Abstract
Nonrigid image registration is an important, but resource demanding and time-consuming task in medical image analysis. This limits its application in time-critical clinical routines. In this report we explore acceleration of two time-consuming parts of a registration algorithm by means of parallel processing using the GPU. We built upon the OpenCL-based GPU image processing framework of the recent ITK4 release, and implemented Gaussian multi-resolution strategies and a general resampling framework. We evaluated the performance gain on two multi-core machines with NVidia GPUs, and compared to an existing ITK4 CPU implementation. A speedup factor of ~2-4 was realized for the multi-resolution strategies and a speedup factor of ~10-46 was achieved for resampling, for larger images (~10^8 voxels).