On The Vascular Modeling And Collision Detection Algorithms For Minimally Invasive Virtual Surgery

Minimally invasive virtual vascular surgery simulation system is a combination of computer graphics, modern medicine and virtual simulation technology. For many clinical doctors and interns,it could reproduce a real scenarios by simulating minimally invasive vascular surgery. This method is not only improve the user’s operation skills, but greatly reduce the operation error at the same time. On the basis of security, minimally invasive surgery system should not only provide enough reality for the procedure, at the same time, should also enable doctors to control the movement of the tool naturally and instinctively,and make full use of its inherent feeling and reaction ability, exert its skilled operation skills. This requires a system with high sense of immediacy. In order to solve the problems of the low verisimilitude of simulation, the imperfection of the behavioral model and contradictory problem between real-time performance and accuracy. In this article, I studies the key technology such as the 3D model reconstruction in minimally invasive virtual vascular surgery simulation system, the optimization of geometric model, physical deformation model simulation, real-time collision detection based on the abdominal aorta vascular as the research object. The main content is summarized as follows:(1) Building the 3D geometry of Blood vessels is the foundation of minimally invasive virtual vascular surgery system. For the problems of the unbalanced angiographic image contrast and the fuzzy vessels’boundaries, the article using the Materialise’s Interactive Medical Image Control System (Mimics) to do such operations on the CT data of vessels, such as data segmentation, editing and 3D model restruction, meanwhile, optimal and test the model. Through the Mimics, adopt the mean of combining the automatic segmentation and manual segmentation to segment the vascular tissue. And through the cavity filling to solve the problem of lacking the information of blood vessels mask that caused by angiography image contrast unbalanced, at the same time, through editing the mask to do boundary enhancement on blood vessel and denoising.(2) In view of the complex tissue blood vessel walls, this paper adopts viscoelastic finite element model based on tetrahedral grid to simulate the deformation of blood vessels. Due to the finite element model requires a lot of calculation in the process of deformation, in the time of ensuring the precision of the abdominal aorta vascular deformation model, combined with real vascular deformation data, this paper mention that blood vessels can be divided into linear elastic deformation and nonlinear deformation part. In the mean time, this paper using Mimics software to analysis and optimize the tetrahedral mesh of finite element, to obtain high quality of the grid.(3) Finite element model of the blood vessels is very complex, leading to a relatively large calculation and calculation time of the collision detection. In order to reduce the test time, improve the detection accuracy, the paper propose a mixed double Bounding Box collision detection algorithm based on the axial Bounding Box (dynamic Axis Bounding Box, AABB) and Discrete direction Bounding Box (Discrete ’Orientation Polytopes, K-DOPs). This method can be divided into two phases includes coarse detection and accurate detection, the upper (outer) adopts the AABB rapid discharge geometrical sets which collision detection may not happen, the lower (inner) adopts the K-DOPs accurately test the geometry sets that the collision may happen and detect the collision point.In this article, through simulating scenes of blood vessels contact with a scalpel in the process of minimally invasive virtual vascular blood vessels surgery, test vascular deformation model and collision detection. The experience show that 3D reconstruction and model optimization by Mimics, not only improve the precision of reconstructed model, also greatly improve the speed of modeling. Improved vascular deformation model and the body grid optimized by Mimics, not only reduces calculation of finite element solution of deformation, at the same time accurately describes the deformation model of blood vessels. The AABB-K-DOPs mix bounding box is effectively reducing the amount of calculation, improving collision detection rate. This lay the foundation of establishing high sense of immediacy minimally invasive virtual surgery system has laid a solid foundation.