An interactive diagnosis and treatment simulation training system based on multi-source clinical pathway and artificial intelligence guidance

Abstract

The application relates to the technical field of medical teaching, and discloses an interactive diagnosis and treatment simulation training system based on a multi-source clinical path and artificial intelligence guidance, which is based on real clinical case data, carries out structured processing on symptom information, sign information, medical history information and examination and test data, constructs a virtual patient model, and uniformly models and dynamically updates a patient state, a disease progression state and a diagnosis and treatment risk state in a diagnosis and treatment simulation process; the system formalizes diagnosis and treatment operation of students into diagnosis and treatment decisions, is used for driving disease evolution, examination and test result generation and risk state change, realizes continuous simulation of a complete diagnosis and treatment process, simultaneously provides intelligent guidance through real-time analysis of a diagnosis and treatment path, and carries out risk monitoring on a medication scheme; after training is completed, the system comprehensively evaluates student performance and outputs results. The application can improve the authenticity and standardization of medical training without involving real patient risks.

Description

Technical Field

[0001] This invention relates to the field of medical teaching technology, and more specifically, to an interactive diagnosis and treatment simulation training system based on multi-source clinical pathways and artificial intelligence guidance. Background Technology

[0002] With the continuous development of medical technology, clinical diagnosis and treatment processes are becoming increasingly complex, placing higher demands on the professional competence, clinical reasoning skills, and risk control capabilities of medical personnel. Especially in standardized residency training, specialist training, and continuing medical education, how to enable trainees to systematically master the complete diagnosis and treatment process under limited teaching resources and clinical conditions remains a real challenge in the field of medical education.

[0003] Currently, clinical teaching mainly relies on the following methods: first, real-world clinical instruction, where senior physicians provide guidance during actual diagnosis and treatment; second, case-based teaching using paper cases, electronic cases, or multimedia courseware; and third, supplementary training through simulation teaching systems or scenario-based training platforms. While these methods can improve trainees' theoretical knowledge and basic operational skills to some extent, they still have significant limitations in practical application.

[0004] First, real-world clinical teaching is limited by factors such as patient safety, medical resources, and workload. Trainees can only participate in a limited number of diagnostic and treatment processes, often only observing parts of the procedure, making it difficult to repeatedly practice different decision-making paths throughout the entire course of a disease. Furthermore, real cases are not reproducible; if trainees miss key diagnostic and treatment points, it is difficult to correct and solidify their understanding through subsequent practice.

[0005] Secondly, existing case-based and multimedia teaching methods are mostly static presentations. Case content usually unfolds according to a predetermined diagnostic and treatment pathway, and trainees can only passively learn existing conclusions, lacking opportunities to observe changes in the condition under different treatment options. This type of teaching method fails to reflect the inherent uncertainties in clinical diagnosis and treatment, and cannot effectively train trainees' clinical reasoning and risk assessment abilities.

[0006] Furthermore, while some simulation teaching systems incorporate virtual patients or scenario-based exercises, they often focus on single-skill training or fixed-scene simulations, such as single-operation training or standard procedure demonstrations. These systems typically operate using pre-set scripts, resulting in limited correlation between changes in patient status and trainee decisions. Consequently, they fail to accurately reflect the combined impact of different diagnostic and treatment decisions on disease progression, examination results, and the risk of complications.

[0007] In existing technologies, the functional modules of diagnostic and treatment simulation systems remain relatively independent. For example, case display, diagnostic and treatment operations, risk warnings, and training assessments are often implemented by different modules, lacking a unified data model and state description method. This makes it difficult for the system to continuously and dynamically simulate the diagnostic and treatment process. Furthermore, existing systems often rely on rule-based prompts or simple question-and-answer methods for intelligent guidance, making it difficult to provide targeted guidance based on specific patient conditions and the trainee's treatment path.

[0008] Furthermore, with the continuous updating of clinical guidelines and medical literature, diagnostic and treatment standards are highly time-sensitive. In existing teaching systems, the problem of outdated medical knowledge and diagnostic and treatment rules is quite common, leading to discrepancies between training content and actual clinical practice, which hinders trainees from developing diagnostic and treatment thinking that conforms to current standards.

