Design and Implementation of Parallel Nonrigid Image Registration Using Off-the-Shelf Supercomputers

Abstract

This paper presents a new parallel algorithm for nonrigid image registration using off-the-shelf supercomputers, or clusters of PCs. Our algorithm realizes scalable registration for high resolution three-dimensional (3-D) images by employing three techniques: (1) data distribution; (2) data-parallel processing; and (3) dynamic load balancing. The experimental results show that our parallel implementation on a cluster of 64 off-the-shelf PCs (with 128 processors) registers liver CT images of 512×512×159 voxels within 8 minutes while a sequential implementation takes 12 hours. Furthermore, our implementation allows processors to use less memory, and thereby enables us to align 1024×1024×590 voxel images, which is not easy for single processor systems due to the restrictions on the memory space and the processing time.KeywordsExecution TimeHigh Performance ComputingNonrigid RegistrationGradient CalculationDynamic Load BalanceThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.