Vascular intervention operation man-machine interaction method and system based on virtual reality

Abstract

A virtual reality-based vascular interventional operation man-machine interaction method is characterized by comprising the following steps: S1, carrying out high-fidelity solution on deformation of a superfine guide wire under a position dynamics framework by adopting a Cosseerat elastic rod in combination with quaternion, and establishing a non-penetrating vascular environment through an expanded surrounding layer, spatial hash and continuous collision detection; s2, constructing a three-level virtual hand model containing bones, soft bodies and adhesion collision particles; s3, binding and adhering collision particles to finger pulps, and realizing stable grabbing and releasing mechanical interaction with guide wire particles through collision filtering; s4, mapping a hallux and index finger twisting displacement difference into a guide wire tip turning speed, and realizing fine control of a vascular fork path by matching with curvature self-adaptive traction; and S5, integrating the modules and realizing real-time gesture and visual interaction feedback through a head-mounted display. The method solves the problem of fine interaction between the virtual hand and the guide wire and the problem of vascular tunnel effect, and can be used for vascular intervention training.

Description

Technical Field

[0001] This invention relates to the field of virtual reality and computer simulation technology, specifically to a human-computer interaction method and system for vascular interventional surgery based on virtual reality. Background Technology

[0002] With changing lifestyles and an aging population, chronic diseases such as cardiovascular disease are seriously impacting human health. Interventional vascular surgery is one of the most effective treatments for cardiovascular and cerebrovascular diseases, and its application in diagnosis and treatment has developed rapidly. Compared to traditional cardiac surgery, it offers advantages such as less trauma, less pain, and faster postoperative recovery. However, interventional vascular surgery typically involves highly complex technical procedures, such as catheter placement and thrombus removal, requiring physicians to master sophisticated interventional techniques and equipment usage skills. Traditional surgical training methods rely on physical models, animal experiments, or clinical observation, which suffer from drawbacks such as high cost, high risk, poor repeatability, and limited training scenarios.

[0003] With the application of virtual reality technology in the medical field, virtual training methods for vascular interventional surgery are of great significance in terms of skills development, training efficiency, and medical education reform. Virtual reality-based training for vascular interventional surgery provides a safe and efficient learning environment, allowing trainees to practice repeatedly in simulated surgical scenarios without posing risks to real patients.

[0004] However, existing research on training systems for vascular interventional surgery based on virtual reality technology largely focuses on the interactive simulation of surgical instruments and organ models, mainly involving the interaction between catheters / guidewires and blood vessels. It neglects the human-computer interaction design between the user and the system, and rarely considers real-time interaction between hand models and guidewire models with ultra-long, ultra-thin, and ultra-flexible characteristics in a virtual environment. This results in the inability to effectively train fine hand manipulation techniques during vascular intervention. Therefore, how to construct a stable mechanical interaction model between the virtual hand and ultra-thin, ultra-flexible guidewires, and on this basis avoid the "tunneling effect" caused by traditional discrete collision detection methods, is a key problem that urgently needs to be solved. Summary of the Invention

[0005] To address the issues of existing virtual reality vascular interventional surgery training systems failing to maintain high-fidelity simulation training of hand movements and stability of hand-guidewire interaction, this invention proposes a virtual reality-based human-computer interaction method for vascular interventional surgery.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: First, a deformation simulation and stable collision environment module for ultra-fine guidewires is constructed. A discretized Cosserat elastic rod model is used in combination with the quaternion method for constraint, and the solution is obtained in the position dynamics framework to achieve high-fidelity simulation of the guidewire. Then, a stable and non-penetrating vascular interaction environment is established by using continuous collision detection technology combined with an extended enclosing layer. Secondly, a multi-layered virtual hand model was created, including a skeletal layer, a deformable soft body layer, and an adhesive collision particle layer, providing a structural basis for mechanical interaction. Then, an interactive mechanical module between the virtual hand and the guide wire is constructed, an interactive mechanical model between the adhesive collision particles and the guide wire particles is established, and the virtual hand can stably grasp and release the guide wire through collision filtering optimization calculation. Subsequently, a fine control module for the guidewire tip based on gesture mapping was developed, which maps the displacement difference of the hand twisting action into the speed and direction command of the guidewire tip. Combined with traction control, it realizes fine active control of the guidewire's direction of travel. Finally, by integrating the aforementioned modules, high-fidelity control simulation of the interaction between the virtual hand model and the guidewire is achieved, while real-time gesture interaction and visual feedback are completed through the head-mounted display device during surgical training.

