-frequency spectrogram, determine the amplitude of each formant on the spectrum envelope, and calculate the average amplitude;
[0034] Determine a filter coefficient according to the average amplitude, input the voice information corresponding to the filter coefficient into the filter for filtering, and obtain preprocessed voice information;
[0035] Image preprocessing module for:
[0036] Performing quality assessment on the image information included in the basic information, and determining an image quality coefficient of the image information;
[0037]
[0038] Where S is the image quality coefficient of the image information; T 1 is the pixel gradient of the image information in the vertical direction; T 2 is the pixel gradient of the image information in the horizontal direction; M θ is the number of pixels with an angle of θ in the image information; T 1(x) is the pixel gradient of the image information at x in the vertical direction; T 2(x) The pixel gradient of the image information at x in the horizontal direction; || || 0 For L 0norm, used to count the number of non-zero vectors;
[0039] Screen out images whose image quality coefficient is less than a preset threshold as images to be processed;
[0040] Perform image enhancement processing on the image to be processed to obtain preprocessed image information;
[0041] The first building module is used to perform information analysis on the preprocessed voice information and the preprocessed image information.
[0042] According to some embodiments of the present invention, it also includes: a network delay processing module, which is used to capture the network logs of the student end and the teacher end when the student end and the teacher end interact with each other in teaching, obtain the delay data in the network logs, and obtain a network delay point set; and perform data update processing on the network delay point set.
[0043] This invention proposes an AI-based virtual digital patient system, which simulates real disease scenarios through highly simulated virtual patient roles, providing medical students with an immersive and interactive consultation practice platform. Medical students can not only communicate freely with virtual patients, but also use medical knowledge to make diagnostic inferences based on the patient's symptom description, medical history inquiry and feedback, effectively training their consultation ability and thinking. It reduces dependence on traditional physical patients and simulation teaching aids, and improves the utilization efficiency of teaching resources.
[0044] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.
[0045] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:
[0047] Figure 1 is a block diagram of an AI-based virtual digital patient system according to an embodiment of the present invention;
[0048] Figure 2 is a block diagram of an acquisition module according to an embodiment of the present invention;
[0049] Figure 3 is a block diagram of a first generating module according to an embodiment of the present invention. DETAILED DESCRIPTION
[0050] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.
[0051] like Figure 1 As shown, the embodiment of the present invention proposes an AI-based virtual digital patient system, including:
[0052] Acquisition module, used to obtain patient portraits in different scenarios;
[0053] The first generation module is used to generate a large patient model based on patient portraits in different scenarios; wherein the large patient model includes several disease cases;
[0054] The second generation module is used to generate a virtual digital patient based on the disease case;
[0055] The collection module is used to collect the interaction information between the student end and the virtual digital patient, and send the interaction information to the teacher end.
[0056] The working principle of the above technical solution is: generate a large patient model based on patient portraits in different scenarios, and generate a virtual digital patient based on the disease cases in the large patient model to facilitate students to interact with the virtual digital patient.
[0057] The beneficial effects of the above technical solution are: through highly simulated virtual patient roles, simulating real disease scenarios, providing medical students with an immersive and interactive consultation practice platform. Medical students can not only communicate freely with virtual patients, but also use medical knowledge to make diagnostic inferences based on the patient's symptom description, medical history inquiry and feedback, effectively training their consultation ability and thinking. Reduce dependence on traditional physical patients and simulation teaching aids, and improve the utilization efficiency of teaching resources.
[0058] According to some embodiments of the present invention, the acquisition module is further used to acquire diagnostic information generated by the student terminal based on the interactive information and send the diagnostic information to the teacher terminal.
[0059] The beneficial effects of the above technical solution are: it is convenient for teachers to view the diagnostic information generated by students based on the interactive information, judge the accuracy and shortcomings of the students' diagnosis, facilitate subsequent targeted teaching, and improve teaching effects.
[0060] like Figure 2 As shown, according to some embodiments of the present invention, the acquisition module includes:
[0061] The first building block is used to:
[0062] Obtain basic information about the patient;
[0063] The basic information is parsed to determine the scene information and content information, and the scene information is divided into a number of scene nodes based on the preset scene categories; the scene information includes the patient in different scenes; the content information includes symptom description, medical records and behavioral observations;
[0064] According to the logical relationship between the scene categories, a node level is assigned to each scene node, and the connection relationship between the scene nodes is determined according to the node level; the connection relationship includes a time sequence and a spatial position relationship;
[0065] Filling the content information into the corresponding scene node to obtain a filled scene node; the filled scene node represents the state of the patient at different times, places or situations;
[0066] Constructing a scene tree according to the filled scene nodes and the connection relationship;
[0067] The first determination module is used to:
[0068] Based on data mining and machine learning, feature extraction is performed on the content data in each scene node in the scene tree to determine the patient labels in different scenes;
[0069] Build patient portraits for different scenarios based on patient labels in different scenarios.