[0009] Therefore, there is an urgent need for an interactive diagnosis and treatment simulation training system based on multi-source clinical pathways and artificial intelligence guidance to solve these problems. Summary of the Invention

[0010] The purpose of this invention is to solve the technical problems mentioned in the background section and to provide an interactive diagnosis and treatment simulation training system based on multi-source clinical pathways and artificial intelligence guidance.

[0011] The objective of this invention is achieved through the following technical solution: An interactive diagnosis and treatment simulation training system based on multi-source clinical pathways and artificial intelligence guidance, the system comprising: The medical record collection and summarization module is used to collect and summarize real patient medical records and save the original data. The intelligent case engine module is used to integrate and structure real patient case data, generate potential etiology maps based on symptoms, signs and medical history, simulate the disease development process on the timeline to extract key medical behavior nodes, and integrate imaging data, test data and genetic data to generate trainable personalized virtual patients. The full-course diagnosis and treatment simulation module is used to drive the virtual patient to exhibit changes in static and dynamic attributes during the training process, and to execute the medical process simulation in the order of outpatient reception, admission examination order, result interpretation, treatment plan determination, doctor-patient communication, case recording, discharge follow-up and remote follow-up. In the medical process simulation, the corresponding examination and test data are matched and updated according to the treatment plan selected by the trainee. The intelligent guidance module is used to provide natural language question-and-answer guidance during the interactive training process of trainees, and to detect and warn of drug interaction risks for trainees' medication prescriptions throughout the entire process; The multi-level knowledge and rules engine module is used to synchronize the latest treatment guidelines, drug instructions and clinical pathway SOPs and store them in a structured database. It automatically associates the current disease with the corresponding research evidence and literature content, and also includes typical cases and expert decision-making logic to support case matching and decision support. The feedback and evaluation module is used to score the trainees' training process from the dimensions of treatment effectiveness, treatment efficiency, economy and safety, and output quantitative evaluation results and personalized improvement suggestions. The community module is used to provide simulation scenario sharing and discussion, and supports online comments from experts on the simulation content.

[0012] As a preferred technical solution of the present invention, the intelligent case engine module includes: a pathogenesis reasoning unit, used to extract and integrate patient symptoms, signs and medical history from diagnosis and treatment process data, and automatically generate a potential pathogenesis map; The disease progression simulation unit is used to simulate the occurrence, development and outcome of a disease over time, and output the disease status and key medical behavior nodes that change over time. The multimodal data fusion unit is used to uniformly structure and represent imaging, laboratory, and genetic data, and thereby generate virtual patient data packages with individual differences.

[0013] As a preferred technical solution of the present invention, the full-course diagnosis and treatment simulation module includes at least: The virtual patient simulation unit is used to define the static and dynamic attributes of the virtual patient; the static attributes include age, gender and genetic background, while the dynamic attributes include symptoms, signs and examination results. The medical process simulation unit is used to simulate the entire process of outpatient reception, admission examination ordering, result interpretation, treatment plan determination, doctor-patient communication, case recording, discharge follow-up and remote follow-up. After the trainee determines the treatment plan, the unit triggers the selection of examination and test items and the update of results according to the treatment plan, so as to realize the plan matching of diagnosis and treatment process data.

[0014] As a preferred technical solution of the present invention, the full-course diagnosis and treatment simulation module further includes: a time dimension advancement unit, which provides a simulated clock and supports time fast forward or rewind, and manages the disease events triggered at the corresponding time points as time changes, including disease deterioration events and examination appointment events, so as to support intervention training for key events.

[0015] As a preferred technical solution of the present invention, the full-course diagnosis and treatment simulation module further includes: a sudden event random unit, used to introduce sudden medical events according to preset probabilities or rules during the training process. Sudden medical events include complication events, drug allergy events and transfer events, thereby generating personalized virtual cases covering routine cases and difficult and severe cases.

