Study Of Soft Tissue Deformation Simulation In Virtual Surgery

With the rapid development of modern science and technology, computer applications have come to pervade every aspect of social life. Virtual surgery is an application of virtual reality in modern medicine, and it presents real reappearance of the operations to users through vision and force feedback. It can be used in surgical training, surgery prediction and image guided surgery.Nowadays most data used in virtual surgery systems are medical data collected by medical appliances such as CT, MRI, PET and type-B ultrasonic, or the Virtual Human Data. The models reconstructed by the data perform well in accuracy and color, but it is very hard for the users to use these models in real-time surgery systems. To solve the real-time problem, simplifying the models is proved to be an efficient method, but the reduction process would also remove some key details on soft tissue and degrades performance in virtual surgery systems. With the purpose of performing in real-time, it is common in virtual surgery system to simulate soft tissue with linear elastic models. However, soft tissues are rendered with obvious viscoelasticity when large deformations happen. Linear elastic models are not applicable in such settings. Nowadays, in research of soft tissue simulation in virtual surgery, the main contradiction is between similarity and real time.To resolve this contradiction, we introduced the physics engine—PhysX. Using PhysX enables a realistic simulation. And physics engine does most of the work of the physical computing to achieve real-time simulation. On this basis, we use more complex geometric model and physical model in order to increase the performance of system simulation.On the FEM mesh automatic generation, this paper presents a tetrahedral mesh generation algorithm. This method not only generates high-performance finite element models which can be used for real-time virtual surgery systems, but also changes the traditional methods, which generate model from medical data obtained from medical equipment(such as CT, MRI, PET, B super or X-ray machine).Models generated by this method can overcome the contradiction between accuracy and real-time. As a result of the introduction of the PhysX, we can use more complex FEM to improve the performance of simulation of soft tissue deformation.On the simulation of tissue deformation, this paper presents the Viscoelastic FEM Deformation Simulation Method. This method overcomes the low accuracy and poor stability of the Mass-Spring Model, and takes the biomechanical properties of soft tissue fully into account. By adjusting parameters, it can simulate tissues with different kinds of mechanical property. Using this method, we can simplify the virtual surgery system, and optimize the real-time performance of the system.At the same time, as a result of accelerated physics advantages of PhysX, we can complement real-time simulation of soft tissue and blood and fluid interaction, which was hard to complement in traditional virtual surgery system.