[0070] Working principle and beneficial effects of the above technical solution: Basic information includes name, age, gender, medical history, etc. The basic information is parsed to identify different scenarios experienced by the patient, such as home, hospital ward, clinic, etc. According to the preset scenario categories (such as daily activities, diagnosis and treatment process, rehabilitation period, etc.), the identified scenario information is divided into different scenario nodes. According to the logical relationship between the scenario categories, a node hierarchy is assigned to each scenario node, and the connection relationship between the various scenario nodes is determined according to the node hierarchy; the connection relationship includes the time sequence and the spatial position relationship; it is ensured that the patient's state changes at different times, places or situations can be accurately reflected. Data mining techniques (such as clustering, association rule mining, etc.) and machine learning algorithms (such as classification, regression, etc.) are used to conduct in-depth analysis of the content data. Key features such as symptom characteristics, behavioral characteristics, environmental characteristics, etc. are extracted for subsequent patient label determination. According to the extracted features, the corresponding patient labels are assigned to the content data under each scenario node. These labels include disease type, severity of the disease, treatment response, etc. Ensure the accuracy and consistency of the labels so that accurate patient portraits can be constructed later. Each patient portrait should include a detailed description of the patient's characteristics and status assessment in that scenario, so as to accurately determine the patient portraits in different scenarios.
[0071] According to some embodiments of the present invention, the method further includes: an updating module, which is used to update the basic information of the patient and update the patient portrait based on the updated basic information of the patient.
[0072] The working principle and beneficial effects of the above technical solution are: it is easy to ensure the accuracy of the patient portrait and improve the accuracy of building a large patient model.
[0073] like Figure 3 As shown, according to some embodiments of the present invention, the first generating module includes:
[0074] The second building block is used to build a convolutional neural network model;
[0075] A training module is used for the patient portraits including static portraits and dynamic portraits; a neural network model is trained based on the static portraits and the dynamic portraits, and during the training process, the neural network model is evaluated based on a cross-validation method, and when the evaluation is qualified, a large patient model is generated.
[0076] The working principle and beneficial effects of the above technical solution: Static portrait: Based on the patient's basic information and long-term medical history data, a static portrait reflecting the patient's basic situation and long-term health status is constructed. This includes information such as the patient's age, gender, occupation, living habits, and medical information such as disease type and severity of the disease. Dynamic portrait: Dynamically update the static portrait based on the patient's real-time health data (such as vital signs monitoring, medication records, behavioral data, etc.). The dynamic portrait can more accurately reflect the patient's current health status and disease change trends. The data set is divided into a training set, a validation set, and a test set. In each round of training, the training set is used to train the model, and then the validation set is used to evaluate the performance of the model. Cross-validation can use methods such as K-fold cross-validation or leave-one-out cross-validation to ensure the stability and reliability of the evaluation results. When the evaluation is qualified, a large model of the patient is generated.
[0077] According to some embodiments of the present invention, it also includes: an analysis module for analyzing disease cases, evaluating disease progression, and providing diagnosis and treatment recommendations.
[0078] The beneficial effects of the above technical solution are as follows: based on the automatic analysis of case studies, the disease progression is evaluated, and diagnosis and treatment recommendations are given, which serve as a relatively standard diagnosis and treatment recommendation, so as to facilitate comparison with the diagnosis information obtained by students and determine whether the student's diagnosis is accurate.
[0079] According to some embodiments of the present invention, the present invention further includes: a second determining module, configured to:
[0080] Compare the diagnostic information with the treatment recommendations and determine the student's deficiencies based on the comparison results;
[0081] Based on the deficiencies, inquiries are made in the patient model, and the corresponding recommended cases are determined and sent to the student end.
[0082] The working principle and beneficial effects of the above technical solution are as follows: Compare the diagnostic information with the treatment recommendations, and determine the deficiencies of the students based on the comparison results; query the patient model based on the deficiencies, determine the corresponding recommended cases and send them to the student end. According to the learning progress and performance of medical students, intelligently recommend appropriate learning content and difficulty levels to promote personalized growth.
[0083] According to some embodiments of the present invention, it also includes: a case management module, which is used to provide a case library management function, and the case library management function includes adding, editing and sharing cases.
[0084] The working principle and beneficial effects of the above technical solution: Teachers can manage the case library and adjust it according to teaching needs.
