A virtual simulation training method and system for occupational protection against needle stick injuries

Abstract

This invention relates to the field of medical education and training technology, and discloses a virtual simulation training method and system for occupational protection against needlestick injuries. It aims to address the problems of existing training programs lacking immersive interaction, insufficient specialized training on high-risk instruments, subjective operational assessment, and a disconnect between emergency response and practical application. The method includes: constructing a virtual model of multiple grasping points for high-risk instruments such as suture needles and blades; real-time monitoring of grasping positions and feedback of erroneous operations; verifying the compliance of instrument transfer direction based on vector angles; simulating needlestick injuries through collision detection between a sphere in the danger zone and a point cloud of the hand; and achieving smooth instrument adsorption and posture calibration at the target location. The system integrates a VR interaction framework, multimodal feedback, and an emergency response module. Through the above solution, this application achieves safe immersive training, accurate operational assessment, closed-loop emergency response, and highly realistic interaction, significantly improving the occupational protection capabilities of medical personnel.

Description

Technical Field

[0001] This invention relates to the field of medical education and training technology, and in particular to a virtual simulation training method and system for occupational protection needlestick injuries. Background Technology

[0002] Occupational safety and protection is a crucial component of the healthcare field. Especially in clinical procedures, healthcare workers frequently come into contact with various sharp instruments, facing a high risk of occupational exposure. Needlestick injuries, as one of the most common occupational injuries, can not only lead to infections with bloodborne pathogens such as hepatitis B, hepatitis C, and HIV, but also significantly impact the mental health of healthcare workers and the operating costs of medical institutions. Therefore, establishing a scientific, efficient, and practical occupational safety training system has become a key element in improving healthcare safety levels.

[0003] Needlestick injuries are most common in surgical instrument transfer, injection procedures, and sharps retrieval, involving the proper use and handover of high-risk instruments such as suture needles, scalpel blades, and scissors. These procedures require high levels of precision, spatial awareness, and procedural awareness, and traditional training methods are insufficient to effectively cover their complexity and inherent dangers.

[0004] Current occupational safety training technologies largely rely on theoretical lectures, video demonstrations, or limited physical simulations, which have significant limitations: First, the lack of immersive and interactive practical environments makes it difficult for trainees to internalize abstract safety regulations into muscle memory. Second, training in real clinical settings carries actual injury risks, which is detrimental to trainee safety and increases the management burden. Third, existing simulation systems mostly focus on routine injection procedures, lacking specialized modeling and training support for high-risk instrument transfer scenarios unique to operating rooms. Furthermore, the training process lacks sophisticated behavior capture and intelligent assessment mechanisms, preventing instructors from obtaining detailed operational data and leading to highly subjective assessment results. Finally, preventative operational training is disconnected from emergency response procedures after needlestick injuries, resulting in insufficient response capabilities for trainees in real-world accidents. These problems collectively restrict the effectiveness and systematic nature of occupational safety training, necessitating a new training method and system that integrates virtual reality, intelligent interaction, and multi-dimensional assessment. Summary of the Invention

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A virtual simulation training method and system for occupational protection against needlestick injuries includes the following specific steps: Step (1) Constructing a refined virtual model of high-risk instruments and configuring multiple gripping points: In the virtual environment, three-dimensional models of high-risk instruments such as suture needles, surgical blades, and scissors are created, and multiple gripping points are configured for each instrument model. Each gripping point corresponds to a safe operating position. The instrument model is bound to the gripping point through a virtual reality interaction framework, so that trainees can only grip through the preset safe gripping point during operation; Step (2) Real-time monitoring of instrument gripping actions and judgment of gripping position correctness: When trainees touch the instrument model with a virtual handle, the system automatically calculates the spatial distance between the handle and each gripping point, selects the nearest gripping point as the actual gripping position, and identifies whether the position conforms to safety regulations through the gripping point name. If the gripping position is incorrect, a visual and operational feedback mechanism is immediately triggered; Step (3) Performing instrument transfer direction compliance verification: During the instrument transfer process, the system obtains the local coordinate vector of the instrument tip and the receiver relative to the instrument in real time. The displacement vector of the center point of the device is used to calculate the cosine value of the angle between the two vectors to generate the compliance coefficient of the transmission direction. When the coefficient is less than zero, it is determined to be a safe transmission direction; otherwise, it is determined to be a dangerous operation and a warning is triggered. Step (4) Implement real-time collision detection of the dangerous area of ​​the needle tip or blade: Model the dangerous part of the device as a sphere with a fixed radius. Its center is updated in real time with the movement of the device. At the same time, simplify the human hand model into a set of point clouds. When the distance between the coordinates of any hand point cloud and the center of the dangerous sphere is less than the sum of the radius of the sphere and the buffer thickness, it is determined to be a needle prick accident, and vibration feedback and sound prompts are immediately activated. Step (5) Achieve smooth adsorption and posture calibration of the device at the target position: When the device approaches the target placement position, the system detects the distance between it and the target Socket. When the distance is less than the preset threshold, cancel the gravity simulation of the physics engine, perform kinetic energy decay mapping on the device speed, and guide the device position to move towards the center of the Socket through the interpolation function. At the same time, check whether the device posture is within the angle tolerance range to complete the safe handover.

