Medical knowledge development and construction platform based on artificial intelligence

Abstract

The invention relates to the technical field of medical artificial intelligence, in particular to a medical knowledge development and construction platform based on artificial intelligence, which integrates functions of dynamic knowledge base construction, personalized recommendation and clinical decision support. The invention discloses medical knowledge development based on artificial intelligence. The medical knowledge development comprises a medical data preprocessing module, a medical knowledge graph construction module and an artificial intelligence model training module, a deep learning algorithm, a reinforcement learning algorithm, a poison training framework, a federal learning framework, a self-research ant combat training module and a self-research weak meat and forceful food algorithm module are integrated, and pre-training data information integration is performed through the medical knowledge graph and medical data preprocessing module; the pre-training data information is integrated and trained to obtain data information digits for multi-model training, and the multi-model training generates an interpretable medical diagnosis model and a personalized treatment recommendation model. The medical knowledge data information is digitally stored, and the security meets the scale.

Description

Technical Field

[0001] The present invention relates to the field of medical artificial intelligence technology, and more specifically, to an artificial intelligence-based medical knowledge development and construction platform that integrates dynamic knowledge base construction, personalized recommendation, and clinical decision support functions. Background Art

[0002] Knowledge needs of doctors: Doctors currently face the following core challenges in clinical practice and continuous learning:

[0003] 1. Knowledge overload and fragmentation: The annual growth rate of medical literature exceeds 8%, making it difficult for doctors to efficiently screen key information (according to statistics from Nature);

[0004] 2. Lagging knowledge updates: Traditional medical textbooks / guidelines have a long update cycle (1-3 years on average), and are unable to synchronize cutting-edge research results in real time;

[0005] 3. Insufficient personalized matching: The general knowledge base cannot adapt to the needs of doctors' specialties (such as interventional cardiology vs. heart failure treatment), practice levels (residents vs. chief physicians), and case scenarios;

[0006] 4. Weak clinical decision support: Existing tools mostly rely on static databases and lack dynamic reasoning capabilities based on individual patient data.

[0007] Disadvantages of existing solutions:

[0008] The prior art is as follows:

[0009] 1. The core flaw of medical literature databases (such as PubMed) is the passive retrieval mode, which requires manual screening and is inefficient.

[0010] 2. Clinical decision support system (CDSS): The core flaw is that the rule engine is rigid and difficult to adapt to complex cases and the latest medical evidence.

[0011] 3. The core flaw of general medical knowledge APP is the homogeneity of content and the lack of deep integration with doctors' practice scenarios.

[0012] 4. The core defects of the hospital's internal training system are slow knowledge updating, poor interactivity, and lack of intelligent push mechanism. Summary of the Invention

[0013] 1. A medical knowledge development and construction platform based on artificial intelligence, characterized by comprising: step 1, a medical data preprocessing module; medical data capture obtains structured information data from multi-source heterogeneous medical data, and targets the data source and credibility level for verification and identification, the first verification and identification data information is verified, the second verification and identification data information is screened and cleaned, and the process is repeated three times to obtain accurate information data for medical knowledge data information digital storage collection, the medical knowledge data information digital storage collection includes: medical literature, clinical diagnosis and treatment records, medical imaging and genomic data, and the first verification and identification data information is verified, the second verification and identification data information is screened and cleaned, and the process is repeated three times to obtain accurate information data for medical knowledge data information digital storage; the medical knowledge data information digital storage is cleaned, de-identified and standardized to generate high-quality training data integration;

[0014] Step 2, medical knowledge graph construction module; based on the integrated natural language processing technology, the pre-processed medical data information is analyzed, and according to the method of step 1 in claim 1, the data information digital parameters are extracted to generate a medical knowledge graph construction module data information integration training system, and the medical knowledge graph construction module data information integration training system of the claim generates a dynamically updated medical knowledge graph, and the knowledge graph contains multi-dimensional associations of diseases, symptoms, drugs, treatment plans and medical research progress;

[0015] Step 3: The artificial intelligence model training module includes: integrating a deep learning algorithm, a reinforcement learning algorithm, a poisoning training framework, a federated learning framework, a self-developed ant combat training module, and a self-developed law of the jungle algorithm module; pre-training data information is integrated through the medical knowledge graph and the medical data pre-processing module; the pre-training data information integration training obtains data information numbers for multi-model training; the multi-model training generates an explainable medical diagnosis model and a personalized treatment recommendation model;

