Virtual-real fusion stoma nursing training system based on AI driving
Through multimodal data acquisition and AI-driven virtual-reality fusion technology, a three-dimensional virtual model and interactive-evaluation mechanism are constructed, which solves the problems of insufficient physiological dynamic simulation and personalized response in traditional ostomy nursing training, and improves teaching efficiency and learning effects.
Patent Information
- Application Number
- CN202510943549.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-09
- Publication Date
- 2025-09-26
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Traditional stoma care training models lack virtual simulation of physiological dynamic changes. The AI-assisted system cannot generate personalized responses in real time. The interaction mode is single and lacks multimodal fusion. It cannot accurately evaluate operational standardization, resulting in low training efficiency.
A three-dimensional virtual model is constructed through multimodal data acquisition equipment, and real-time evaluation is carried out in combination with AI standardized patient interaction equipment. An interaction-evaluation mechanism is established to generate multi-dimensional evaluation reports and automatically adjust teaching content. AR, VR and pressure sensors are used to simulate physiological processes and interaction scenarios.
It improves the efficiency of stoma care teaching, enhances learners' cognition of the relationship between operations and physiological reactions, realizes personalized interaction and accurate evaluation, and enhances the immersion of training and learning effect.
Smart Images
Figure CN120707355A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of medical education equipment technology, and in particular to an AI-driven virtual-reality fusion ostomy nursing training system. Background Art
[0002] Stoma care involves surgically creating an opening in the abdominal wall for the bowel or urinary tract, aiming to manage excretion and prevent complications. Based on functional needs, stomas are categorized into three types: colostomy, ileostomy, and urostomy. Core care includes standardized pouch changes, skin protection, and complication monitoring. The nursing process encompasses preoperative positioning guidance, daily postoperative assessment of the stoma, a standardized changing process (cleaning, measuring, cutting, and fitting), and personalized dietary and exercise guidance.
[0003] In related technologies, traditional stoma physical models only provide mechanical operation training, lack virtual simulation of the physiological dynamic changes of the stoma, and cannot display internal physiological processes such as intestinal peristalsis and fecal flow, which makes it difficult for learners to establish the correlation between operation and physiological response. In addition, existing AI-assisted systems mostly rely on preset dialogue scripts and cannot generate personalized patient responses in real time based on operations. The interaction mode is single and lacks multimodal fusion interaction such as voice, gestures, and expressions, making it difficult to simulate real-life doctor-patient communication scenarios. In addition, traditional training models lack accurate operation evaluation mechanisms and cannot use computer vision and other technologies to analyze the standardization of nursing actions in real time. Learners find it difficult to obtain timely error correction guidance. At the same time, there is a lack of data-driven personalized learning path planning, resulting in low training efficiency and an inability to improve skill shortcomings in a targeted manner, thereby reducing the efficiency of stoma nursing teaching. There is room for improvement. Summary of the Invention
[0004] In response to the shortcomings of existing technologies, this application provides an AI-driven virtual-reality fusion ostomy care training system.
[0005] In the first aspect, the present application provides an AI-driven virtual-real fusion stoma care training method, comprising the following steps: Step S1: multi-dimensional data acquisition is performed on the stoma physical model using a multimodal data acquisition device, and a three-dimensional virtual model is constructed based on the acquired data; Step S2: constructing a virtual simulation scene, which includes real-time simulation of the physiological state of the stoma, dynamic generation of virtual cases according to teaching needs, dynamic control of excrement, and linkage of internal organs; Step S3: Collecting interaction data through the AI standardized patient interaction device, and performing evaluation based on the interaction data, thereby forming an interaction-evaluation mechanism; Step S4: construct a multidimensional evaluation indicator system based on the interaction-evaluation mechanism, generate a multidimensional evaluation report based on the multidimensional evaluation indicator system, analyze the multidimensional evaluation report through a machine learning model, and automatically adjust the teaching content based on the analysis results.
[0006] Preferably, the multimodal data acquisition device includes AR glasses, VR gloves and pressure sensors.