[0085] According to some embodiments of the present invention, it further includes: a preprocessing module, which is used to preprocess the basic information before the first building module performs information parsing on the basic information;
[0086] The preprocessing module comprises:
[0087] Speech preprocessing module for:
[0088] Inputting the speech information included in the basic information into a pre-trained speech model to determine a time-frequency spectrogram corresponding to the speech information;
[0089] Determine the spectrum envelope of the speech information according to the time-frequency spectrogram, determine the amplitude of each formant on the spectrum envelope, and calculate the average amplitude;
[0090] Determine a filter coefficient according to the average amplitude, input the voice information corresponding to the filter coefficient into the filter for filtering, and obtain preprocessed voice information;
[0091] Image preprocessing module for:
[0092] Performing quality assessment on the image information included in the basic information, and determining an image quality coefficient of the image information;
[0093]
[0094] Where S is the image quality coefficient of the image information; T 1 is the pixel gradient of the image information in the vertical direction; T 2 is the pixel gradient of the image information in the horizontal direction; M θ is the number of pixels with an angle of θ in the image information; T1(x) is the pixel gradient of the image information at x in the vertical direction; T 2(x) The pixel gradient of the image information at x in the horizontal direction; || || 0 For L 0 norm, used to count the number of non-zero vectors;
[0095] Screen out images whose image quality coefficient is less than a preset threshold as images to be processed;
[0096] Perform image enhancement processing on the image to be processed to obtain preprocessed image information;
[0097] The first building module is used to perform information analysis on the preprocessed voice information and the preprocessed image information.
[0098] The working principle of the above technical solution is: the input speech information is sent to a pre-trained speech model. This model is based on deep learning (such as autoencoder, recurrent neural network RNN, etc.), which is used to convert the original audio waveform into a data that is easy to analyze and generate a time-frequency spectrogram: the speech signal is converted from the time domain to the time-frequency domain through short-time Fourier transform (STFT) or other time-frequency analysis methods to generate a time-frequency spectrogram. The time-frequency spectrogram shows the frequency distribution of the speech signal over time. Determine the spectrum envelope: On the time-frequency spectrogram, the spectrum envelope is estimated by smoothing or fitting techniques (such as linear predictive coding LPC). The spectrum envelope describes the spectral profile of the speech signal and is crucial for identifying features such as vowels and consonants in speech. Formant detection: On the spectrum envelope, prominent frequency points, namely formants, are identified. These points correspond to the resonance enhancement of the sound frequency by the shape of the vocal tract. Calculate the average amplitude: Calculate the average of all detected formant amplitudes as a benchmark for subsequent processing. According to the calculated average amplitude, design or adjust the filter coefficient of the filter. These coefficients determine how the filter modifies the spectrum of the input signal through frequency response. For example, a bandpass filter can be used to enhance or suppress sounds in a specific frequency range. The filter coefficients are applied to the input voice information and filtered through the filter. This step aims to improve the signal-to-noise ratio of the voice signal, remove unnecessary frequency components, or emphasize certain specific voice features. Based on the image preprocessing module, the image quality coefficient of the image information is accurately calculated, and images with image quality coefficients less than a preset threshold are screened out as images to be processed; image enhancement processing is performed on the images to be processed, including histogram equalization, contrast adjustment, sharpening, etc., to improve the contrast, clarity and visibility of the image.
[0099] The beneficial effects of the above technical solution are: obtaining preprocessed voice information and preprocessed image information, that is, obtaining accurate basic information, and improving the efficiency and accuracy of information analysis when analyzing the basic information.
[0100] According to some embodiments of the present invention, it also includes: a network delay processing module, which is used to capture the network logs of the student end and the teacher end when the student end and the teacher end interact with each other in teaching, obtain the delay data in the network logs, and obtain a network delay point set; and perform data update processing on the network delay point set.
[0101] The working principle and beneficial effects of the above technical solution are as follows: when the student end and the teacher end are conducting teaching interaction, the network logs of the student end and the teacher end are captured, the delay data in the network logs are obtained, and a network delay point set is obtained; data update processing is performed on the network delay point set, that is, the influence of network delay on the teaching interaction is eliminated, the efficiency of the teaching interaction is improved, and the teaching effect is improved.
[0102] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.
Claims
1. An AI-based virtual digital patient system, It is characterized in that include: Acquisition module, used to obtain patient portraits in different scenarios; The first generation module is used to generate a large patient model based on patient portraits in different scenarios; The patient model includes several disease cases; The second generation module is used to generate a virtual digital patient based on the disease case; The collection module is used to collect the interaction information between the student end and the virtual digital patient, and send the interaction information to the teacher end.
2. The AI-based virtual digital patient system according to claim 1, It is characterized in that The acquisition module is also used to acquire the diagnostic information generated by the student terminal according to the interactive information and send it to the teacher terminal.