[0016] As a preferred technical solution of the present invention, the intelligent guidance module includes: a context-aware prompting unit, which is used to identify the treatment path based on the student's operational behavior in each stage of diagnosis and treatment, compare the student's treatment path with the standard treatment path to identify deviations and issue deviation reminders, and push corresponding medical knowledge prompts in conjunction with the operational behavior to consolidate training; The AI ​​tutor Q&A unit supports natural language interactive Q&A, answers students' questions in the diagnosis, examination, medication and treatment stages, and simulates the role of a senior physician to provide guidance on key decisions.

[0017] As a preferred technical solution of the present invention, the intelligent guidance module further includes: a drug interaction early warning unit, used to monitor the patient's medication prescriptions throughout the entire course of the disease in real time, identify drug interaction risks and output early warning information, wherein the early warning information includes at least the drug pair of the source of the risk, the type of risk and the suggested alternative.

[0018] As a preferred technical solution of the present invention, the multi-level knowledge and rule engine module includes: The guideline dynamic update unit is used to automatically capture and synchronize the latest treatment guidelines, drug instructions and clinical pathway SOPs, complete the structured storage, and update the standard treatment pathways and recommended decision-making schemes in the system in real time based on the updated content. The Expert Experience Unit is used to collect typical cases and expert decision-making logic, establish a searchable experience entry library, and support case matching. The real-time literature retrieval unit is used to automatically link relevant research evidence for the current disease, assisting trainees in literature retrieval and interpretation.

[0019] As a preferred technical solution of the present invention, the feedback evaluation module includes: The multi-dimensional scoring unit is used to compare trainees' training results with standard path maps and excellent hospital path maps, outputting differences and forming treatment effect scores, treatment efficiency scores, cost-effectiveness scores, and safety scores. The quantitative survey unit is used to quantitatively evaluate the effectiveness, economy, security and integrity of the training process and generate expression packages; The personalized improvement suggestion unit is used to record the learner's historical performance and generate a skill growth curve, recommending learning content and training cases based on weaknesses.

[0020] As a preferred technical solution of the present invention, the community module includes: a scene sharing and discussion unit, which is used to publish the simulated scene and support multi-user discussion, and intelligently organize the collected content and liked content and return them to the simulation module to form a reusable training scene. The online expert review section allows experts to provide online feedback on the simulation content and the participants' process, and offer suggestions for improvement.

[0021] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention constructs a unified joint state model to centrally describe the virtual patient state, disease progression state, and treatment risk state, and continuously updates this model during the treatment simulation process. This enables the system to drive multiple functional modules, such as case construction, disease progression, treatment simulation, intelligent guidance, and training evaluation, with a unified data foundation. This approach avoids the problems of data fragmentation and inconsistent states between modules in existing treatment simulation systems, allowing the treatment process to proceed continuously according to real clinical logic, thus improving the overall consistency and feasibility of the treatment simulation.

[0022] 2. This invention constructs a virtual patient model based on real clinical cases and introduces etiological reasoning, uncertainty modeling, and a dynamic disease progression update mechanism into the diagnosis and treatment simulation process. This allows the virtual patient's status to change in real time according to the trainees' diagnosis and treatment decisions. Different diagnosis and treatment decisions will directly affect the direction of disease development, examination and test results, and the occurrence of risk events. This allows trainees to intuitively feel the impact of their own decisions on the diagnosis and treatment outcome, which helps to train trainees' clinical thinking ability, risk assessment ability, and comprehensive decision-making ability, and improves the authenticity and relevance of teaching and training.

[0023] 3. This invention introduces intelligent guidance and risk warning mechanisms into the diagnosis and treatment simulation process. Through continuous analysis of the trainee's treatment path, combined with the current patient status and disease progression, it provides trainees with knowledge prompts and decision-making suggestions that match the specific treatment stage. Simultaneously, through continuous monitoring of medication regimens, it provides real-time alerts on drug interaction risks, helping to strengthen trainees' understanding of treatment guidelines and medication safety, and reducing the risk of developing incorrect treatment habits during training.

[0024] 4. This invention, by establishing training evaluation and community feedback mechanisms, provides multi-dimensional evaluation of the trainees' training process. It not only focuses on the final treatment outcome but also comprehensively considers the standardization of treatment pathways, risk control capabilities, and resource utilization, thereby generating more objective and comprehensive training evaluation results. The expert commentary and typical training scenario feedback mechanism enables the system to continuously accumulate high-quality training cases, enhancing the reference value for subsequent training and the overall scalability of the system.