[0007] Preferably, the step S1 specifically includes: Step S11: Using AR glasses to perform a three-dimensional spatial scan of the stoma physical model to obtain stoma data, including stoma location, stoma size, and spatial relationship data corresponding to the tissue surrounding the stoma. In addition, infrared sensors are used to collect temperature and humidity distribution data corresponding to the skin surface. Step S12: normalizing the tactile feedback data obtained through the VR gloves, confirming the tactile feedback data based on the normalized processing results, and collecting pressure distribution data when the ostomy bag is attached in combination with the pressure sensor; Step S13: using a three-dimensional modeling algorithm to fuse the stoma data, temperature distribution data, humidity distribution data, tactile feedback data and pressure distribution data, and then construct a three-dimensional virtual model.
[0008] Preferably, constructing a virtual simulation scene specifically includes: Building a personalized case generator based on a deep learning model, inputting teaching requirement parameters based on the personalized case generator, and then generating a virtual case based on the teaching requirement parameters; The pressure sensor collects the corresponding pressure change data in the ostomy bag in real time, and inputs the pressure change data into the preset AI algorithm model to generate excrement data. The excrement data includes excrement flow, excrement characteristics and excrement abnormality data, and displays physiological parameter data in real time. When the pressure change data exceeds the preset pressure threshold, the odor simulation module releases excrement odor; The physiological parameter data is linked to the three-dimensional virtual model, and the flexible intestinal simulator built into the stoma physical model is driven to perform intestinal simulation.
[0009] Preferably, the corresponding pressure change data in the ostomy bag is collected in real time by a pressure sensor, which specifically includes: Acquire historical clinical data, extract a correlation database corresponding to fecal characteristics and pressure changes from the historical clinical data, and perform spatiotemporal distribution analysis on current pressure change data collected by the pressure sensor data to obtain pressure change rate and peak value data; The risk of stoma injury is graded according to the pressure change rate and peak data, and abnormal stool data is generated based on the graded results; The odor simulation module is used to release the corresponding concentration of excrement odor according to the abnormal excrement data.
[0010] Preferably, the step S3 specifically includes: Step S31: constructing a dialogue generation system using natural language processing technology, inputting dialogue content data corresponding to the target learner into the dialogue generation system, and then generating response data based on the dialogue content data; Step S32: Capturing the target learner's corresponding facial micro-expressions through a camera device, analyzing the learner's corresponding facial micro-expressions through a facial recognition algorithm, evaluating the degree of empathy in combination with a voice emotion analysis model, and generating an emotional interaction report; Step S33: using computer vision technology to analyze the ostomy bag replacement process frame by frame, confirming normative evaluation data based on the analysis results, and marking the error location with an AR cursor based on the normative evaluation data; Step S34: constructing an interaction-evaluation mechanism based on the response data, the emotional interaction report and the normative evaluation data.
[0011] Preferably, computer vision technology is used to analyze the ostomy bag replacement process frame by frame, and normative evaluation data is determined based on the analysis results, specifically including: Motion capture of the standard ostomy bag changing process was performed to construct a 3D motion feature vector database; Computer vision technology is used to track the real-time skeleton points of the target learner's ostomy bag replacement process and generate the feature vector corresponding to the current action; The cosine similarity calculation is performed between the feature vector corresponding to the current action and the standard feature vector corresponding to the three-dimensional action feature vector database to confirm the normative evaluation data, and the error detection mechanism is triggered based on the normative evaluation data.
[0012] In a second aspect, the present application provides an AI-driven virtual-real fusion stoma care training system, comprising: A data acquisition module is used to collect multi-dimensional data of the stoma physical model through a multimodal data acquisition device, and to construct a three-dimensional virtual model based on the collected data; A construction module for constructing a virtual simulation scenario, which includes real-time simulation of the physiological state of the stoma, dynamic generation of virtual cases according to teaching needs, dynamic control of excretion, and linkage of internal organs; An evaluation module, configured to collect interaction data through an AI standardized patient interaction device and perform evaluation based on the interaction data, thereby forming an interaction-evaluation mechanism; An adjustment module is used to construct a multidimensional evaluation indicator system based on the interaction-evaluation mechanism, generate a multidimensional evaluation report based on the multidimensional evaluation indicator system, analyze the multidimensional evaluation report through a machine learning model, and automatically adjust the teaching content based on the analysis results.