[0007] Preferably, in step (1), each high-risk device model is attached with a graspable component. This component is attached to multiple empty objects in the virtual engine through a sub-object method. Each empty object represents a grasp point and is attached with a grasp point component. All grasp points form a list for the system to access dynamically during runtime. The spatial position of the grasp point strictly corresponds to the actual safe gripping area of ​​the device.

[0008] Preferably, in step (2), the grasping offset compensation algorithm calculates the initial transformation matrix of the device model coordinate system relative to the handle coordinate system at the moment of grasping trigger, and uses a smooth following formula to update the device position in each subsequent frame update, wherein the smoothing factor is set to a value within a predetermined range to suppress sensor jitter of virtual reality device, ensure stable and continuous movement trajectory of device, and avoid model clipping or visual jump.

[0009] Preferably, the criteria for determining the compliance coefficient of the transmission direction in step (3) are as follows: when the cosine value is less than zero, it means that the tip of the instrument is pointing to the receiver, which complies with the safe operation specifications; when the cosine value is greater than or equal to zero, it means that the tip is pointing to the operator or deviating from the safe direction, the system determines it as a violation of the operation, and displays a red "×" icon and operation error prompt through the user interface module.

[0010] Preferably, in step (4), the radius of the sphere in the danger zone is set according to the type of equipment, the buffer thickness is uniformly set to a preset value, the point cloud set is generated by sampling the vertices of the hand model surface, and the sampling density is not lower than the predetermined density to ensure the sensitivity and coverage integrity of the collision detection.

[0011] Preferably, in step (5), the Socket component is attached to the empty object at the target position. It contains a collider and a preset posture. When the device enters the collider range, the system compares the Euler angle deviation between the current posture of the device and the preset posture of the Socket. If the angle deviation of each axis is less than the preset angle tolerance, the posture matching is successful and the placement operation is allowed to be completed.

[0012] Preferably, the highlighting and visual feedback mechanism calls the activation method of the highlighting component for the currently operated device at the start of training, so that it produces a periodic flashing light effect to guide the trainee's attention. When the device is correctly grasped, the stop method is automatically called to turn off the highlighting. If the trainee grasps the wrong position, the OperateStateUI module immediately calls the operation error display method, presenting a visually impactful interface composed of a red background and warning icons.

[0013] Preferably, after determining that a needlestick injury has occurred, the system not only triggers vibration feedback and sound prompts, but also automatically records the time of the accident, the type of instrument, the contact site, and the operational context, and then switches to the emergency response training module to guide trainees to complete standard procedures such as wound rinsing, report registration, and serological testing, thereby achieving integrated training in prevention and treatment.

[0014] Preferably, the angle tolerance parameter can be dynamically adjusted according to different instrument types. The system manages the tolerance values ​​of various instruments uniformly through configuration files to ensure the scientific nature of the evaluation standards and the adaptability to different scenarios.

[0015] Preferably, the adsorption coefficient k in the kinetic energy decay mapping is a predetermined value, the kinetic energy decay coefficient is another predetermined value, and the interpolation function adopts linear interpolation to ensure that the device decelerates smoothly and is accurately aligned within a predetermined time period, simulating the stable feel and operating rhythm of device handover in real clinical practice.

[0016] Compared with the prior art, the present invention has the following beneficial effects:

[0017] 1. Safe and immersive training environment

[0018] This invention constructs a risk-free operational training scenario using high-fidelity virtual reality technology, completely avoiding actual needlestick injuries caused by operational errors in traditional clinical internships. It reduces training risks and management costs for medical institutions while ensuring trainees' safety. The system supports specialized training with various high-risk operating room instruments such as suture needles, blades, and scissors, filling a gap in existing simulation systems for the transfer of complex instruments.