[0025] In summary, this invention can dynamically simulate and systematically train the entire diagnosis and treatment process without increasing real clinical risks, effectively improving the practicality, continuity, and standardization of medical training, and has good application prospects and promotional value. Attached Figure Description

[0026] Figure 1 This is a system block diagram of the present invention; Figure 2 This is a diagram showing the composition of the disease course diagnosis and treatment simulation module of the present invention; Figure 3 This is a diagram showing the composition of the feedback evaluation module of the present invention; Figure 4 This is a diagram showing the composition of the community module of the present invention. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this invention clearer, the following description is provided in conjunction with embodiments and appendices. Figures 1-4 The present invention will be further described in detail below. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0028] Example 1: This example provides a specific application process of an interactive diagnosis and treatment simulation training system based on multi-source clinical pathways and artificial intelligence guidance, which is used to illustrate the implementation of the technical solution of the present invention in a real medical training scenario.

[0029] In this embodiment, the system is deployed on the server side of a medical teaching platform and provides trainees with an interactive diagnostic and treatment simulation training environment through terminal devices. The system first collects and pre-loads real-source clinical case data through a medical record collection and aggregation module, and saves the raw data. This case data includes the patient's basic information, chief complaint, present illness, past medical history, physical examination records, imaging results, laboratory test indicators, and available genetic testing information. After entering the system, the above case data is processed uniformly by the intelligent case engine module.

[0030] Specifically, the intelligent case engine module standardizes and parses case data, converting unstructured text information into structured data and classifying and organizing symptom, sign, and medical history information. Simultaneously, the system uniformly encodes imaging, laboratory, and genetic data, enabling different types of data to be accessed within the same virtual patient model. Based on these processing results, the system constructs the initial state of the virtual patient, describing the patient's overall condition at the beginning of the treatment phase.

[0031] After constructing the initial state of the virtual patient, the system further infers the causes of possible disease types. Instead of directly providing a single, definitive diagnosis, the system generates multiple potential causes and their corresponding probabilities based on the virtual patient's state information, thus creating an uncertain initial state of disease within the system. This setup allows different trainees, even when faced with the same initial case, to guide the disease progression in different directions due to different decisions during training.

[0032] Subsequently, the full-course diagnosis and treatment simulation module was activated, and the virtual patient officially entered the simulation process. The system configures static and dynamic attributes for the virtual patient. Static attributes include age, gender, and genetic background, used to define the patient's basic characteristics; dynamic attributes include symptom presentation, changes in physical signs, and examination and test results, used to reflect the patient's real-time status during the diagnosis and treatment process. The dynamic attributes are not fixed in advance but are continuously updated as the diagnosis and treatment process progresses.

[0033] During the training, trainees log into the system as clinicians and perform diagnostic and treatment operations based on the information displayed to the virtual patient. The medical process simulation unit organizes the training steps according to the real clinical process, including outpatient reception, ordering of examinations and tests, interpretation of test results, diagnosis, treatment plan formulation, doctor-patient communication, case recording, discharge assessment, and follow-up management. The actions taken by the trainees at each stage are recorded by the system and used as an important basis for subsequent simulation calculations.

[0034] After a trainee selects a treatment plan, the system does not return preset, fixed test results. Instead, it dynamically matches and updates subsequent test results based on the current disease state and the chosen treatment plan. For example, under a reasonable treatment plan, the system will simulate symptom relief and improvement in test indicators; while under unreasonable or delayed treatment, it may simulate disease progression or abnormal test results. In this way, the system allows trainees to intuitively understand the impact of their decisions on the patient's condition.

[0035] During the simulated diagnosis and treatment process, the system manages the disease progression through time-based progression units. The system internally sets up a simulated timeline, with different diagnostic and treatment operations producing corresponding effects at different time points. For example, some treatments require a certain amount of time to show their effects, and some examinations also have time delays in scheduling and result return. Through time-based control, the system can train trainees to make diagnostic and treatment decisions at appropriate times, rather than completing all operations at once.

