Unity3D-based Virtual Liver Surgical Deformation Modeling And Collision Detection Study

Virtual surgery,as an important application of virtual reality technology in modern medicine,is a technology that combines multiple disciplines such as computer graphics and material mechanics,and can build a surgical training scenario that integrates multidimensional perceptions such as vision,hearing and touch for the trainer.Compared with traditional surgical training,it can not only reduce the cost of surgical training,but also avoid the ethical and moral issues in traditional surgical training.An effective virtual surgery simulation system needs to have two major characteristics:real-time interaction and accurate soft tissue deformation simulation,and realistic soft tissue model and real-time accurate collision detection algorithm are the important factors affecting these two characteristics.Therefore,this paper conducts a research related to soft tissue modeling technology and collision detection technology in virtual liver surgery simulation system,and the main research contents are as follows:First,a Biometric Novel Hybrid Model(BNHM)based on biomechanics is proposed in order to establish a realistic and efficient soft tissue model.The model introduces virtual springs,damped springs and nonlinear springs on the basis of the traditional mass-spring model,and better simulates the viscoelastic and nonlinear biological characteristics of real soft tissues by combining multiple spring structures.In addition,a redefined finite element strain energy function is used to accurately describe the relationship between strain and stress within the model,and to effectively model the quasi-incompressibility and anisotropy exhibited by soft tissues when subjected to external forces.The experimental results show that BNHM can effectively simulate the four major biomechanical properties of real soft tissues in real time.Secondly,for the problems of traditional discrete detection algorithm miss detection error detection and traditional continuous collision algorithm computationally costly,this paper proposes a new continuous collision detection algorithm(New Continuous Collision detection,N-CCD)based on GPU accelerated spatial hashing algorithm and hybrid hierarchical wraparound box algorithm.In order to determine the collision information more accurately and quickly,the collision detection is divided into two stages: fast detection and accurate detection.In the fast detection phase,the GPU-accelerated spatial hashing algorithm and the hybrid hierarchical wraparound box algorithm are used to quickly eliminate the disjoint elements,and in the accurate detection phase,the M?ller-Trumbore algorithm is used to detect the intersection of the basic triangular cells after the accurate prediction of the collision information by the improved fold-and-half time search method.The experimental results show that the N-CCD has higher detection efficiency.Finally,in order to verify the practicality of the proposed soft tissue modeling method and collision detection algorithm,a virtual liver surgery simulation system was built in this paper in combination with Unity3 D engine technology,in which the squeezing,clamping and cutting operations on the soft tissue model of liver were realized in the virtual environment.The interactive simulation experiments were completed in the system,and the experimental results verified the effectiveness of BNHM and N-CCD,which can meet the practical needs of the virtual surgery system.