[0019] 2. Refined operation capture and intelligent evaluation

[0020] By employing multi-grasp point configuration, a grasp offset compensation algorithm, and a transmission direction verification mechanism, the system can accurately identify subtle deviations in trainees' operations, such as incorrect grip position or improper tip pointing, which are considered key risk behaviors. The evaluation results are generated based on objective data, avoiding the arbitrariness of subjective judgment in traditional training and providing a quantitative basis for teaching feedback.

[0021] 3. Real-time needlestick injury simulation and closed-loop treatment training

[0022] The hazardous area collision detection model monitors unauthorized contact in real time with high precision. Once triggered, it enhances risk perception through multimodal feedback (vibration, sound, and vision). More importantly, the system seamlessly integrates with standard emergency response procedures after an accident, allowing trainees to experience the entire "prevention-exposure-response" process in a virtual environment, significantly improving their emergency response capabilities in real-world scenarios.

[0023] 4. Highly realistic interactive experience and smooth operation

[0024] The combined effect of grasp offset compensation and Socket adsorption kinetic energy attenuation model effectively eliminates operational instability caused by virtual device jitter, while simulating the physical feel of real instrument handover. The angle tolerance mechanism balances operational standardization with clinical flexibility, making training both rigorous and consistent with real-world work logic, thus enhancing skill transfer.

[0025] 5. Full-process data recording and teaching optimization support

[0026] The system automatically records complete behavioral data for each operation, including the selection of grab points, transmission trajectories, collision events, and response measures, forming a structured training archive. Teachers can use this data for individualized guidance or for large-scale training effectiveness analysis, promoting the standardization and intelligent upgrading of occupational safety education. Attached Figure Description

[0027] Figure 1 A flowchart of a virtual simulation training system for occupational protection needlestick injuries provided in an embodiment of the present invention.

[0028] Figure 2 This is a diagram of the user interface of the present invention. Detailed Implementation

[0029] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” or “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0030] This embodiment is applied to a high-risk instrument delivery and occupational safety training scenario in the operating room. It aims to construct an immersive, interactive, and intelligently assessable needlestick injury prevention training system using virtual reality (VR) technology. The system is deployed on a high-performance workstation platform, integrating a head-mounted VR display, a six-degrees-of-freedom (6-DoF) controller, haptic feedback devices, and a multi-channel audio output unit to form a complete closed-loop human-computer interaction training environment.

[0031] At the system architecture level, the hardware platform used in this embodiment includes: a graphics workstation equipped with an Intel Core i9-13900K processor, NVIDIA RTX 4090 GPU, 64GB DDR5 memory, and 2TB NVMe SSD, serving as the main computing node; two Valve Index or Meta Quest Pro-level VR headsets, used for the simulation of the main operator's and receiver's perspectives respectively (supporting single-person dual-role switching mode); a matching Lighthouse 2.0 base station or Inside-Out tracking system, providing sub-millimeter-level spatial positioning accuracy; a controller with a built-in linear resonant actuator (LRA), supporting programmable vibration feedback in the 0-100Hz frequency range; and an audio system using a 7.1-channel surround sound card in conjunction with headphones to achieve spatial sound cues.

[0032] The software system is developed based on the Unity 2022 LTS engine, integrating the HurricaneVR interaction framework as the core VR interaction middleware, and supplemented by five self-developed functional modules: HighRiskInstrumentManager, CollisionMonitor, DirectionValidator, SocketAligner, and OperateStateUI. Each module is deeply coupled with the HurricaneVR underlying API through an event-driven mechanism, forming a unified closed loop of operation logic.

[0033] like Figure 1 The virtual simulation training system proposed in this invention adopts a modular layered architecture design, mainly consisting of four layers: user management layer, scene control layer, interaction execution layer, and data evaluation layer.

[0034] System Overall Architecture

[0035] User Management

[0036] This layer is responsible for managing student information and initializing the system. The LoginStart module implements the student login interface, which displays a list of students through a drop-down menu (TMP_Dropdown). Students can select a student to view their personal historical scores and training records.