[0013] In a third aspect, the present application provides a computer-readable storage medium storing instructions. When the instructions are executed on a computer, the computer executes any one of the above-mentioned AI-driven virtual-reality fusion ostomy care training methods.
[0014] In summary, this application has the following beneficial technical effects: The present application provides an AI-driven virtual-reality fusion stoma nursing training method, which collects multi-dimensional data of the stoma physical model through multimodal data acquisition equipment, constructs a three-dimensional virtual model and a virtual simulation scene, and effectively simulates the dynamic physiological changes of the stoma, thereby reducing the situation where learners have difficulty in establishing the correlation between operations and physiological reactions due to the inability to display internal physiological processes such as intestinal peristalsis and excretion flow. The application also collects interaction data through AI standardized patient interaction equipment, and evaluates the interaction data based on the interaction data, thereby forming an interaction-evaluation mechanism, and effectively interacts. A multi-dimensional evaluation index system is constructed based on the interaction-evaluation mechanism, and a multi-dimensional evaluation report is generated according to the multi-dimensional evaluation index system. The multi-dimensional evaluation report is analyzed by a machine learning model, and the teaching content is automatically adjusted according to the analysis results, thereby effectively improving the efficiency of stoma nursing teaching. BRIEF DESCRIPTION OF THE DRAWINGS
[0015] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0016] Figure 1 This is a flow chart of the method for AI-driven virtual-reality fusion stoma care training in an embodiment of the present application.
[0017] Figure 2 This is a system diagram of AI-driven virtual-reality fusion stoma care training in an embodiment of the present application. DETAILED DESCRIPTION
[0018] The following is combined with Figure 1-2 This application is described in further detail.
[0019] Example 1 The embodiments of the present application disclose an AI-driven virtual-real fusion stoma care training method.
[0020] Reference Figure 1 The AI-driven virtual-real integration stoma care training method includes the following steps: Step S1: multi-dimensional data acquisition is performed on the stoma physical model using a multimodal data acquisition device, and a three-dimensional virtual model is constructed based on the acquired data; Step S2: constructing a virtual simulation scene, which includes real-time simulation of the physiological state of the stoma, dynamic generation of virtual cases according to teaching needs, dynamic control of excrement, and linkage of internal organs; Step S3: Collecting interaction data through the AI standardized patient interaction device, and performing evaluation based on the interaction data, thereby forming an interaction-evaluation mechanism; Step S4: construct a multidimensional evaluation indicator system based on the interaction-evaluation mechanism, generate a multidimensional evaluation report based on the multidimensional evaluation indicator system, analyze the multidimensional evaluation report through a machine learning model, and automatically adjust the teaching content based on the analysis results.
[0021] Specifically, in an AI-driven, virtual-reality fusion stoma care training method, multimodal devices such as AR glasses, VR gloves, and pressure sensors collect spatial coordinates and mucosal tactile data from the physical stoma model. This creates a three-dimensional virtual model with dynamic anatomical structures. For example, AR glasses can scan the stoma location and overlay virtual intestinal vascular structures, allowing learners to intuitively visualize the connection between the stoma and the intestine. Secondly, the virtual simulation scenario uses AI algorithms and pressure sensors to simulate changes in fecal flow. For example, when the ostomy bag is applied too forcefully, the system triggers an abnormal "bleeding" state. Simultaneously, the flexible intestinal simulator contracts, and the odor module releases a corresponding odor. Virtual cases such as urostomy-related dermatitis can also be generated based on teaching needs. Then, the AI standardized patient captures the learner's gestures through a camera, such as angle deviation when cutting the stoma bag. Errors are annotated using AR and combined with natural language processing to simulate anxious patient conversations. For example, if a misoperation occurs, the virtual patient will say, "The wound hurts a little bit." Micro-expressions are analyzed to assess empathy, forming a closed loop of "operation-feedback-error correction." Finally, an evaluation report is generated based on dimensions such as operational accuracy and communication skills. If it is found that the learner's stoma measurement error is large, the machine learning model will push targeted training modules and automatically adjust the teaching content from basic measurement to complex complication treatment. The above method not only makes the training closer to real clinical scenarios, but also improves learning efficiency through data-driven and accurately strengthens weak links.
