A fast and efficient parallel implementation for 3D-CBCT image reconstruction

In the framework of nondestructive testing and evaluation (NDT/NDE), this paper deals with the problem of improving the speed of image reconstruction in three-dimensional cone-beam computed tomography (3D-CBCT). Image reconstruction of 3D-CBCT using FDK algorithm requires huge computational resource, due to the voluminous data to handle, and the long processing times. Parallel implementation, on multi-core computer, seems to be a good approach to solve the problem. In this study, we introduce Z-line data first reconstruction algorithm, which is based on the FDK algorithm, and execute it with SIMD and multi-core technology. In order to assess the effectiveness of the proposed approach, an illustrative result from representative numerical test cases is presented and discussed.