[0037] The ExcelRead module is designed using the singleton pattern and is responsible for reading student information from external Excel files, including student ID, name, and historical grades. The system parses the Excel data into a StudentType structure and stores it in a studyList for other modules to access.

[0038] This design, based on external data files, allows the system to flexibly manage large amounts of student information, adding or updating student data without modifying the program code.

[0039] Scene control layer

[0040] This layer is responsible for scheduling the training process and managing the scene. The SelTask ​​module is the core of scene control. It implements the task selection interface through Toggle control groups, with each Toggle corresponding to a training task. The system adopts a progressive unlocking mechanism. The student must complete the current task before the Toggle for the next task will be activated (interactable=true) and the lock icon will be removed.

[0041] The Level series classes (Level00-Level04) are the concrete implementations of each training level. Each Level class inherits from MonoBehaviour and contains the complete training logic for that level. The Level classes implement flow control through coroutines and use the DOTween animation library to implement perspective switching and character animation.

[0042] Scene switching is implemented using Unity's SceneManager. The system has multiple preset scenes such as "operating room," "ward," and "treatment room," allowing trainees to switch between different scenes according to their training progress.

[0043] Interactive Execution Layer

[0044] This layer is the core execution layer of the system, responsible for implementing VR interaction and multimodal feedback.

[0045] The knowledge presentation module includes IntroWindows and DialogWin. IntroWindows displays illustrated knowledge introductions and supports customizable display duration and a shutdown callback function. DialogWin implements NPC character dialogue, allowing users to set character facial animations, dialogue content, button text, and response events.

[0046] The VR interaction module is based on the HurricaneVR framework. The HVRGrabbable component gives virtual objects the property of being grabbable. Each grabbable object can be configured with multiple GrabPoints and corresponding HandPoser gestures. The HVRSocket component defines the target position of the object. The Socketed event is only triggered when the Grabbable object is moved into the space of the Socket and the gesture matches.

[0047] The multi-sensory feedback module includes three dimensions of feedback: visual (OperateStateUI displays correct / incorrect icons, Highlighter makes objects highlight and flash), auditory (PlayPYAudio plays dubbing and sound effects), and tactile (vibration via VR controllers).

[0048] Data evaluation layer

[0049] This layer is responsible for recording and analyzing trainees' training data in real time.

[0050] The ScoreItems class is the core of the scoring system. It uses the singleton pattern and is marked as DontDestroyOnLoad to ensure that data is not lost when switching scenes. This class maintains a baseline score of 100 points and a RecordScore list. Each element in the list is a ScoreType structure that records the error number, deduction value, error description, error type, and occurrence time.

[0051] The TimeCounter class records the total training time, while the RecTime class implements the countdown and timeout detection for single-step operations. The system uses coroutines to implement the timing function, and automatically executes the preset deduction callback when a timeout is triggered.

[0052] The ScoreManager class is responsible for generating the final evaluation report. It reads all record data from ScoreItems, calculates the total score, assigns a grade, counts the number of errors of each type, and presents it to the student in a visual way through a UI interface.

[0053] Specifically, such as Figure 2 The system's high-risk surgical instrument model library consists of typical surgical instruments such as suture needles, scalpel blades, tissue scissors, and needle holders. Each instrument is modeled in high-fidelity 3D using Blender, with a face count between 5,000 and 15,000 triangles and a texture resolution of at least 2K. Surface detail realism is enhanced through normal mapping and PBR materials. Each instrument model is instantiated as a GameObject in the Unity scene and an HVRGrabbable component is attached, giving it grabbable properties. Under this GameObject, multiple empty child objects are created, each representing a safe gripping point and attaching an HVRGrabPoint component. For example, for a needle holder, the system configures three gripping points: the proximal handle area (GrabPoint_Hold_Base), the mid-section anti-slip groove area (GrabPoint_Hold_Mid), and the distal ring area (GrabPoint_Hold_Ring), corresponding to three safe gripping methods in clinical guidelines. The spatial coordinates of all gripping points are strictly set according to the ergonomic gripping area of ​​the actual device, with the error controlled within ±1mm.

[0054] At runtime, the HighRiskInstrumentManager module dynamically obtains a list of GrabPoints (of type List) for each instrument through reflection. <hvrgrabpoint>Before training begins, the corresponding instrument configuration file (JSON format) is loaded based on the current task. This file contains parameters such as a whitelist of grab point names, the radius of the hazard zone sphere, and the angle tolerance threshold. For example, the configuration items for the suture needle are: { "allowedGrabPoints": ["Tip_Guard", "Shaft_Safe"], "hazardRadius_mm": 2, "angleTolerance_deg": 10}.