[0007] Furthermore, the deformation simulation and stable collision environment module of the ultra-fine guidewire achieves high-fidelity restoration of the guidewire and establishes a stable, non-penetrating virtual vascular environment. The specific steps for constructing the deformation simulation and stable collision environment of the ultrafine guide wire include: (1) Physical simulation of ultra-fine guidewire: The discretized Cosserat elastic rod model is combined with the quaternion method for constraint, and the solution is discretized in the position dynamics framework to realize the deformation simulation of the guidewire model, and efficiently and accurately simulate its bending and torsion behavior in actual surgery. (2) Stable collision environment: Construct axis-aligned bounding boxes for all initial objects, perform coarse collision detection, use spatial hashing algorithm to filter out possible intersecting objects more quickly, and then apply continuous collision detection technology to improve the efficiency and accuracy of detection, avoid the failure of traditional discrete collision detection methods caused by the small diameter of the guidewire, high relative velocity and simulation time step during rapid intervention, and establish a stable, non-penetrating blood vessel interaction environment. Furthermore, the physical simulation of constructing the ultrafine guidewire specifically includes the following steps: Cosserat elastic rod model: Based on the Cosserat elastic rod theory, the mechanical behavior of the guidewire is accurately modeled using parameters. s The parameter represents the continuity of the elastic rod from its starting point to its ending point. s =0, parameter s=1 corresponds to the start and end points of the elastic rod, respectively, and at each point a vector consisting of three orthogonal basis vectors is attached. d 1, d 2, d 3] constitutes a control frame, in which d 2 and d 3 represents the normal vector and tangent vector of the particle, respectively. d 1 represents the product of the two. The Darboux vector is used to describe the rate of change of the frame's rotation along the arc length: (1) in, Indicates to The derivative, k =1, 2, 3, These represent three different orthogonal basis direction vectors. The bending and torsional strains of the curve can be obtained by projecting the Darboux vector onto the coordinates of the control points. (2) Quaternion method: Using unit quaternions to represent the rotation of the control frame ensures the smoothness and physical accuracy of the rotation state when the guidewire undergoes a wide range of bending. The Darboux vector of the control frame can be parameterized as a direction quaternion function. , Represents the rotation quaternion in the current state, where This represents the scalar part, which is related to the rotation angle of the continuous rod model. x , y , z Let ) represent the vector part. The strain between the current state and the resting state can be expressed as: (3) in Represents a rotational quaternion in a resting state. A directional quaternion function in a resting state; Position dynamics framework: Using position dynamics as the core solution framework for guide wire deformation simulation, it can efficiently handle large deformations and complex constraints, discretizing the continuous guide wire model into particles. P n Furthermore, properties such as elasticity and instretchability are expressed as constraint functions relating to the particle's position. C ( P =0. In each simulation time step, the particle position is directly corrected by iteratively solving the constraint projection equations, integrating the strain constraints of the Cosserat elastic rod into the position dynamics framework. Specifically, this is achieved through discretized rod elements, obtaining the discretized form based on the rod elements from adjacent quaternions through arithmetic mean interpolation. (4) in, The vector representing the imaginary part of a quaternion. and This represents the quaternion in two adjacent link elements. and Represents the quaternion in its resting state; Furthermore, the construction of a stable collision environment specifically includes the following steps: Spatial hashing algorithm: Construct axis-aligned bounding boxes for all initial objects. The bounding boxes are hashed to be assigned to each cubic cell for filtering. Then, the spatial hashing algorithm is used to assign the box indices to a hash index array. By traversing this hash array, collision detection is performed on each element with the same index, and collision detection is not performed on elements with different indices, thereby reducing the computational complexity of the collision process and avoiding unnecessary computation time. Continuous collision detection: For potential collision pairs selected through wide phase filtering, the state of the objects at the beginning and end of the time step is detected, and the states of both are calculated over the entire Δ time step using linear or nonlinear interpolation. t The movement path within the space fundamentally avoids the "tunneling effect" caused by excessive displacement or high speed of an object, which is the problem of an object moving instantaneously from one side of another object to the other without triggering a collision. Extended bounding layer technique: A geometric offset layer is constructed in the normal direction of the surface of an extremely thin blood vessel model. All collision detection is performed between the extended bounding layer and the guidewire particles, so that even if calculation errors are caused by discrete time steps when the guidewire passes through narrow bends at high speed, the extended layer can reliably capture them, thereby avoiding the "tunneling effect".