[0016] Step 4, the medical knowledge data information is digitally stored and dynamically updated. The data information of the first dynamic update is identified by multi-model training for pre-training. The data information of the second dynamic update is identified by the medical knowledge graph construction module for artificial intelligence multi-model training and verification module. The verification module is a training and learning module obtained by digitally verifying, identifying and calculating the information data of the medical data preprocessing module and the medical knowledge graph construction module according to steps 1 and 2 of the claim. The training and learning module monitors the medical database update and user feedback data information in real time, identifies and calculates the active learning mechanism to optimize the digital parameters of the knowledge graph and model data information, and evaluates the model performance based on the digital integration of clinical verification data information;

[0017] Step 5: Security compliance module. The security compliance module has a built-in HIPAA and GDPR-compliant multi-model training system. The multi-model training system identifies user ID information data and performs multi-model training.

[0018] First dimension module: used to receive access requests for user ID information data. The built-in HIPAA and GDPR-compliant multi-model training system obtains the access request and implements data anonymization and encrypted transmission identification verification through the identification privacy enhancement technology (PETs) system.

[0019] The second dimension module: through the privacy enhancement technology (PETs) system, is used to receive the user ID described by the first dimension module for requesting access information data verification and identification, and for second verification identification of ID information data through the requesting access information data verification and identification, and for screening target marking by calculating the third identification verification ID data information through the first and second verification information data identification, and the screening target marking marks the user ID tracking information data for requesting access.

[0020] The third dimension module: ID access request information enters the identification verification information data digital claim step 1 to step 4 information data including: medical data preprocessing module, medical knowledge data information digital reserve collection, medical knowledge data information digital storage, generation of high-quality training data integration, medical knowledge graph construction module, artificial intelligence model training module, data information digital integration for model performance evaluation, and security compliance module.

[0021] 2. The platform according to claim 1, wherein the personalized learning engine module comprises:

[0022] Generate personalized learning paths based on the doctor's professional field, skill level and historical learning data, and dynamically adjust the learning content and difficulty through reinforcement learning algorithms.

[0023] 3. The platform according to claim 1 is characterized in that the intelligent recommendation module: based on the knowledge graph and learning engine, recommends matching medical cases, the latest research results and simulated diagnosis and treatment tasks to doctors, jointly optimizes disease prediction, treatment plan recommendation and prognosis analysis tasks through a multi-task learning framework, provides a model interpretability interface, and outputs key medical features and literature supporting evidence for diagnosis.

[0024] 4. The platform according to claim 1 is characterized in that the simulation diagnosis and treatment and feedback module provides a virtual patient interaction interface, generates simulated cases based on real clinical data, and provides real-time evaluation and feedback on doctors' diagnosis and treatment decisions.

[0025] 5. The platform according to claim 1, wherein the personalized learning engine module comprises:

[0026] 5.1 Skills Assessment Unit: quantifies the doctor's professional competence level by analyzing the doctor's decision-making accuracy, response time, and logical integrity in simulated diagnosis and treatment;

[0027] 5.2 Adaptive learning unit, dynamically adjusts the depth, breadth and presentation of learning content based on evaluation results.

[0028] 6. The platform according to claim 1 is characterized in that the intelligent recommendation module further includes: a multi-objective optimization algorithm for balancing the timeliness, clinical relevance and doctor interest preferences of the recommended content, and a collaborative filtering unit that combines anonymous learning data of the doctor group to generate a recommendation list driven by group intelligence.

[0029] 7. The platform according to claim 1 is characterized in that the simulation diagnosis and feedback module includes: a virtual case generation unit based on generative adversarial network (GAN) to simulate diverse clinical scenarios; a decision tree analysis unit to perform retrospective analysis on the doctor's diagnostic path and generate visual improvement suggestions.

[0030] 8. The platform according to claim 1, further comprising:

[0031] 8.1: Privacy protection module, which uses federated learning technology to conduct distributed training on doctor learning data to ensure that the original data does not leave the local terminal. The data desensitization unit encrypts and anonymizes sensitive information in case data.

[0032] 8.2: The platform according to any one of claims 1-7 is characterized in that the platform is further integrated with medical equipment, including: an interface unit for medical imaging equipment, used to obtain imaging data in real time and associate it with diagnostic criteria in the knowledge graph; and a surgical robot collaboration unit, which trains the robot-assisted decision-making model by simulating diagnosis and treatment data.