3. The AI-based virtual digital patient system according to claim 1, It is characterized in that The acquisition module comprises: The first building block is used to: Obtain basic information about the patient; The basic information is parsed to determine the scene information and content information, and the scene information is divided into a number of scene nodes based on the preset scene categories; the scene information includes the patient in different scenes; the content information includes symptom description, medical records and behavioral observations; According to the logical relationship between the scene categories, a node level is assigned to each scene node, and the connection relationship between the scene nodes is determined according to the node level; the connection relationship includes a time sequence and a spatial position relationship; Filling the content information into the corresponding scene node to obtain a filled scene node; the filled scene node represents the state of the patient at different times, places or situations; Constructing a scene tree according to the filled scene nodes and the connection relationship; The first determination module is used to: Based on data mining and machine learning, feature extraction is performed on the content data in each scene node in the scene tree to determine the patient labels in different scenes; Build patient portraits for different scenarios based on patient labels in different scenarios.
4. The AI-based virtual digital patient system according to claim 1, It is characterized in that Also includes: The update module is used to update the patient's basic information and update the patient's portrait based on the updated patient's basic information.
5. The AI-based virtual digital patient system according to claim 1, It is characterized in that The first generation module includes: The second building block is used to build a convolutional neural network model; A training module is used for the patient portraits including static portraits and dynamic portraits; a neural network model is trained based on the static portraits and the dynamic portraits, and during the training process, the neural network model is evaluated based on a cross-validation method, and when the evaluation is qualified, a large patient model is generated.
6. The AI-based virtual digital patient system according to claim 2, It is characterized in that Also includes: The analysis module is used to analyze disease cases, evaluate disease progression, and provide diagnosis and treatment recommendations.
7. The AI-based virtual digital patient system according to claim 6, It is characterized in that Also includes: The second determination module is used for: Compare the diagnostic information with the treatment recommendations and determine the student's deficiencies based on the comparison results; Based on the deficiencies, a query is made in the patient model, and the corresponding recommended cases are determined and sent to the student end.
8. The AI-based virtual digital patient system according to claim 1, It is characterized in that Also includes: The case management module is used to provide case library management functions, which include adding, editing and sharing cases.
9. The AI-based virtual digital patient system according to claim 3, It is characterized in that Also includes: A preprocessing module, used for preprocessing the basic information before the first building module parses the basic information; The preprocessing module comprises: Speech preprocessing module, used for: Inputting the speech information included in the basic information into a pre-trained speech model to determine a time-frequency spectrogram corresponding to the speech information; Determine the spectrum envelope of the speech information according to the time-frequency spectrogram, determine the amplitude of each formant on the spectrum envelope, and calculate the average amplitude; Determine a filter coefficient according to the average amplitude, input the voice information corresponding to the filter coefficient into the filter for filtering, and obtain preprocessed voice information; Image preprocessing module for: Performing quality assessment on the image information included in the basic information, and determining an image quality coefficient of the image information; Where S is the image quality coefficient of the image information; T 1 is the pixel gradient of the image information in the vertical direction; T 2 is the pixel gradient of the image information in the horizontal direction; M θ is the number of pixels with an angle of θ in the image information; T 1(x) is the pixel gradient of the image information at x in the vertical direction; T 2(x) The pixel gradient of the image information at x in the horizontal direction; || || 0 For L 0 norm, used to count the number of non-zero vectors; Screen out images whose image quality coefficient is less than a preset threshold as images to be processed; Perform image enhancement processing on the image to be processed to obtain preprocessed image information; The first building module is used to perform information analysis on the preprocessed voice information and the preprocessed image information.
10. The AI-based virtual digital patient system according to claim 1, It is characterized in that Also includes: The network delay processing module is used to capture the network logs of the student end and the teacher end when the student end and the teacher end interact with each other in teaching, obtain the delay data in the network logs, and obtain the network delay point set; and perform data update processing on the network delay point set.
Citation Information
Patent Citations
Portable target-controlled infusion intravenous anesthesia device and target-controlled infusion control method
CN113648484A
Wristwatch maintenance management system based on Internet technology
CN114548442A
Human-computer interaction method for hypertension inquiry and follow-up visit scene
CN117235241A
Digital patient-based general practice doctor-patient communication training assessment method
CN117593932A
General practice doctor-patient communication training examination method based on virtual reality technology
CN117854341A
Cited By
A diagnosis and treatment simulation method, system, medium and program product based on digital human
CN119741173A
A diagnosis and treatment simulation method, system, medium and program product based on digital humans
CN119741173B
Teaching target feedback method and system based on artificial intelligence model
CN120015327A
Intelligent evaluation and feedback system for simulating patient teaching based on dialogue analysis
CN120655466A
Standardized training system based on large model and virtual generation technology
CN121050570A