[0008] Furthermore, the creation of the multi-layered virtual hand model specifically includes: (1) Skeletal layer: A hand mesh model is constructed using 3D modeling software, and it is associated with a predefined skeletal structure through bone weight binding technology, serving as the basic skeleton and visual representation to drive the entire hand movement; (2) Deformable soft layer: The soft layer located on the fingertip is simulated by distance constraints and volume constraints in position dynamics to simulate the deformation of biological soft tissue when in contact with an object; (3) Adhesive collision particle layer: The collision particle system with adhesive properties is bound to the fingertip bones of the thumb and index finger so as to perform mechanical interaction with the guide wire; The deformable soft layer is implemented by specifically including the following steps: Based on the principle of positional dynamics: the deformable fingertip is discretized into a set of particles, and the physical deformation characteristics of the model are expressed as constraint functions. C ( u Solving for the condition using the constraint projection step.C ( u Particle position correction Δ = 0 u It directly updates particle positions, achieving efficient and stable deformation simulation; Distance constraint: For adjacent particles in the fingertip model, the constraint function is defined as the difference between the current distance and the initial distance. Through the projection of this constraint, the particles can recover the initial spacing after deformation, thereby simulating the elastic properties of the fingertip tissue. Volume constraint: For the fingertip tetrahedral mesh element, its constraint function is defined as the difference between the current volume and the initial volume. Through the projection of this constraint, it is ensured that the volume of each tetrahedron remains basically unchanged during deformation, thereby realistically simulating the incompressibility of biological soft tissue. Furthermore, the interaction mechanics module for constructing the virtual hand and guidewire, in which the adhesion collision particle mechanics model solves the problem that the interaction effect of the virtual hand and guidewire cannot reach the accuracy of the interaction between the real hand and guidewire, and ensures the efficiency and stability of the interaction through collision filtering; The implementation of the interactive mechanical module between the virtual hand and the guidewire specifically includes the following steps: Adhesive Collision Particle Mechanical Model: A set of collision particles with adhesive properties are bound to the fingertip bones of the virtual thumb and index finger. A contact mechanical model is established between these particles and the particles on the guide wire surface, including tangential forces. f t Axial force f z The combined effect of these factors determines the adhesion and fixation of hand particles on the guidewire surface, and the adhesive force of the hand-collision particles. f The tangential and axial forces are expressed as: (5) (6) in, N ( x , y It is the pressure at any given contact point. κ The adhesion coefficient, It is a constant. A It is the contact area, when the adhesion parameter κ When the value is 0.4, the virtual hand model can have good interactive effects in the process of extracting the guidewire model; Collision filtering: Independent collision categories (category 1, 2, 3) are assigned to deformable fingertip soft bodies, adhesive collision particles, and guide wire models respectively. Collision mask rules are configured: guide wires (masks containing 1 and 2) can collide with both fingertip soft bodies (category 1) and adhesive collision particles (category 2); while fingertip soft bodies (masks containing only 3) and adhesive collision particles (masks containing only 3) are set to not collide with each other, thereby avoiding internal interference and optimizing computational resources.