[0033] 9. An application method based on the platform according to any one of claims 1 to 8, characterized in that it comprises the following steps:

[0034] Step S1: Collect the professional background data and historical learning records of the target doctor;

[0035] Step S2: Generate a medical knowledge network associated with its professional field through the knowledge graph construction module;

[0036] Step S3: Formulate phased learning goals and tasks based on the personalized learning engine;

[0037] Step S4: Complete case practice in the simulated diagnosis and treatment module and optimize the learning path based on the feedback results.

[0038] 10. The method according to claim 9, wherein the learning task in step S3 comprises:

[0039] 10.1: Mandatory learning content: required knowledge updated according to the latest guidelines or regulations in the medical industry. Optional learning content: customized learning modules based on the doctor's interests and career plans.

[0040] 10.2: The method according to claim 8 is characterized in that the platform uses blockchain technology to store doctors' learning records, simulated diagnosis and treatment results, and ability assessment results, and generates an unalterable electronic ability certificate. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a flowchart developed based on artificial intelligence medical knowledge in an embodiment of the present invention.

[0042] Figure 2 This is a flowchart of a security compliance module in the development of artificial intelligence medical knowledge in an embodiment of the present invention.

[0043] Figure 3 This is a flow chart of a platform for building artificial intelligence medical knowledge in an embodiment of the present invention.

[0044] Figure 4 This is an application method in an artificial intelligence medical knowledge building platform in an embodiment of the present invention. DETAILED DESCRIPTION

[0045] like Figure 1 As shown, an embodiment of the present invention provides a flowchart based on artificial intelligence medical knowledge development, including:

[0046] Step 1, medical data preprocessing module; medical data capture obtains structured information data from multi-source heterogeneous medical data, and targets the data source and credibility level for verification and identification. The first verification and identification data is verified, the second verification and identification data are screened and cleaned, and the cycle is repeated three times to obtain accurate information data for medical knowledge data digital storage collection. The medical knowledge data digital storage collection includes: medical literature, clinical diagnosis and treatment records, medical imaging and genomic data, and the first verification and identification data are verified, the second verification and identification data are screened and cleaned, and the cycle is repeated three times to obtain accurate information data for medical knowledge data digital storage; the medical knowledge data digital storage is cleaned, de-identified and standardized to generate high-quality training data integration;

[0047] Step 2, medical knowledge graph construction module; based on the integrated natural language processing technology, the pre-processed medical data information is analyzed, and according to the method of step 1 in claim 1, the data information digital parameters are extracted to generate a medical knowledge graph construction module data information integration training system, and the medical knowledge graph construction module data information integration training system of the claim generates a dynamically updated medical knowledge graph, and the knowledge graph contains multi-dimensional associations of diseases, symptoms, drugs, treatment plans and medical research progress;

[0048] Step 3: The artificial intelligence model training module includes: integrating a deep learning algorithm, a reinforcement learning algorithm, a poisoning training framework, a federated learning framework, a self-developed ant combat training module, and a self-developed law of the jungle algorithm module; pre-training data information is integrated through the medical knowledge graph and the medical data pre-processing module; the pre-training data information integration training obtains data information numbers for multi-model training; the multi-model training generates an explainable medical diagnosis model and a personalized treatment recommendation model;

[0049] Step 4, the medical knowledge data information is digitally stored and dynamically updated. The data information of the first dynamic update is identified by multi-model training for pre-training. The data information of the second dynamic update is identified by the medical knowledge graph construction module for artificial intelligence multi-model training and verification module. The verification module is a training and learning module obtained by digitally verifying, identifying and calculating the information data of the medical data preprocessing module and the medical knowledge graph construction module according to steps 1 and 2 of the claim. The training and learning module monitors the medical database update and user feedback data information in real time, identifies and calculates the active learning mechanism to optimize the digital parameters of the knowledge graph and model data information, and evaluates the model performance based on the digital integration of clinical verification data information;

[0050] Step 5: Security compliance module. The security compliance module has a built-in HIPAA and GDPR-compliant multi-model training system. The multi-model training system identifies user ID information data and performs multi-model training.