[0036] To further simulate a real clinical environment, the system incorporates an emergency event mechanism during the diagnosis and treatment simulation. The emergency event random unit, based on the current disease state and treatment progress, may trigger complications, adverse drug reactions, or situations requiring transfer to another department. When an emergency occurs, the system updates the virtual patient's status accordingly and requires trainees to reassess the treatment plan, thereby increasing the training difficulty and comprehensive handling capabilities.

[0037] Throughout the training process, the intelligent guidance module continuously analyzes the trainee's actions. The system compares the trainee's actual treatment path with the system's built-in standard treatment path. When significant deviations are detected, the system provides targeted medical knowledge prompts or decision reminders to the trainee through context-aware prompts. In addition, trainees can ask questions to the system in natural language during training. The AI ​​tutor question-and-answer unit will combine the current condition and medical knowledge base to provide explanatory answers to the trainee's questions, simulating the guidance process from a senior physician to a junior physician.

[0038] In medication-related procedures, the drug interaction early warning unit continuously monitors the medication regimens developed by trainees throughout the entire treatment process. When the system identifies potential drug interaction risks or medication safety hazards, it will promptly issue early warnings to trainees, reminding them to adjust their medication regimens or further verify the evidence, thereby strengthening their awareness of medication safety.

[0039] A multi-level knowledge and rules engine module provides knowledge support to each module during system operation. This module, through a dynamic guideline update unit, regularly synchronizes the latest treatment guidelines, drug instructions, and clinical pathway standards, and updates the standard treatment pathways within the system. Simultaneously, the system can automatically link relevant medical research literature based on the current case, allowing trainees to consult and reference it. Furthermore, the system stores expert experience cases to provide trainees with comparative references during training.

[0040] After training, the feedback and evaluation module comprehensively assesses the trainees' overall performance. The evaluation includes not only the appropriateness of the final treatment outcome, but also the standardization of the treatment pathway, the adequacy of risk control, and the rationality of medical resource utilization. Based on the evaluation results, the system generates a quantitative score and provides personalized improvement suggestions in conjunction with the trainees' historical training records, helping them identify their weaknesses.

[0041] Furthermore, the system supports the sharing and exchange of training content through a community module. Trainees can post their simulated scenarios to the community for discussion, and experts can provide online feedback on the trainees' diagnostic and treatment processes. The system will compile representative training scenarios from the community and feed them back into the training system as reusable training cases for subsequent trainees, thereby achieving continuous optimization of the training content.

[0042] It also includes a medical record collection and summary module to save the original data.

[0043] As can be seen from the above embodiments, the present invention organically combines virtual patient modeling, disease progression simulation, diagnosis and treatment process simulation, intelligent guidance, and training evaluation to construct a complete, implementable, and scalable interactive diagnosis and treatment simulation training system, which can effectively improve the authenticity and systematic nature of medical training.

[0044] Example 2: In this example, an interactive diagnosis and treatment simulation training system based on multi-source clinical pathways and artificial intelligence guidance is provided. This system is used in medical teaching and clinical skills training scenarios to virtually reconstruct and interactively train real-world diagnosis and treatment processes. By uniformly modeling the virtual patient's state, disease progression state, and diagnosis and treatment risk state, the system ensures that every diagnosis and treatment decision made by the trainee during training directly impacts subsequent disease progression, examination and test results, and training evaluation.

[0045] During system operation, to ensure data consistency and coordinated operation among various functional modules, the system at any simulation moment... Internally, a joint state variable is maintained to describe the overall state of the virtual patient in the current training scenario. This joint state variable is denoted as... , where subscript This indicates the current simulation time point. Joint state variables. From the patient's condition Disease progression status and the status of diagnosis and treatment risks Together they constitute the patient's condition. Used to describe the symptoms, signs, and test results of a virtual patient at the current moment; disease progression status. Used to describe the stage of a disease and its development trend; status of diagnostic and treatment risks. Used to describe the risk of complications, medication risks, and medical resource consumption related to the diagnosis and treatment process, the joint state variable serves as the state carrier within the system and is continuously read and updated during subsequent diagnosis and treatment simulations, intelligent guidance, and training evaluations.

