| This research focuses on a preoperative planning system for minimally invasive vascular interventional surgery based on virtual reality,and a virtual reality imaging structure based on polarized light stereo vision is built.Additionally,an algorithm for extracting blood vessel centerline based on CT angiography image is proposed,and a cost map-based vascular intervention path planning algorithm is proposed.This overall work aims to provide doctors with intuitive and scientific image guidance,improving the efficiency and quality of vascular interventional surgery planning.The main contents and results are as follows.First,an algorithm for extracting blood vessel centerline based on CT angiography images is proposed.By analyzing the eigenvectors of angiography-enhanced angiography images,the vascular structure images are enhanced to improve the reconstruction quality of the vascular three-dimensional model.The existing maximum inscribed sphere algorithm is improved by introducing virtual points in the centerline extraction process,which simplifies the section extraction process.At the same time,post-processing such as offset inspection and correction is performed on the initially extracted centerline.A smooth and continuous centerline path model is obtained,which provides a basis for path planning.Second,a cost map-based vascular intervention path planning algorithm is proposed.Based on the analytic hierarchy process,the weights of each index affecting the interventional operation in the vascular structure are quantitatively analyzed.The cost map is encoded by the reconstructed vascular centerline model,which shows the risk degree of each point in the vascular intervention path in a quantitative manner.Based on the cost map,a path search algorithm is used to obtain the optimal path for vascular intervention.By superimposing the 3D model of the blood vessel with the cost map,the structural parameters of the blood vessel can be visually presented to guide the surgical planning.Finally,a reflective marker tracking algorithm based on binocular vision is proposed,which provides stable and efficient space registration interaction support for virtual reality systems.Combined with the space registration algorithm,the virtual reality display of 3D models will be completed.A virtual reality vascular intervention path planning system is built to view the path planning results.The planning results is presented three-dimensionally in the virtual reality environment,so that doctors have a more intuitive understanding of the complex vascular structure.The system provides surgical planning intuitive and effective guidance.In this paper,the virtual reality surgical planning system can quantify the risk of interventional paths according to complex vascular structural parameters,and provide interventional path recommendations,which improves the scientificity and credibility of vascular path planning.The virtual reality system intuitively presents the planning results of the superimposed cost map,which can provide doctors with clear and intuitive surgical planning assistance.The sysytem has a strong potential clinical application value. |