[0051] First dimension module: used to receive access requests for user ID information data. The built-in HIPAA and GDPR-compliant multi-model training system obtains the access request and implements data anonymization and encrypted transmission identification verification through the identification privacy enhancement technology (PETs) system.

[0052] The second dimension module: through the privacy enhancement technology (PETs) system, is used to receive the user ID described by the first dimension module for requesting access information data verification and identification, and for second verification identification of ID information data through the requesting access information data verification and identification, and for screening target marking by calculating the third identification verification ID data information through the first and second verification information data identification, and the screening target marking marks the user ID tracking information data for requesting access.

[0053] The third dimension module: ID access request information enters the identification verification information data digital claim step 1 to step 4 information data including: medical data preprocessing module, medical knowledge data information digital reserve collection, medical knowledge data information digital storage, generation of high-quality training data integration, medical knowledge graph construction module, artificial intelligence model training module, data information digital integration for model performance evaluation, and security compliance module.

[0054] like Figure 2 This is a flow chart of a security compliance module in the development of artificial intelligence medical knowledge in an embodiment of the present invention: the security compliance module has a built-in HIPAA and GDPR-compliant multi-model training system, which identifies user ID information data and performs multi-model training.

[0055] First dimension module: used to receive access requests for user ID information data. The built-in HIPAA and GDPR-compliant multi-model training system obtains the access request and implements data anonymization and encrypted transmission identification verification through the identification privacy enhancement technology (PETs) system.

[0056] The second dimension module: through the privacy enhancement technology (PETs) system, is used to receive the user ID described by the first dimension module for requesting access information data verification and identification, and for second verification identification of ID information data through the requesting access information data verification and identification, and for screening target marking by calculating the third identification verification ID data information through the first and second verification information data identification, and the screening target marking marks the user ID tracking information data for requesting access.

[0057] The third dimension module: ID access request information enters the identification verification information data digital claim step 1 to step 4 information data including: medical data preprocessing module, medical knowledge data information digital reserve collection, medical knowledge data information digital storage, generation of high-quality training data integration, medical knowledge graph construction module, artificial intelligence model training module, data information digital integration for model performance evaluation, and security compliance module.

[0058] like Figure 3This is a flow chart of a platform for building artificial intelligence medical knowledge in an embodiment of the present invention, including:

[0059] The personalized learning engine modules include:

[0060] Generate personalized learning paths based on the doctor's professional field, skill level and historical learning data, and dynamically adjust the learning content and difficulty through reinforcement learning algorithms.

[0061] Smart recommendation module:

[0062] Based on the knowledge graph and learning engine, it recommends matching medical cases, the latest research results and simulated diagnosis and treatment tasks to doctors. Through the multi-task learning framework, it jointly optimizes disease prediction, treatment plan recommendation and prognosis analysis tasks, provides a model interpretability interface, and outputs key medical features and literature supporting evidence for diagnosis.

[0063] Simulated diagnosis and treatment and feedback module: provides a virtual patient interaction interface, generates simulated cases based on real clinical data, and provides real-time evaluation and feedback on doctors' diagnosis and treatment decisions.

[0064] The personalized learning engine modules include:

[0065] (1): Skill assessment unit, which quantifies the professional ability level of doctors by analyzing their decision-making accuracy, response time and logical integrity in simulated diagnosis and treatment;

[0066] (2): Adaptive learning unit, dynamically adjusts the depth, breadth and presentation of learning content based on evaluation results.

[0067] The intelligent recommendation module further includes: a multi-objective optimization algorithm for balancing the timeliness, clinical relevance and doctor interest preferences of the recommended content; a collaborative filtering unit that combines anonymous learning data from the doctor group to generate a recommendation list driven by group intelligence.

[0068] The simulation diagnosis and treatment and feedback module includes: a virtual case generation unit based on generative adversarial networks (GANs) to simulate diverse clinical scenarios; a decision tree analysis unit to retrospectively analyze the doctor's diagnostic path and generate visual improvement suggestions.

[0069] The privacy protection module uses federated learning technology to perform distributed training on doctor learning data to ensure that the original data does not leave the local terminal. The data desensitization unit encrypts and anonymizes sensitive information in case data.

[0070] The platform according to any one of claims 1 to 7 is characterized in that the platform is further integrated with medical equipment, including: an interface unit for medical imaging equipment, used to acquire imaging data in real time and associate it with diagnostic criteria in the knowledge graph; and a surgical robot collaboration unit, which trains a robot-assisted decision-making model by simulating diagnosis and treatment data.