[0046] In the initial training phase, the system processes real clinical case data through an intelligent case engine module. This real clinical case data includes patient symptom information, physical signs, medical history, imaging results, laboratory test data, and genetic testing information. The system first standardizes and structures this data, representing symptom information as a set of symptom features. Representing vital signs information as a set of vital sign features Representing medical history information as a set of medical history features .in, , and All of these are feature sets that have undergone unified encoding and quantization.

[0047] The system constructs the initial patient state of the virtual patient based on the aforementioned feature set, denoted as... The subscript "O" indicates the start time of training. The initial patient state is generated through a patient state mapping function, and its calculation relationship is expressed as follows: ; in, This represents a patient state mapping function, used to transform discrete clinical feature information into a patient state description that can be processed internally by the system.

[0048] In obtaining the initial patient status Then, the system infers potential causes. The system pre-defines a set of candidate causes. ,in Indicates the first One candidate etiology, This indicates the number of candidate causes. The system is based on the initial patient state. Calculate the probability value corresponding to each candidate etiology, and denote the probability value as follows: , used to indicate that the virtual patient belongs to the cause of the disease The probability of is calculated as follows: ; The system does not directly select the cause with the highest probability as the definitive diagnosis. Instead, it generates the initial progression state of the disease based on the set of probabilities of the stated causes, denoted as... The initial disease progression state is determined by disease state templates corresponding to each etiology. It is obtained by probability weighting, and its calculation relationship is as follows: ; The system does not directly select the cause with the highest probability as the definitive diagnosis. Instead, it generates the initial disease progression state based on the set of probabilities of the stated causes, denoted as... The initial disease progression state is determined by disease state templates corresponding to each etiology. It is obtained by probability weighting, and its calculation relationship is as follows: ; In this way, the system introduces clinical diagnostic uncertainty at the beginning of the training phase, allowing for multiple possible development paths in the subsequent disease progression.

[0049] After completing the initial patient status and disease progression status After its construction, the system forms the initial joint state for training. And enter the full course of diagnosis and treatment simulation process.

[0050] During the clinical simulation, trainees are based on the current collaborative state. Perform diagnostic and treatment procedures. The system will record the trainee's progress in real time. The diagnostic and treatment procedures performed are uniformly represented as diagnostic and treatment decision variables. The diagnostic and treatment decision variables include at least diagnostic judgment operations, examination and test order operations, treatment plan selection operations, and doctor-patient communication operations, which are formally represented as follows: ; in, This represents the diagnostic decision mapping function, which is used to associate student actions with the current patient status and disease status.

[0051] Upon receiving diagnostic and treatment decision variables Then, the disease progression simulation unit simulates the current disease progression status. The aforementioned diagnostic and treatment decision variables are used to update the disease progression status and generate the disease progression status for the next time step. Its update relationship is as follows: ; in, This represents the disease evolution function, used to describe the impact of different diagnostic and treatment decisions on the direction and speed of disease progression.

[0052] Subsequently, the system updates the disease progression status. and treatment decision variables The corresponding inspection and testing results are generated, and the inspection and testing results are denoted as... Its generation relationship is: ; in, This represents a function that generates test results, reflecting the impact of treatment plans and changes in the patient's condition on these test indicators. The system writes the test results back to the patient's status, updates the patient's status, and generates a new patient status. Its update relationship is: ; During the diagnostic simulation, the system is based on the updated disease progression status. Determine whether a medical emergency has been triggered, and the medical emergency is denoted as... These include complication events, adverse drug reaction events, or transfer events, and their triggering relationships are as follows: ; in, This represents a function that triggers sudden medical events, used to determine whether complications, adverse drug reactions, or transfer events have occurred based on the disease progression status.