[0022] Furthermore, the multimodal data acquisition device includes AR glasses, VR gloves and pressure sensors.
[0023] It should be noted that the step S1 specifically includes: Step S11: Using AR glasses to perform a three-dimensional spatial scan of the stoma physical model to obtain stoma data, including stoma location, stoma size, and spatial relationship data corresponding to the tissue surrounding the stoma. In addition, infrared sensors are used to collect temperature and humidity distribution data corresponding to the skin surface. Step S12: normalizing the tactile feedback data obtained through the VR gloves, confirming the tactile feedback data based on the normalized processing results, and collecting pressure distribution data when the ostomy bag is attached in combination with the pressure sensor; Step S13: using a three-dimensional modeling algorithm to fuse the stoma data, temperature distribution data, humidity distribution data, tactile feedback data and pressure distribution data, and then construct a three-dimensional virtual model.
[0024] Specifically, AR glasses use structured light scanning technology to perform millimeter-level three-dimensional spatial positioning of the stoma physical model. For example, when scanning a stoma component made of medical silicone, the position data of the stoma with a diameter of about 3 cm and located 5 cm away from the navel on the lower left side of the abdominal wall can be obtained in real time. At the same time, the infrared sensor captures the distribution characteristics of the skin temperature around the stoma at about 32°C, providing spatial and physiological basic data for the virtual model. The force-sensitive resistor array built into the VR gloves converts the touch into electrical signals, which are converted into quantitative parameters such as mucosal hardness and humidity after normalization. For example, when a learner touches the stoma mucosa, The resistance data fed back by the gloves will be standardized into a computable tactile vector, and combined with the distribution data of the edge pressure of 15N when the ostomy bag is pasted collected by the pressure sensor to form a multi-dimensional interactive parameter set. Using the Unity3D physics engine, the stoma space coordinates, temperature field distribution, tactile vector and pressure data are integrated into a three-dimensional virtual model. The three-dimensional virtual model can display the vascular texture and elastic deformation of the stoma mucosa in real time. When the learner sticks the ostomy bag with a force of 20N, the three-dimensional virtual model will simultaneously show the color of the mucosa becoming lighter after being compressed, intuitively showing the impact of the operation on the physiological structure.
[0025] The advantages of the above-mentioned multimodal modeling method are: on the one hand, through the fusion of infrared temperature and tactile data, the virtual model has physiological realism. For example, the temperature abnormality of the infected area can be highlighted in the model through color coding; on the other hand, the linkage between pressure distribution data and three-dimensional deformation can accurately simulate the mucosal ischemia scenario caused by the over-tightening of the ostomy bag in clinical practice, helping learners to establish the correlation between operating force and physiological reactions, significantly improving the immersion of training and the efficiency of knowledge conversion compared with the traditional single physical model.
[0026] It should be noted that building a virtual simulation scene specifically includes: Building a personalized case generator based on a deep learning model, inputting teaching requirement parameters based on the personalized case generator, and then generating a virtual case based on the teaching requirement parameters; The pressure sensor collects the corresponding pressure change data in the ostomy bag in real time, and inputs the pressure change data into the preset AI algorithm model to generate excrement data. The excrement data includes excrement flow, excrement characteristics and excrement abnormality data, and displays physiological parameter data in real time. When the pressure change data exceeds the preset pressure threshold, the odor simulation module releases excrement odor; The physiological parameter data is linked to the three-dimensional virtual model, and the flexible intestinal simulator built into the stoma physical model is driven to perform intestinal simulation.