[0055] Once the trainee puts on the VR headset and starts the training program, the system first enters the guidance phase. At this time, the OperateStateUI module calls the TweenStart() method of the Highlighter component, causing the surface of the instrument to be operated (such as a suture needle) to produce a bluish-white highlight effect with a period of 800ms and a gradual change in brightness, guiding the trainee's attention to the target object. At the same time, the voice prompt module plays the instruction "Please use a safe grip to pick up the suture needle".

[0056] Once the grasping phase begins, the trainee uses the VR controller to approach the sewing needle model. The HurricaneVR framework continuously monitors the collision status between the controller Collider and the instrument's HVRGrabbable component. Upon triggering the OnGrabbed event, the system immediately performs the following actions:

[0057] First, the grab point identification sub-process begins: The system iterates through the GrabPoints list of the sewing needle, calculates the Euclidean distance between the handle center point (i.e., HVRGrabberBase.transform.position) and the world coordinates of each grab point, and selects the one with the smallest distance as the actual grab point. Assuming the nearest point is "Shaft_Safe", the system reads its name attribute and compares it with the allowedGrabPoints whitelist in the configuration file. If a match is found, the grab is considered correct; otherwise, an error feedback is triggered.

[0058] Secondly, the grasping offset compensation algorithm is activated synchronously. At the instant of grasping (t=0), the system calculates the coordinate system of the instrument model. Relative to the handle coordinate system Transformation matrix :

[0059]

[0060] In each subsequent frame update, the instrument's real-time position Follow the following smooth following formula:

[0061] Parameter definition: Smoothing factor It is used to eliminate sensor jitter in VR devices and ensure visual stability of the instrument during transmission.

[0062] Among them, smoothing factor Setting it to 0.2 effectively suppresses model jumps or clipping caused by high-frequency jitter of the VR sensor, ensuring the continuous and stable motion trajectory of the device. This algorithm is called in real time by the underlying motion prediction subsystem of the DirectionValidator module, providing reliable pose data for subsequent orientation determination.

[0063] If the gripping position is incorrect (e.g., the student accidentally grips the area near the needle tip), the OperateStateUI module immediately calls the ShowUIOperateWronge() method, popping up a full-screen red warning interface in the center of the student's field of vision. The background color value is #FF0000, and a flashing white "×" icon and the text "Incorrect gripping position! Please use the safe area to grip!" are superimposed. At the same time, the LRA controller performs three short vibrations (50ms each, 100ms interval) to form strong multimodal feedback.

[0064] During the instrument transfer phase, the system initiates a compliance verification of the transfer direction. It is assumed that the current operator is passing a suture needle to a virtual recipient (whose hand model is represented by another HVRGrabbable, with a fixed position or controlled by AI).

[0065] During the "scalpel / scissors passing" process, it is essential to ensure that the tip of the instrument is pointed at the injured person, not yourself.

[0066] Formula description:

[0067] Let the center point of the instrument be O, the local coordinate vector of the tip (or dangerous end) be d, and the displacement vector of the receiver relative to the center point of the instrument be... .

[0068] Transmission direction compliance coefficient The calculation is as follows:

[0069] Judgment criteria:

[0070] Technical Explanation: This formula determines the angle between vectors using the cosine value. .like obtuse angle ( If the value is negative, it indicates that the tip is pointing in a safe direction. Immediately display a red "×" icon in the upper right corner of the UI and play a low-frequency warning sound (440Hz, lasting 1 second).

[0071] Meanwhile, the CollisionMonitor module continuously performs real-time collision detection of the needle tip's hazardous area. This module models the suture needle tip as a sphere collider with a radius of 2 mm, and its center C(t) is updated in real time as the instrument moves. The surface of the receiver's and operator's hand models is sampled as a point cloud set with a sampling density of 10 points per square centimeter (higher than the minimum requirement of 8 points / square centimeter), and the coordinates of each point are denoted as P_i. The system traverses all hand points every frame and calculates their distance from the center of the hazardous sphere.

