X-ray computed tomography cone-beam analytic reconstruction using GPU

Abstract

Computed tomography (CT) is medical imaging technology to acquire 3D data. Since it was invented 1967, CT has experienced tremendous growth over the years. Even though many other modalities have appeared CT is still competitive. Compared to other medical imaging methods, the strength of CT is imaging acquisition time. However, achieving exact CT reconstruction image remains a challenge. Due to the complexity of the operation, real-time reconstruction is difficult. Also it requires too much computation cost. To solve this issue, we are motivated to adapt the GPU (Graphic Processing Unit) which is well developed for parallel processing. NVIDA has developed CUDA (Computer Unified Device Architecture) which allows user to use GPU as a CPU (Central Processing Unit) resource. In this study, our goal is to validate the effectiveness of GPU implementation using CUDA in a CT exact cone-beam reconstruction method.