Study Of Soft Tissue Deformation In Virtual Surgery

Virtual surgery is a comprehensive application that amalgamates virtual reality with modern biomedicine.It utilizes computer technology to reproduce multiperceived surgical scenes and achieve medical education,surgical training and guidance.It is of great significance to help doctors improve surgical skills,enhance the overall level of the medical industry,and promote medical advancement.In virtual surgery,the soft tissue deformation caused by frequent interaction between surgical instruments and virtual soft tissue,directly influences the authenticity of the simulation system.In order to realistically simulate visual and tactile interactions,this thesis has studied the following aspects.First of all,for the sake of obtaining a three-dimensional model of human soft tissue,we have reconstructed it based on medical CT images triumphed the conversion of CT images from 2D images to 3D model.Moreover,in order to obtain a threedimensional model with internal nodes,we use a tetrahedral subdivision algorithm to discretize the three-dimensional grid model and construct the virtual soft tissue model with physical properties.Then,only virtual soft tissues exhibit more gradual and more accurate deformation behaviors and interactive operations can be more realistic.The large deformation becomes difficult to recover and the deformation area is not smooth are disadvantages of the traditional Mass Spring Model(MSM).Therefore,this thesis proposes a method of elasticity compensation mass spring to determine the corresponding compensation elastic force by calculating the initial and real-time spatial position between the mass point and the adjacent mass point.Further,this method also avoids the super-elasticity of the spring by pulling the mass from its current position back to its original position.In addition,this proposed method improves the problem of insufficient recovery and distortion of the traditional particle spring model.We verified the effectiveness of the new method through experiments.It improves the problem of insufficient recovery and distortion of the traditional particle spring model.Finally,collision detection is a prerequisite for deformation and feedback calculations.We propose a vector-based collision detection algorithm to achieve a fine and accurate collision between a virtual instrument and a soft tissue model during surgical interaction.We transform the interactive behavior between surgical instruments and soft tissue collisions into that between surgical instrument tip and soft tissue surface.The position of the ball in the virtual space and a vertex of the triangle constitute a vector.The angle between this vector and the normal vector of the triangle can determine whether the collision occurs.This method has a small amount of calculation.Its efficiency is more than twice as fast as the continuous collision detection algorithm,and its implementation simple.It is suitable for high-precision and slow-speed surgical scenes.The experimental results show that the method can quickly and accurately detect the touch behavior in simulated surgery.