[0009] Furthermore, the construction of the gesture mapping-based guidewire tip fine control module is specifically achieved through guidewire tip steering and guidewire tip traction: Guidewire tip rotation: Real-time monitoring and calculation of the vertical direction of the thumb and index finger pads ( y The displacement difference on the axis is used as a control signal and mapped to the displacement of the guidewire tip particle on the horizontal plane. x shaft and y The velocity vector on the axis. Through this mapping relationship, the user can precisely control the movement direction of the guidewire tip by twisting gestures, thereby selecting the path at the bifurcation of the blood vessel; Guidewire tip traction: A base advance speed along its axis is applied to the guidewire tip particles. This speed value is adaptively adjusted according to the anatomical curvature of the target blood vessel. During the intervention, the system calculates the collision state between the guidewire and the blood vessel wall in real time. When it detects obstruction at a large bend, the speed can be dynamically fine-tuned within a preset range to help the guidewire overcome resistance and advance smoothly, while avoiding penetration of the blood vessel wall due to excessive speed.

[0010] Furthermore, by integrating the aforementioned modules, high-fidelity control simulation of the interaction between the virtual hand model and the guidewire is achieved. At the same time, during surgical training, real-time gesture interaction and visual feedback are completed through the head-mounted display device, thus constructing a complete, usable, and user-friendly virtual reality vascular interventional surgery training system. Gesture Interaction: Direct contact gesture interaction is adopted. Users can directly grasp, move and rotate virtual surgical instrument models through virtual hand models, and directly click user interface buttons that are floating in space through virtual fingers. The interaction method is consistent with the operation in the real world, reducing the learning cost and cognitive load of users in the virtual environment. Visual feedback: The system provides instant and clear visual feedback for all interactive operations to confirm the operation status and guide the user. When the virtual finger touches the interactive button, the button will show dynamic color changes and be accompanied by a slight pressing animation to simulate the effect of a real button being pressed, and provide status prompt feedback.

[0011] Compared with the prior art, the beneficial effects of the technical solution of the present invention are: (1) This invention enables more precise interaction between the virtual hand model and ultra-long, ultra-thin, and ultra-flexible guidewires. It overcomes the limitations of interaction between the virtual hand and traditional rigid bodies, demonstrating the effective control of the flexible guidewire's motion by the hand's adhering and colliding particles, and the precision of the interaction control with the guidewire. The guidewire's advance distance in the blood vessel remains consistent with the hand's advance distance in three-dimensional space at the same time, achieving a stable interaction effect between the hand model and the guidewire. This solves the interaction problem with ultra-long and ultra-flexible guidewires, enabling the virtual hand to grasp the flexible guidewire model.

[0012] (2) Enables ultra-long, ultra-flexible virtual guidewire models to pass through highly curved blood vessels. This invention analyzes the interactive movements of the hand during the twisting process, constructs a multi-layered virtual hand model, and maps the finger displacement difference to the control signal of the guidewire, realizing the steering and traction design of the guidewire tip. This solves the problem that the guidewire is difficult to pass through highly curved blood vessels and realizes high-precision vascular interventional surgery training of the hand in a virtual surgical environment. Attached Figure Description

[0013] Figure 1 This is a schematic diagram of the human-computer interaction method for virtual reality vascular interventional surgery according to the present invention; Figure 2 This is a schematic diagram of the virtual reality vascular interventional surgery human-computer interaction system module of the present invention; Figure 3 This is a schematic diagram illustrating the principle of the extended enclosing layer technology of the present invention; Figure 4 This is a schematic diagram of the adhesion collision particle mechanics model of the present invention; Figure 5 This is a schematic diagram illustrating the principle of the collision filtering mechanism of the present invention. Detailed Implementation

[0014] To further understand the invention's content, features, and effects, the following embodiments are provided, along with the accompanying drawings, for further detailed explanation. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments described below can be combined with each other as long as they do not conflict with each other.