[0027] Specifically, the personalized case generator is built based on the DeepSeek medical model or other AI models. When the teaching requirement parameters of "elderly patients + colostomy + stenosis complications" are input, relevant features will be extracted from the clinical case library to generate a virtual case with physical signs such as the stoma diameter reduced to 1cm and redness and swelling of the surrounding skin. At the same time, the healing difficulty is set by associating the patient's "diabetes history". The pressure sensor collects pressure changes in the ostomy bag at a frequency of 100Hz. For example, when the target learner applies uneven force to the ostomy bag, resulting in a local pressure of 25N, the AI algorithm model generates loose stool characteristics data with a flow rate of 50ml / h based on fluid mechanics formulas and historical data, and triggers abnormal bleeding parameters. At this time, the AR interface displays the dynamic curve of electrolyte levels in real time, and the odor simulation module releases volatile gases similar to fecal water. The linkage between physiological parameters and three-dimensional models is manifested as follows: the intestinal peristalsis frequency is superimposed on the virtual intestine through AR, and the flexible intestinal simulator in the physical model will synchronously produce rhythmic contractions. When the pressure exceeds the threshold of 30N, the simulator will locally bulge to simulate intestinal wall edema, forming a virtual-real mapping with the mucosal congestion mark on the AR interface.
[0028] It should be noted that the corresponding pressure change data in the ostomy bag is collected in real time through the pressure sensor, which specifically includes: Acquire historical clinical data, extract a correlation database corresponding to fecal characteristics and pressure changes from the historical clinical data, and perform spatiotemporal distribution analysis on current pressure change data collected by the pressure sensor data to obtain pressure change rate and peak value data; The risk of stoma injury is graded according to the pressure change rate and peak data, and abnormal stool data is generated based on the graded results; The odor simulation module is used to release the corresponding concentration of excrement odor according to the abnormal excrement data.
[0029] It should be noted that step S3 specifically includes: Step S31: constructing a dialogue generation system using natural language processing technology, inputting dialogue content data corresponding to the target learner into the dialogue generation system, and then generating response data based on the dialogue content data; Step S32: Capturing the target learner's corresponding facial micro-expressions through a camera device, analyzing the learner's corresponding facial micro-expressions through a facial recognition algorithm, evaluating the degree of empathy in combination with a voice emotion analysis model, and generating an emotional interaction report; Step S33: using computer vision technology to analyze the ostomy bag replacement process frame by frame, confirming normative evaluation data based on the analysis results, and marking the error location with an AR cursor based on the normative evaluation data; Step S34: constructing an interaction-evaluation mechanism based on the response data, the emotional interaction report and the normative evaluation data.
[0030] Specifically, a medical-specific GPT model is used to build a dialogue system. When a learner asks, "What should I do if the skin around the stoma becomes red?", the dermatitis treatment process will be retrieved from the clinical guidelines, and the response "It is recommended to clean it with saline first, and then apply a protective agent" will be generated. The complexity of the expression will be adjusted according to the learner's operation progress. For example, a novice trainee will receive a life-like explanation of "just like applying ointment to a wound". The camera device captures the learner's micro-expressions such as frowning and avoiding eyes at 60 frames per second, and combines the voice and tone to evaluate the empathy index through the emotional computing model. For example, when an operational error causes the virtual patient to feel "pain", if the learner does not observe the increase in heart rate on the AR interface, it will be marked as "lack of humanistic care" in the emotional report, and a motion recognition algorithm will be used to analyze the ostomy bag cutting process frame by frame. When it is detected that the cutting diameter is 5mm larger than the stoma, the AR cursor will highlight the error area in red and display the prompt "Cutting too large may cause leakage", while playing back the standard cutting gesture in slow motion.
