The Design And Implementation Of 3D Visualization Of Plug And Imaging Based On Digital PET Image

The digital positron emission tomography(PET)detection platform based on Plug and Imaging(PnI)was a novel application-adaptive PET system development platform based on all-digital PET technology and core devices.It could help researchers build new PET systems quickly and conveniently.With the flexible hardwork/software framewok,it was possible for users to build PET systems with various geometic configurations according to actual needs.The PnI-based digital PET composed of two parts: hardware and software.The hardware part could be randomly combined by the modularized digital PET detector,and the software part could be formed by selecting the corresponding module on the general software platform.Currently PnI software platform adopted the traditional two-dimensional method for image analysis,which depends on the users’ knowledge of anatomy and their clinical capability.It could not only differentiate the lesion from the surrounding tissues,also could not carry on the aid diagnosis and the surgery planning.With the development of computer graphics,3D visualization technology had emerged.Through 3D visualization technology,the position of region of interest and the surrounding tissues could be clearly detected,which was of great significance to the diagnosis and surgical planning of diseases.At present,PnI users could only use third-party software for 3D visualization,which brought inconvenience to them.Users urgently needed to realize 3D visualization through PnI’s software platform 3D visualization module and conduct corresponding research.To solve the above problems,in this thesis,I implemented the 3D visualization module of digital PET images based on PnI the optimization and improvement of the existing 3D visualization algorithm.The main works were summarized as follows:(1)This paper studied the basic process and methods of 3D visualization,emphasising on marching cube algorithm and contour connection algorithm in surface rendering and ray casting algorithm in volume rendering.In order to further improve the 3D visualization rendering speed,the method of reducing triangles to improve marching cube algorithm and increasing the sampling spacing to accelerate the ray casting algorithm.The experiment showed that reducing 30% triangles and 0.5～3 mm sampling spacing can shorten the rendering time while ensuring the rendering effect.(2)In order to solve the problems of slow rendering speed and fuzzy interaction in traditional ray casting algorithm,the parallelism of graphics processing unit(GPU)was used to accelerate ray casting algorithm.Compared with traditional ray casting algorithm,the ray casting algorithm based on GPU could significantly improve rendering efficiency and improve the fuzzy problem in the process of interaction.(3)Because the region of interest was not in one plane,the multi-plane reconstruction technology was studied,and the multi-plane reconstruction function of digital PET image based on PnI was realized.User could observe the region of interest from different planes according to their needs.(4)In order to meet the requirements of users quickly visualization,this paper compared the marching cube algorithm and contour connection algorithm in surface rendering,the ray casting algorithm and ray casting based on GPU algorithm from three aspects of drawing time,drawing effect and interaction speed.Considering all factors,this paper selected the contour connection algorithm in surface rendering and the ray casting based on GPU algorithm in volume rendering to realize 3D visualization.(5)The 3D visualization module of digital PET image based on PnI was realized by combining 3D visualization tools VTK and C++.Users could realize the functions of reading and displaying fault sequence images,marking regions of interest,surface rendering,volume rendering,and multi-plane reconstruction through this module.The implementation of the 3D visualization of digital PET based on PnI was a supplement to the software platform,which greatly met the needs of users and had great value in the application of auxiliary diagnosis,and laid a foundation for the further function expansion and follow-up research of the platform.