[0015] Please see Figures 1-5 This invention protects a human-computer interaction method and system for vascular interventional surgery based on virtual reality: 1. Construct a deformation simulation and stable collision environment module for ultra-fine guidewires. This module consists of a physical simulation module for ultra-fine guidewires and a stable collision environment module, which realizes high-fidelity simulation of guidewires and establishes a stable, non-penetrating blood vessel interaction environment. (1) Physical simulation of ultra-fine guidewire: A discretized Cosserat elastic rod model is used in combination with the quaternion method for constraint, and the solution is discretized in the position dynamics framework to realize the deformation simulation of the guidewire model, and to efficiently and accurately simulate its bending and torsional behavior in actual surgery: Cosserat elastic rod model: Based on the Cosserat elastic rod theory, the mechanical behavior of the guidewire is accurately modeled using parameters. s The parameter represents the continuity of the elastic rod from its starting point to its ending point. s =0, parameter s =1 corresponds to the start and end points of the elastic rod, respectively, and at each point a vector consisting of three orthogonal basis vectors is attached. d 1, d 2, d 3] constitutes a control frame, in which d 2 and d 3 represents the normal vector and tangent vector of the particle, respectively. d 1 is the product of the two. The Darboux vector is used to describe the rate of change of the frame along the arc length, as shown in formula (1). The bending and torsional strain of the curve can be obtained by projecting the Darboux vector onto the coordinates of the control points, as shown in formula (2). Quaternion method: Using unit quaternions to represent the rotation of the control frame ensures the smoothness and physical correctness of the rotation state when the guidewire undergoes a wide range of bending. The Darboux vector of the control frame can be parameterized as a direction quaternion function. , Represents the rotation quaternion in the current state, where This represents the scalar part, which is related to the rotation angle of the continuous rod model. x , y , z ) represents the vector part, and the strain between the current state and the resting state can be expressed as formula (3); Position dynamics framework: Position dynamics is used as the core solution framework for guide wire deformation simulation, which can efficiently handle large deformations and complex constraints, and discretize the continuous guide wire model into particles. P n Furthermore, properties such as elasticity and instretchability are expressed as constraint functions relating to the particle's position. C ( P =0. In each simulation time step, the particle position is directly corrected by iteratively solving the constraint projection equation. The strain constraint of the Cosserat elastic rod is integrated into the position dynamics framework. Specifically, it is achieved by discretized rod elements. The discretized form based on the rod element is obtained by arithmetic mean interpolation from adjacent quaternions, i.e., formula (4). (2) Stable collision environment: Axis-aligned bounding boxes are used for coarse collision detection. Spatial hashing algorithm is used to filter out possible intersecting objects more quickly. Then, continuous collision detection technology is applied to improve the efficiency and accuracy of detection. This avoids the failure of traditional discrete collision detection methods caused by the small diameter of the guidewire, high relative speed and simulation time step during rapid intervention. A stable and non-penetrating blood vessel interaction environment is established. Spatial hashing algorithm: Construct axis-aligned bounding boxes for all initial objects. The bounding boxes are hashed to be assigned to each cubic cell for filtering. Then, the spatial hashing algorithm is used to assign the box indices to a hash index array. By traversing this hash array, collision detection is performed on each element with the same index, and collision detection is not performed on elements with different indices, thereby reducing the computational complexity of the collision process and avoiding unnecessary computation time. Continuous collision detection: For potential collision pairs selected through wide phase filtering, the state of the objects at the beginning and end of the time step is detected, and the states of both are calculated over the entire Δ time step using linear or nonlinear interpolation. t The movement path within the space fundamentally avoids the "tunneling effect" caused by excessive displacement or high speed of an object, which is the problem of an object moving instantaneously from one side of another object to the other without triggering a collision. Extended bounding layer technique: Constructing a geometrically offset layer in the normal direction of the extremely thin blood vessel model surface, such as... Figure 3 As shown in the figure, the collision state without the extended confinement layer, the extended confinement layer in the direction of the outer normal of the blood vessel wall, and the non-collision state with the extended confinement layer are illustrated. In this extended confinement layer, all collision detection is performed with the guidewire particles, so that even if there are calculation errors due to the discrete time step when passing through narrow bends at high speed, the extended layer can reliably capture them, thereby avoiding the "tunneling effect". This design can further improve the stability and robustness of this scheme compared with the traditional discrete collision detection method.

