Research On Hybrid Soft Tissue Modeling And Collision Detection Based On RPIM-IXPBD In Virtual Surgery Simulation

Virtual surgery,as one of the typical applications of virtual reality technology,uses computer graphics to establish virtual scenes and combines interactive devices to train trainee surgeons in surgery,which can effectively improve the surgical level of trainee surgeons,which has an important role in transforming traditional surgical training methods.In the surgery process,the doctor operates the surgical instruments to touch soft tissue and cause soft tissue to interact with deformation response,which is the most basic operation process when the doctor performs surgery.How to establish an effective soft tissue model and collision detection algorithm to realize the real-time and accurate interactive operation of virtual surgery simulation,which is one of the key and difficult points of virtual surgery simulation.Therefore,the following research work is carried out in this paper for virtual soft tissue modeling and collision detection.Firstly,we propose a hybrid soft tissue modeling method based on the Wilsontheta implicit integration method with radial base point interpolation(RPIM)and improved extended positional dynamics(IXPBD).The method dynamically divides the whole model into surgical and non-surgical regions.In the surgical region,the proposed method uses RPIM based on the Wilson-theta implicit integration method for soft tissue modeling,aiming to provide a realistic simulation of soft tissue deformation.In the non-surgical region,since accurate deformation simulation is not required,we propose an IXPBD modeling method with faster computational speed,aiming to reduce the computational cost of the whole model and reduce the amount of model computation at interaction.Note that the proposed IXPBD method incorporates soft tissue biomechanical properties into extended positional dynamics method.Compared with the conventional extended position dynamics method,the proposed IXPBD method has a higher degree of realism.In addition,to avoid the fragmentation phenomenon between the surgical and non-surgical regions,we established a smooth transition method based on Newtonian mechanics between the two regions.The experimental results show that the effectiveness of the hybrid modeling method.Secondly,we propose a hybrid dynamic interpolation collision detection algorithm.The method uses spatial hashing and hybrid bounding box continuous collision detection in the coarse detection stage,which can quickly extract the set of primitives where collisions may occur.In the fine detection stage,the dynamic interpolation collision detection avoids easy occurrence of missed detection or puncture during the interaction.The collision detection of virtual surgery scenes with a large number of primitives and fast deformation speed is achieved by this staged approach.The experimental results show that our method avoids the phenomenon of missed detection or puncture when surgical instrument interact with soft tissue model,and also has a high real-time performance.Finally,based on the above method,we built a virtual surgical liver deformation interaction simulation system.The system contains a force feedback module,a visual rendering module and a dynamic interaction calculation module,which are used to simulate touch deformation operation of real liver surgery.The experiments of deformation interaction simulation were conducted in the system.The experimental results show that the system has good deformation interaction effect and high realtime performance,which meets the practical training requirements of the virtual surgery system.