[0071] An application method based on the platform according to any one of claims 1 to 8, characterized in that it comprises the following steps:

[0072] Step S1: Collect the professional background data and historical learning records of the target doctor;

[0073] Step S2: Generate a medical knowledge network associated with its professional field through the knowledge graph construction module;

[0074] Step S3: Formulate phased learning goals and tasks based on the personalized learning engine;

[0075] Step S4: Complete case practice in the simulated diagnosis and treatment module and optimize the learning path based on the feedback results.

[0076] The method according to claim 9, wherein the learning task in step S3 comprises:

[0077] (1): Mandatory learning content, required knowledge updated according to the latest guidelines or regulations in the medical industry; elective learning content: customized learning modules based on the doctor's interests and career plans.

[0078] (2): The method according to claim 8 is characterized in that the platform uses blockchain technology to store the doctor's learning records, simulated diagnosis and treatment results and ability assessment results, and generates an unalterable electronic ability certificate.

[0079] like Figure 4 An application method for building a platform based on artificial intelligence medical knowledge in an embodiment of the present invention includes:

[0080] Step S1: Collect the professional background data and historical learning records of the target doctor;

[0081] Step S2: Generate a medical knowledge network associated with its professional field through the knowledge graph construction module;

[0082] Step S3: Formulate phased learning goals and tasks based on the personalized learning engine;

[0083] Step S4: Complete case practice in the simulated diagnosis and treatment module and optimize the learning path based on the feedback results.

[0084] In addition, it should be noted that although the present invention is disclosed above, the scope of protection of the invention is not limited thereto. Those skilled in the art may make various changes and modifications without departing from the spirit and scope of the present invention, and these changes and modifications all fall within the scope of protection of the present invention.

Claims

1. A platform for developing and building medical knowledge based on artificial intelligence, characterized by: include: Step 1, medical data preprocessing module; Medical data capture obtains structured information data from multi-source heterogeneous medical data, and targets the data source and credibility level for verification and identification. The first verification and identification data are verified, and the second verification and identification data are screened and cleaned. The process is repeated three times to obtain accurate information data for medical knowledge data digital storage collection. Medical knowledge data digital storage collection includes: medical literature, clinical diagnosis and treatment records, medical imaging and genomic data. The first verification and identification data are verified, and the second verification and identification data are screened and cleaned. The process is repeated three times to obtain accurate information data for medical knowledge data digital storage; medical knowledge data digital storage undergoes data cleaning, de-identification and standardization to generate high-quality training data integration; Step 2, medical knowledge graph construction module; based on the integrated natural language processing technology, the pre-processed medical data information is analyzed, and according to the method of step 1 in claim 1, the data information digital parameters are extracted to generate a medical knowledge graph construction module data information integration training system, and the medical knowledge graph construction module data information integration training system of the claim generates a dynamically updated medical knowledge graph, and the knowledge graph contains multi-dimensional associations of diseases, symptoms, drugs, treatment plans and medical research progress; Step 3: The artificial intelligence model training module includes: integrating a deep learning algorithm, a reinforcement learning algorithm, a poisoning training framework, a federated learning framework, a self-developed ant combat training module, and a self-developed law of the jungle algorithm module; pre-training data information is integrated through the medical knowledge graph and the medical data pre-processing module; the pre-training data information integration training obtains data information numbers for multi-model training; the multi-model training generates an explainable medical diagnosis model and a personalized treatment recommendation model; Step 4, the medical knowledge data information is digitally stored and dynamically updated. The data information of the first dynamic update is identified by multi-model training for pre-training. The data information of the second dynamic update is identified by the medical knowledge graph construction module for artificial intelligence multi-model training and verification module. The verification module is a training and learning module obtained by digitally verifying, identifying and calculating the information data of the medical data preprocessing module and the medical knowledge graph construction module according to steps 1 and 2 of the claim. The training and learning module monitors the medical database update and user feedback data information in real time, identifies and calculates the active learning mechanism to optimize the digital parameters of the knowledge graph and model data information, and evaluates the model performance based on the digital integration of clinical verification data information; Step 5: Security compliance module. The security compliance module has a built-in HIPAA and GDPR-compliant multi-model training system. The multi-model training system identifies user ID information data and performs multi-model training. First dimension module: used to receive access requests for user ID information data. The built-in HIPAA and GDPR-compliant multi-model training system obtains the access request and implements data anonymization and encrypted transmission identification verification through the identification privacy enhancement technology (PETs) system. The second dimension module: through the privacy enhancement technology (PETs) system, is used to receive the user ID described by the first dimension module for requesting access information data verification and identification, and for second verification identification of ID information data through the requesting access information data verification and identification, and for screening target marking by calculating the third identification verification ID data information through the first and second verification information data identification, and the screening target marking marks the user ID tracking information data for requesting access. The third dimension module: ID access request information enters the identification verification information data digital claim step 1 to step 4 information data including: medical data preprocessing module, medical knowledge data information digital reserve collection, medical knowledge data information digital storage, generation of high-quality training data integration, medical knowledge graph construction module, artificial intelligence model training module, data information digital integration for model performance evaluation, and security compliance module.