[0016] 2. Create a multi-layered virtual hand model, which consists of three layers: a skeletal layer, a deformable soft body layer, and an adhesive collision particle layer, providing a structural foundation for mechanical interaction: (1) Skeletal layer: A hand mesh model is constructed using 3D modeling software, and it is associated with a predefined skeletal structure through bone weight binding technology, serving as the basic skeleton and visual representation to drive the entire hand movement; (2) Deformable soft layer: The soft layer located on the fingertip enables the biological soft tissue deformation that occurs when the fingertip contacts an object. Based on the principle of positional dynamics: the deformable fingertip is discretized into a set of particles, and the physical deformation characteristics of the model are expressed as constraint functions. C ( u Solving for the condition using the constraint projection step.C ( u Particle position correction Δ = 0 u It directly updates particle positions, achieving efficient and stable deformation simulation: Distance constraint: For adjacent particles in the fingertip model, the constraint function is defined as the difference between the current distance and the initial distance. Through the projection of this constraint, the particles can recover the initial spacing after deformation, thereby simulating the elastic properties of the fingertip tissue. Volume constraint: For fingertip tetrahedral mesh elements, the constraint function is defined as the difference between the current volume and the initial volume. Through the projection of this constraint, it is ensured that each tetrahedron maintains a substantially constant volume during deformation, thus realistically simulating the incompressibility of biological soft tissue. (3) Adhesive collision particle layer: The collision particle system with adhesive properties is bound to the fingertip bones of the thumb and index finger so as to perform mechanical interaction with the guide wire; 3. Construct an interactive mechanics module between the virtual hand and the guidewire, establish an interactive mechanics model between adhesive collision particles and guidewire particles, and achieve stable grasping and release of the guidewire by the virtual hand through collision filtering optimization calculations: Adhesive Collision Particle Mechanical Model: A set of collision particles with adhesive properties are bound to the fingertip bones of the virtual thumb and index finger. A contact mechanical model is established between these particles and the particles on the guide wire surface, including tangential forces. f t Axial force f z The combined effect of these factors determines the adhesion and fixation of the hand particles on the guidewire surface, as shown in the mechanical decomposition diagram below. Figure 4 The image shows the force vector arrows, including tangential forces, when the virtual hand is bound to adhesive collision particles and the guide wire interacts with the adhesive collision particles. f t Axial force f z ,pressure f n Gravity G, Adhesion force f The mechanical decomposition diagram is the physical basis for the virtual hand to simulate the fine movements of real fingers, such as "twisting" and "pushing," and the adhesive force of the collision particles in the hand. f The tangential and axial forces are expressed by formulas (5) and (6); Collision filtering: Separate collision categories are assigned to deformable fingertip soft bodies, adhesive collision particles, and guide wire models, such as... Figure 5As shown, the categories are 1, 2, and 3. The collision mask rules are configured as follows: guide wires (masks containing 1 and 2) can collide with both fingertip soft objects (category 1) and adherent collision particles (category 2); while fingertip soft objects (masks containing only 3) and adherent collision particles (masks containing only 3) are set to not collide with each other, thus avoiding internal interference and optimizing computational resources.

[0017] 4. Construct a fine-tuning module for the guidewire tip based on gesture mapping, which achieves this through guidewire tip steering and guidewire tip traction: Guidewire tip rotation: Real-time monitoring and calculation of the vertical direction of the thumb and index finger pads ( y The displacement difference on the axis is used as a control signal and mapped to the displacement of the guidewire tip particle on the horizontal plane. x shaft and y The velocity vector on the axis. Through this mapping relationship, the user can precisely control the movement direction of the guidewire tip by twisting gestures, thereby selecting the path at the bifurcation of the blood vessel; Guidewire tip traction: A base advance speed along its axis is applied to the guidewire tip particles. This speed value is adaptively adjusted according to the anatomical curvature of the target blood vessel. During the intervention, the system calculates the collision state between the guidewire and the blood vessel wall in real time. When it detects obstruction at a large bend, the speed can be dynamically fine-tuned within a preset range to help the guidewire overcome resistance and advance smoothly, while avoiding penetration of the blood vessel wall due to excessive speed.

