Study On Medical Image Volume Rendering Technology Based On VTK

Medical Image 3D Reconstruction and Visualization is a multi-disciplinary subject. Now, this technology, one of the most important hotspots in the field of medical imaging, is not only important to the theory research but also valuable for clinic diagnosing.This thesis focuses on the direct volume rendering algorithms, especially the ray casting algorithm which is the most important and common one among direct volume rendering algorithms. To solve the existing problems of traditional ray casting algorithm, including high computational complexity and large number of empty voxels, which result in slow rendering velocity, a hybrid approach of accelerating traditional ray casting algorithm is proposed. In this thesis, some important knowledge about Medical Image 3D Reconstruction is introduced, including characteristics of common medical images and pretreatment methods, different methods of Medical Imaging Visualization, their advantages and disadvantages and their accelerating algorithms. Then principles and details of both medical image 3D reconstruction and direct volume rendering are described and analyzed respectively and systematically. Finally, the ray casting algorithm is studied deeply and the existing problems of traditional ray casting algorithm, including high computational complexity and large number of empty voxels are pointed out. So, a new hybrid approach of accelerating traditional ray casting algorithm is proposed. Simulation results demonstrate that this new approach has achieved noticeably better performance than traditional ray casting algorithm with moderately decreased image quality.