[0031] Furthermore, computer vision technology is used to analyze the ostomy bag replacement process frame by frame, and normative evaluation data is confirmed based on the analysis results, including: Motion capture of the standard ostomy bag changing process was performed to construct a 3D motion feature vector database; Computer vision technology is used to track the real-time skeleton points of the target learner's ostomy bag replacement process and generate the feature vector corresponding to the current action; The cosine similarity calculation is performed between the feature vector corresponding to the current action and the standard feature vector corresponding to the three-dimensional action feature vector database to confirm the normative evaluation data, and the error detection mechanism is triggered based on the normative evaluation data.
[0032] Example 2 The embodiments of the present application also disclose an AI-driven virtual-reality fusion stoma care training system.
[0033] Reference Figure 2 , an AI-driven virtual-real fusion stoma care training system, including: A data acquisition module is used to collect multi-dimensional data of the stoma physical model through a multimodal data acquisition device, and to construct a three-dimensional virtual model based on the collected data; A construction module for constructing a virtual simulation scenario, which includes real-time simulation of the physiological state of the stoma, dynamic generation of virtual cases according to teaching needs, dynamic control of excretion, and linkage of internal organs; An evaluation module, configured to collect interaction data through an AI standardized patient interaction device and perform evaluation based on the interaction data, thereby forming an interaction-evaluation mechanism; An adjustment module is used to construct a multidimensional evaluation indicator system based on the interaction-evaluation mechanism, generate a multidimensional evaluation report based on the multidimensional evaluation indicator system, analyze the multidimensional evaluation report through a machine learning model, and automatically adjust the teaching content based on the analysis results.
[0034] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the described specific embodiments or replace them in similar ways. As long as they do not deviate from the concept of the invention, they should all fall within the scope of protection of the present invention.
[0035] Throughout this specification, references to terms such as "one embodiment," "example," or "specific example" indicate that the specific features, structures, materials, or characteristics described in conjunction with that embodiment or example are included in at least one embodiment or example of the present invention. In this specification, schematic representations of these terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
[0036] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the present invention to specific embodiments. Obviously, many modifications and variations are possible based on the contents of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention.
Claims
1. An AI-driven virtual-real fusion stoma care training method, characterized by: The following steps are involved: Step S1: multi-dimensional data acquisition is performed on the stoma physical model using a multimodal data acquisition device, and a three-dimensional virtual model is constructed based on the acquired data; Step S2: constructing a virtual simulation scene, which includes real-time simulation of the physiological state of the stoma, dynamic generation of virtual cases according to teaching needs, dynamic control of excrement, and linkage of internal organs; Step S3: Collecting interaction data through the AI standardized patient interaction device, and performing evaluation based on the interaction data, thereby forming an interaction-evaluation mechanism; Step S4: construct a multidimensional evaluation indicator system based on the interaction-evaluation mechanism, generate a multidimensional evaluation report based on the multidimensional evaluation indicator system, analyze the multidimensional evaluation report through a machine learning model, and automatically adjust the teaching content based on the analysis results.
2. The AI-driven virtual-real fusion stoma nursing training method according to claim 1 is characterized in that: The multimodal data acquisition device includes AR glasses, VR gloves and pressure sensors.
3. The AI-driven virtual-real fusion stoma nursing training method according to claim 2 is characterized in that: The step S1 specifically includes: Step S11: Using AR glasses to perform a three-dimensional spatial scan of the stoma physical model to obtain stoma data, including stoma location, stoma size, and spatial relationship data corresponding to the tissue surrounding the stoma. In addition, infrared sensors are used to collect temperature and humidity distribution data corresponding to the skin surface. Step S12: normalizing the tactile feedback data obtained through the VR gloves, confirming the tactile feedback data based on the normalized processing results, and collecting pressure distribution data when the ostomy bag is attached in combination with the pressure sensor; Step S13: using a three-dimensional modeling algorithm to fuse the stoma data, temperature distribution data, humidity distribution data, tactile feedback data and pressure distribution data, and then construct a three-dimensional virtual model.