[0053] When a sudden medical event occurs, the system combines diagnostic and treatment decision variables. In addition, the results of emergencies are used to update the diagnosis and treatment risk status and generate new risk statuses. Its update relationship is: ; in, This represents the risk status update function, used to update the treatment risk status based on comprehensive medical decision-making and unexpected medical events. After updating the patient status, disease status, and risk status, the system forms a new joint state. This information is then used as input for the next round of diagnosis and treatment simulation, intelligent guidance, and training evaluation, thereby enabling multiple rounds of continuous diagnosis and treatment training.

[0054] Throughout the training process, the intelligent guidance module continuously analyzes the trainee's treatment pathways. The system extracts the trainee's actual treatment pathway based on the joint state sequence and compares it with the system's built-in standard treatment pathways. When a deviation is detected, the system provides the trainee with medical knowledge prompts or decision suggestions matching the current stage of treatment through context-aware prompts. Simultaneously, the system supports responding to trainee questions in natural language, simulating the guidance process of a senior physician on key treatment decisions.

[0055] In medication-related operations, the system uses diagnostic and treatment decision variables. The system extracts the medication regimens developed by trainees and continuously monitors these regimens. Based on a drug interaction knowledge base, the drug interaction early warning unit identifies potential risks in the medication regimens and issues warnings to trainees when risks are detected.

[0056] After training, the feedback and evaluation module comprehensively assesses trainees' performance based on the joint state evolution process from the start to the end of training. The evaluation includes the rationality of treatment outcomes, the standardization of treatment pathways, risk control capabilities, and resource utilization, generating quantitative scores and personalized improvement suggestions accordingly.

[0057] Furthermore, the system supports the sharing and exchange of training scenarios through a community module, allowing experts to provide online feedback on trainees' diagnosis and treatment processes. Representative training scenarios and expert comments are then compiled and fed back into the training system as reusable cases for subsequent training.

[0058] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An interactive diagnosis and treatment simulation training system based on multi-source clinical pathways and artificial intelligence guidance, characterized in that, The system includes: The medical record collection and summarization module is used to collect and summarize real patient medical records and save the original data. The intelligent case engine module is used to integrate and structure real patient case data, generate potential etiology maps based on symptoms, signs and medical history, simulate the disease development process on the timeline to extract key medical behavior nodes, and integrate imaging data, test data and genetic data to generate trainable personalized virtual patients. The full-course diagnosis and treatment simulation module is used to drive the virtual patient to present static and dynamic attribute changes during the training process, and to execute the medical process simulation in the order of outpatient reception, admission examination order, result interpretation, treatment plan determination, doctor-patient communication, case recording, discharge follow-up and remote follow-up. In the medical process simulation, the corresponding examination and test data are matched and updated according to the treatment plan selected by the trainee. The intelligent guidance module is used to provide natural language question-and-answer guidance during the interactive training process of trainees, and to detect and warn of drug interaction risks for trainees' medication prescriptions throughout the entire process; The multi-level knowledge and rules engine module is used to synchronize the latest treatment guidelines, drug instructions and clinical pathway SOPs and store them in a structured database. It automatically associates the current disease with the corresponding research evidence and literature content, and also includes typical cases and expert decision-making logic to support case matching and decision support. The feedback and evaluation module is used to score the trainees' training process from the dimensions of treatment effectiveness, treatment efficiency, economy and safety, and output quantitative evaluation results and personalized improvement suggestions. The community module is used to provide simulation scenario sharing and discussion, and supports online comments from experts on the simulation content.