2. The platform according to claim 1, characterized in that The personalized learning engine module includes: Generate personalized learning paths based on the doctor's professional field, skill level and historical learning data, and dynamically adjust the learning content and difficulty through reinforcement learning algorithms.

3. The platform according to claim 1, characterized in that The intelligent recommendation module: based on the knowledge graph and learning engine, recommends matching medical cases, the latest research results and simulated diagnosis and treatment tasks to doctors, jointly optimizes disease prediction, treatment plan recommendation and prognosis analysis tasks through a multi-task learning framework, provides a model interpretability interface, and outputs key medical features and literature supporting evidence for diagnosis.

4. The platform according to claim 1, characterized in that The simulated diagnosis and treatment and feedback module provides a virtual patient interaction interface, generates simulated cases in combination with real clinical data, and provides real-time evaluation and feedback on doctors' diagnosis and treatment decisions.

5. The platform according to claim 1, characterized in that The personalized learning engine module includes: 5.1 Skills Assessment Unit: quantifies the doctor's professional competence level by analyzing the doctor's decision-making accuracy, response time, and logical integrity in simulated diagnosis and treatment; 5.2 Adaptive learning unit, dynamically adjusts the depth, breadth and presentation of learning content based on evaluation results.

6. The platform according to claim 1, characterized in that The intelligent recommendation module further includes: a multi-objective optimization algorithm for balancing the timeliness, clinical relevance and doctor interest preferences of the recommended content; a collaborative filtering unit for generating a recommendation list driven by group intelligence by combining anonymous learning data of the doctor group.

7. The platform according to claim 1, characterized in that The simulated diagnosis and treatment and feedback module includes: a virtual case generation unit based on a generative adversarial network (GAN) to simulate diverse clinical scenarios; a decision tree analysis unit to retrospectively analyze the doctor's diagnostic path and generate visual improvement suggestions.

8. The platform according to claim 1, characterized in that The platform also includes: 8.1: Privacy protection module, which uses federated learning technology to conduct distributed training on doctor learning data to ensure that the original data does not leave the local terminal. The data desensitization unit encrypts and anonymizes sensitive information in case data. 8.2: The platform according to any one of claims 1-7 is characterized in that the platform is further integrated with medical equipment, including: an interface unit for medical imaging equipment, used to obtain imaging data in real time and associate it with diagnostic criteria in the knowledge graph; and a surgical robot collaboration unit, which trains the robot-assisted decision-making model by simulating diagnosis and treatment data.

9. An application method based on the platform according to any one of claims 1 to 8, characterized in that: The following steps are involved: Step S1: Collect the professional background data and historical learning records of the target doctor; Step S2: Generate a medical knowledge network associated with its professional field through the knowledge graph construction module; Step S3: Formulate phased learning goals and tasks based on the personalized learning engine; Step S4: Complete case practice in the simulated diagnosis and treatment module and optimize the learning path based on the feedback results.

10. The method according to claim 9, characterized in that The learning tasks in step S3 include: 10.1: Mandatory learning content: required knowledge updated according to the latest guidelines or regulations in the medical industry. Optional learning content: customized learning modules based on the doctor's interests and career plans. 10.2: The method according to claim 8, characterized in that The platform uses blockchain technology to store doctors' learning records, simulated diagnosis and treatment results, and competency assessment results, and generates tamper-proof electronic competency certificates.

Images