[0018] 5. Integrate the aforementioned modules, such as... Figure 2 As shown in the system module diagram, the system consists of a physical simulation layer comprised of a virtual hand model, a guidewire deformation simulation unit, and a collision detection unit, achieving high-fidelity simulation of the surgical environment. A collaborative control layer, composed of a virtual hand interaction unit and a guidewire tip control unit, maps user gestures to mechanical interactions and path manipulation of the flexible guidewire. Finally, a rendering and feedback unit completes real-time 3D rendering and visual feedback, constructing a complete, usable, and user-friendly virtual reality vascular interventional surgery training system. Gesture Interaction: Direct contact gesture interaction is adopted. Users can directly grasp, move and rotate virtual surgical instrument models through virtual hand models, and directly click user interface buttons that are floating in space through virtual fingers. The interaction method is consistent with the operation in the real world, reducing the learning cost and cognitive load of users in the virtual environment. Visual feedback: The system provides instant and clear visual feedback for all interactive operations to confirm the operation status and guide the user. When the virtual finger touches the interactive button, the button will show dynamic color changes and be accompanied by a slight pressing animation to simulate the effect of a real button being pressed, and provide status prompt feedback.

[0019] The present invention has been shown and described in detail above with reference to the accompanying drawings and preferred embodiments. However, the present invention is not limited to these disclosed embodiments. Based on the above multiple embodiments, those skilled in the art will know that more embodiments of the present invention can be obtained by combining the code review methods in the different embodiments above. These embodiments are also within the protection scope of the present invention.

Claims

1. A human-computer interaction method for vascular interventional surgery based on virtual reality, characterized in that, Includes the following steps: Step S1: Construct a guidewire deformation model and an interaction environment with a non-penetrating blood vessel. The discretized Cosserat elastic rod model is used in combination with quaternions for constraint, and the deformation of the ultra-fine guidewire is solved within the position dynamics framework. An extended bounding layer is set on the surface of the blood vessel model, and a stable collision environment is constructed to prevent the guidewire from penetrating the blood vessel wall by combining spatial hashing algorithm and continuous collision detection technology. Step S2: Create a multi-layered virtual hand model containing a bone layer, a deformable soft body layer, and an adhesive collision particle layer; Step S3: By using the adhesive collision particles set on the fingertips of the virtual hand's thumb and index finger, mechanical interaction is performed with the particles on the guide wire surface, and combined with the collision filtering mechanism, the virtual hand can stably grasp and release the guide wire. Step S4: Real-time detection of the twisting motion of the thumb and index finger, mapping the displacement difference between the two fingers into the steering control command of the guidewire tip, and combining it with adaptive traction based on vascular curvature to achieve precise control of the guidewire travel path. Step S5: Integrate the above modules to provide users with real-time gesture interaction and visual feedback through a head-mounted display device.

2. The method according to claim 1, characterized in that, The extended bounding layer, spatial hashing algorithm, and continuous collision detection technology work together to eliminate the "tunneling effect" caused by high-speed guidewire movement or excessively large simulation time steps.

3. The method according to claim 1, characterized in that, In step S3, the adhesion force of the adhesive collision particles is determined by the contact pressure, contact area, and adhesion coefficient.

4. The method according to claim 1, characterized in that, In step S3, the collision filtering mechanism is implemented by assigning different collision categories to the guide wire, deformable soft layer and adhesive collision particles and configuring collision masks to avoid invalid collision detection between internal components of the virtual hand.

5. A human-computer interaction system for vascular interventional surgery based on virtual reality, characterized in that, include: Module S1: Guidewire deformation simulation unit, used to calculate guidewire deformation based on discretized Cosserat elastic rod combined with quaternion model; Module S2: Collision detection unit, used to determine the collision between the guidewire and the blood vessel within the extended bounding layer through spatial hashing and continuous collision detection; Module S3: Virtual hand interaction unit, used to drive a multi-layered virtual hand and force couple with guide wire particles through adhesive collision particles; Module S4: Guidewire tip control unit, used to acquire the rotational displacement of the thumb and forefinger and convert it into commands for the speed and direction of the guidewire tip, and cooperate with curvature adaptive traction to enable the guidewire to pass smoothly through the greater curvature of the blood vessel; Module S5: Rendering and Feedback Unit, used to output real-time visual feedback to the head-mounted display device.

Images