[0072] If a needlestick injury occurs, the system immediately triggers a triple response: (1) the LRA handle performs a high-intensity long vibration (300ms, 100Hz); (2) the audio system plays a sharp and piercing "ding!" sound (sampled from a real metal impact sound); (3) the OperateStateUI switches to the accident emergency interface and automatically loads the emergency response training module, guiding trainees to complete the standard procedures such as "immediately rinse the wound → report to the head nurse → fill in the occupational exposure registration form → draw blood for testing" in sequence. Each step must be confirmed through VR interaction to form a closed-loop training of "prevention-exposure-response".

[0073] When the instrument approaches the target placement location (such as a socket on the instrument tray), the SocketAligner module initiates a smooth attachment and attitude calibration process. The target socket is defined by an empty object with an attached HVRSocket component, which contains a spherical trigger with a radius of 30mm. When the instrument's center point enters this range (distance <30mm), the system performs the following operations:

[0074] 1. Cancel gravity simulation: Set the useGravity property of the Rigidbody to false to avoid falling interference;

[0075] 2. Kinetic energy decay: Apply decay mapping to the current velocity V of the instrument, with a kinetic energy decay coefficient of 0.9;

[0076] 3. Position Interpolation: Position P is interpolated towards the center of the Socket according to the interpolation function. move:

[0077] Technical Note: k is the adsorption coefficient. This represents the kinetic energy decay coefficient. This algorithm ensures a smooth transition of the device into the socket, simulating the real-world process of steadily taking over the device. The adsorption coefficient k=0.8 ensures smooth deceleration within the last 100ms.

[0078] 4. Posture Verification: Calculate the deviation between the current Euler angles of the instrument and the preset posture of the Socket. If the deviations of the X, Y, and Z axes are all less than 10 degrees (needle tolerance), the posture is considered to be successfully matched and placement is allowed; otherwise, the instrument will slightly spring back and display the message "Angle mismatch, please adjust the grip direction".

[0079] The data flow throughout the training process is uniformly scheduled by the central event bus: grab events, orientation verification results, collision detection signals, and adsorption states are all published in the form of structured events (such as GrabEvent, DirectionViolationEvent, NeedlestickEvent). The DataLogger module serializes the complete context of each operation (timestamp, device ID, grab point name, transfer vector, collision point coordinates, and response time) into JSON format and stores it in a local SQLite database for teachers to perform retrospective analysis and personalized teaching.

[0080] In summary, this embodiment constructs a virtual simulation training system for occupational protection needlestick injuries that combines high realism, strong interactivity, and intelligent assessment capabilities through precise hardware selection, modular software architecture, and deeply integrated step logic. It effectively solves the core pain points of traditional training, such as high risk, weak feedback, and fragmented processes.

[0081] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.< / hvrgrabpoint>

Claims

1. A virtual simulation training method for occupational needlestick injuries, characterized in that, Includes the following steps: In a virtual environment, 3D models of high-risk instruments such as suture needles, surgical blades, or scissors are constructed, and multiple gripping points are configured for each instrument model. Each gripping point corresponds to a safe operating position. The instrument model is bound to the gripping points through a virtual reality interaction framework, so that trainees can only grip through preset safe gripping points. When a trainee touches the equipment model with a virtual handle, the system calculates the spatial distance between the handle and each gripping point, selects the nearest gripping point as the actual gripping position, and identifies whether the position meets the safety specifications by the gripping point name. If it does not meet the safety specifications, the system triggers a visual and operational feedback mechanism. During the instrument transfer process, the system acquires the local coordinate vector of the instrument tip and the displacement vector of the receiver relative to the instrument center point in real time, calculates the cosine value of the angle between the two vectors to generate a transfer direction compliance coefficient. When the coefficient is less than zero, it is determined to be a safe transfer direction; otherwise, it is determined to be a dangerous operation and a warning is triggered. The dangerous parts of the instrument are modeled as spheres with a fixed radius, and the center of the sphere is updated in real time as the instrument moves. At the same time, the human hand model is simplified into a set of point clouds. When the distance between the coordinates of any hand point cloud and the center of the dangerous sphere is less than the sum of the radius of the sphere and the buffer thickness, it is determined that a needlestick injury has occurred, and vibration feedback and sound prompts are immediately activated. When the device approaches the target placement location, the system detects its distance from the target Socket. If the distance is less than a preset threshold, the physics engine gravity simulation is canceled, kinetic energy decay mapping is performed on the device speed, and the device position is guided to move towards the center of the Socket through an interpolation function. At the same time, the device posture is checked to see if it is within the angle tolerance range, thus completing the safe handover.