4. The AI-driven virtual-real fusion stoma nursing training method according to claim 2 is characterized in that: Building a virtual simulation scene specifically includes: Building a personalized case generator based on a deep learning model, inputting teaching requirement parameters based on the personalized case generator, and then generating a virtual case based on the teaching requirement parameters; The pressure sensor collects the corresponding pressure change data in the ostomy bag in real time, and inputs the pressure change data into the preset AI algorithm model to generate excrement data. The excrement data includes excrement flow, excrement characteristics and excrement abnormality data, and displays physiological parameter data in real time. When the pressure change data exceeds the preset pressure threshold, the odor simulation module releases excrement odor; The physiological parameter data is linked to the three-dimensional virtual model, and the flexible intestinal simulator built into the stoma physical model is driven to perform intestinal simulation.
5. The AI-driven virtual-real fusion stoma nursing training method according to claim 4 is characterized in that: The corresponding pressure change data in the ostomy bag is collected in real time through the pressure sensor, including: Acquire historical clinical data, extract a correlation database corresponding to fecal characteristics and pressure changes from the historical clinical data, and perform spatiotemporal distribution analysis on current pressure change data collected by the pressure sensor data to obtain pressure change rate and peak value data; The risk of stoma injury is graded according to the pressure change rate and peak data, and abnormal stool data is generated based on the graded results; The odor simulation module is used to release the corresponding concentration of excrement odor according to the abnormal excrement data.
6. The AI-driven virtual-real fusion stoma nursing training method according to claim 1 is characterized in that: The step S3 specifically includes: Step S31: constructing a dialogue generation system using natural language processing technology, inputting dialogue content data corresponding to the target learner into the dialogue generation system, and then generating response data based on the dialogue content data; Step S32: Capturing the target learner's corresponding facial micro-expressions through a camera device, analyzing the learner's corresponding facial micro-expressions through a facial recognition algorithm, evaluating the degree of empathy in combination with a voice emotion analysis model, and generating an emotional interaction report; Step S33: using computer vision technology to analyze the ostomy bag replacement process frame by frame, confirming normative evaluation data based on the analysis results, and marking the error location with an AR cursor based on the normative evaluation data; Step S34: constructing an interaction-evaluation mechanism based on the response data, the emotional interaction report and the normative evaluation data.
7. The AI-driven virtual-real fusion stoma nursing training method according to claim 6 is characterized in that: Computer vision technology is used to analyze the pouch changing process frame by frame, and normative evaluation data is confirmed based on the analysis results, including: Motion capture of the standard ostomy bag changing process was performed to construct a 3D motion feature vector database; Computer vision technology is used to track the real-time skeleton points of the target learner's ostomy bag replacement process and generate the feature vector corresponding to the current action; The cosine similarity calculation is performed between the feature vector corresponding to the current action and the standard feature vector corresponding to the three-dimensional action feature vector database to confirm the normative evaluation data, and the error detection mechanism is triggered based on the normative evaluation data.
8. An AI-driven virtual-real fusion stoma nursing training system, applied to the AI-driven virtual-real fusion stoma nursing training method according to any one of claims 1 to 7, characterized in that: include: A data acquisition module is used to collect multi-dimensional data of the stoma physical model through a multimodal data acquisition device, and to construct a three-dimensional virtual model based on the collected data; A construction module for constructing a virtual simulation scenario, which includes real-time simulation of the physiological state of the stoma, dynamic generation of virtual cases according to teaching needs, dynamic control of excretion, and linkage of internal organs; An evaluation module, configured to collect interaction data through an AI standardized patient interaction device and perform evaluation based on the interaction data, thereby forming an interaction-evaluation mechanism; An adjustment module is used to construct a multidimensional evaluation indicator system based on the interaction-evaluation mechanism, generate a multidimensional evaluation report based on the multidimensional evaluation indicator system, analyze the multidimensional evaluation report through a machine learning model, and automatically adjust the teaching content based on the analysis results.
9. A computer-readable storage medium, characterized in that: Instructions are stored, and when the instructions are run on a computer, the computer is caused to execute the AI-driven virtual-real fusion ostomy care training method according to any one of claims 1 to 7.