2. The interactive diagnosis and treatment simulation training system based on multi-source clinical pathways and artificial intelligence guidance according to claim 1, characterized in that, The intelligent medical record engine module includes: The etiology reasoning unit is used to extract and integrate patient symptoms, signs and medical history from the diagnosis and treatment process data, and automatically generate potential etiology maps; The disease progression simulation unit is used to simulate the occurrence, development and outcome of a disease over time, and output the disease status and key medical behavior nodes that change over time. The multimodal data fusion unit is used to uniformly structure and represent imaging, laboratory, and genetic data, and thereby generate virtual patient data packages with individual differences. 3.The multi-source clinical pathway and artificial intelligence guided interactive diagnosis and treatment simulation training system according to claim 1, wherein, The full-course diagnosis and treatment simulation module includes at least: a virtual patient simulation unit, used to define the static and dynamic attributes of the virtual patient; wherein the static attributes include age, gender and genetic background, and the dynamic attributes include symptoms, signs and examination results. The medical process simulation unit is used to simulate the entire process of outpatient reception, admission examination ordering, result interpretation, treatment plan determination, doctor-patient communication, case recording, discharge follow-up and remote follow-up. After the trainee determines the treatment plan, the unit triggers the selection of examination and test items and the update of results according to the treatment plan, so as to realize the plan matching of diagnosis and treatment process data. 4.The multi-source clinical pathway and artificial intelligence guided interactive diagnosis and treatment simulation training system according to claim 3, wherein, The full-course diagnosis and treatment simulation module also includes a time dimension advancement unit, which provides a simulated clock and supports time fast forward or rewind, and manages the disease events triggered at the corresponding time points as time changes, including disease deterioration events and examination appointment events, to support intervention training for key events. 5.The multi-source clinical pathway and artificial intelligence guided interactive diagnosis and treatment simulation training system according to claim 4, wherein, The full-course diagnosis and treatment simulation module also includes: The randomized unit for sudden events is used to introduce sudden medical events according to preset probabilities or rules during the training process. Sudden medical events include complication events, drug allergy events, and transfer events, thereby generating personalized virtual cases covering routine cases and difficult and critical illness scenarios. 6.The multi-source clinical pathway and artificial intelligence guided interactive diagnosis and treatment simulation training system according to claim 1, wherein, The intelligent guidance module includes: a context-aware prompting unit, which is used to identify the treatment path based on the trainee's operational behavior in each stage of diagnosis and treatment, compare the trainee's treatment path with the standard treatment path to identify deviations and issue deviation reminders, and at the same time push corresponding medical knowledge prompts in combination with the operational behavior to consolidate training. The AI ​​tutor Q&A unit supports natural language interactive Q&A, answers students' questions in the diagnosis, examination, medication and treatment stages, and simulates the role of a senior physician to provide guidance on key decisions. 7.The multi-source clinical pathway and artificial intelligence guided interactive diagnosis and treatment simulation training system according to claim 6, wherein, The intelligent guidance module also includes a drug interaction early warning unit, which is used to monitor the student's medication prescriptions throughout the entire course of the disease in real time, identify drug interaction risks and output early warning information. The early warning information includes at least the drug pair from which the risk originates, the type of risk and recommended alternatives. 8.The multi-source clinical pathway and artificial intelligence guided interactive diagnosis and treatment simulation training system according to claim 1, wherein, The multi-level knowledge and rule engine module includes: The guideline dynamic update unit is used to automatically capture and synchronize the latest treatment guidelines, drug instructions and clinical pathway SOPs, complete the structured storage, and update the standard treatment pathways and recommended decision-making schemes in the system in real time based on the updated content. The Expert Experience Unit is used to collect typical cases and expert decision-making logic, establish a searchable experience entry library, and support case matching. The real-time literature retrieval unit is used to automatically link relevant research evidence for the current disease, assisting trainees in literature retrieval and interpretation. 9.The multi-source clinical pathway and artificial intelligence guided interactive diagnosis and treatment simulation training system according to claim 1, wherein, The feedback evaluation module includes: The multi-dimensional scoring unit is used to compare trainees' training results with standard path maps and excellent hospital path maps, outputting differences and forming treatment effect scores, treatment efficiency scores, cost-effectiveness scores, and safety scores. The quantitative survey unit is used to quantitatively evaluate the effectiveness, economy, security and integrity of the training process and generate expression packages; The personalized improvement suggestion unit is used to record the learner's historical performance and generate a skill growth curve, recommending learning content and training cases based on weaknesses. 10.The multi-source clinical pathway and artificial intelligence guided interactive diagnosis and treatment simulation training system according to claim 1, wherein, The community module includes a scenario sharing and discussion unit, which is used to publish simulated scenarios and support multi-user discussions, while intelligently organizing and feeding back collected and liked content to the simulation module to form reusable training scenarios. The online expert review section allows experts to provide online feedback on the simulation content and the participants' process, and offer suggestions for improvement.

Images