2. The occupational protection needlestick injury virtual simulation training method according to claim 1, characterized in that, Each high-risk medical device model is attached with a graspable component. In the virtual engine, this component is attached to multiple empty objects through sub-objects. Each empty object represents a grasp point and has a grasp point component attached. All grasp points form a list for the system to access dynamically at runtime. The spatial location of the grasp points strictly corresponds to the actual safe gripping area of ​​the medical device.

3. The occupational protection needlestick injury virtual simulation training method according to claim 1, characterized in that, The initial transformation matrix of the device model coordinate system relative to the handle coordinate system is calculated at the moment of triggering. In each subsequent frame update, the device position is updated using a smooth following formula. The smoothing factor is set to a value within a predetermined range to suppress sensor jitter of the virtual reality device and ensure that the device's motion trajectory is stable and continuous.

4. The occupational protection needlestick injury virtual simulation training method according to claim 1, characterized in that, The criteria for determining the compliance coefficient of the transfer direction are as follows: when the cosine value is less than zero, it means that the tip of the instrument is pointing towards the receiver, which complies with the safe operating procedure; when the cosine value is greater than or equal to zero, it means that the tip is pointing towards the operator or deviating from the safe direction, and the system determines it as a violation of the operation and displays a red "×" icon and operation error prompt through the user interface module.

5. The occupational protection needlestick injury virtual simulation training method according to claim 1, characterized in that, The radius of the sphere in the danger zone is set according to the type of equipment, the buffer thickness is uniformly set to a preset value, and the point cloud set is generated by sampling the vertices of the hand model surface. The sampling density is not lower than the predetermined density to ensure the sensitivity and coverage integrity of collision detection.

6. The occupational protection needlestick injury virtual simulation training method according to claim 1, characterized in that, The Socket component is attached to an empty object at the target location. It contains a collider and a preset orientation. When the device enters the collider's range, the system compares the Euler angle deviation between the device's current orientation and the Socket's preset orientation. If the angle deviation of each axis is less than the preset angle tolerance, the orientation is considered to be successfully matched, and the placement operation is allowed to be completed.

7. The occupational protection needlestick injury virtual simulation training method according to claim 1, characterized in that, At the start of training, the activation method of the highlight component is called for the currently operated device, causing it to produce a periodic flashing light effect to guide the trainee's attention. When the device is correctly grasped, the stop method is automatically called to turn off the highlight. If the trainee grasps the wrong position, the OperateStateUI module immediately calls the operation error display method, presenting a visually striking interface with a red background and warning icons.

8. The occupational protection needlestick injury virtual simulation training method according to claim 1, characterized in that, After determining that a needlestick injury has occurred, the system not only triggers vibration feedback and sound prompts, but also automatically records the time of the accident, the type of instrument, the contact site, and the context of the operation, and then switches to the emergency response training module to guide trainees to complete the standard procedures of wound rinsing, reporting and registration, and serological testing.

9. The occupational protection needlestick injury virtual simulation training method according to claim 6, characterized in that, The angle tolerance parameter is dynamically adjusted according to the type of instrument. The system manages the tolerance values ​​of various instruments in a unified manner through configuration files to ensure the scientific nature of the evaluation standards and the adaptability to the scenario.

10. A virtual simulation training system for occupational protection needlestick injuries based on the method of claim 1, characterized in that, It includes a high-risk device modeling module, a gripping point configuration module, a gripping action monitoring module, a transfer direction verification module, a collision detection module, an adsorption calibration module, and an operation status feedback module; The high-risk medical device modeling module is used to construct three-dimensional models of suture needles, surgical blades, or scissors. The gripping point configuration module configures multiple gripping points for each instrument model and binds them to the virtual reality interaction framework; The grasping motion monitoring module calculates the distance between the handle and the grasping point to identify the correctness of the grasping position; the transmission direction verification module judges the compliance of the transmission direction based on the cosine value of the instrument tip vector and the receiver displacement vector. The collision detection module models the dangerous area as a sphere and compares the distance with the hand point cloud to determine the needle prick injury. The adsorption calibration module performs kinetic energy decay, position interpolation, and attitude verification when the instrument approaches the target socket. The operation status feedback module triggers visual, vibration, or sound feedback based on the operation result, and initiates the emergency response training process when an accident occurs.

Images