Ray-driven 기반 고속 프로젝션 영상 생성을 위한 GPU 적용 기법 연구

We study GPU-based parallel processing methods for computing ray driven projection images. We implemented three parallel processing methods; parallelization based on detector bin, parallelization based on projection angle using global memory and parallelization based on projection angle using shared memory. We report the computation time for the three methods using NVDIA TITAN X.