Parallel Mutual Information Based 3 D Non-Rigid Registration on a Multi-Core Platform

For many clinical applications, non-rigid registration of medical images demands cost-effective high performance solutions in order to be usable. We present a parallel design and implementation of a wellknown algorithm based on mutual information and a B-Spline transformation model for the Cell Broadband Engine(Cell/B.E.) platform. It exploits the different levels of parallelism offered by this heterogeneous multi-core processor and scales almost linearly on a system containing two chips or a total of 16 accelerator cores. A speedup of more than 40x was reached compared to a sequential implementation and registration time for a 512x512x100 voxel image was 60 seconds.

[1]  Michael Unser,et al.  Fast parametric elastic image registration , 2003, IEEE Trans. Image Process..

[2]  Daniel Rueckert,et al.  Nonrigid registration using free-form deformations: application to breast MR images , 1999, IEEE Transactions on Medical Imaging.

[3]  Jason N. Dale,et al.  Cell Broadband Engine Architecture and its first implementation - A performance view , 2007, IBM J. Res. Dev..

[4]  O. Dandekar,et al.  Hardware Implementation of Hierarchical Volume Subdivision-based Elastic Registration , 2006, 2006 International Conference of the IEEE Engineering in Medicine and Biology Society.

[5]  M. Unser,et al.  Spline Pyramids for Inter-Modal Image Registration Using Mutual Information , 1997 .

[6]  H. Peter Hofstee,et al.  Introduction to the Cell multiprocessor , 2005, IBM J. Res. Dev..

[7]  Michael Unser,et al.  Unwarping of unidirectionally distorted EPI images , 2000, IEEE Transactions on Medical Imaging.

[8]  Ashwini K. Nanda,et al.  Cell/B.E. blades: Building blocks for scalable, real-time, interactive, and digital media servers , 2007, IBM J. Res. Dev..

[9]  Guy Marchal,et al.  Multimodality image registration by maximization of mutual information , 1997, IEEE Transactions on Medical Imaging.

[10]  David R. Haynor,et al.  PET-CT image registration in the chest using free-form deformations , 2003, IEEE Transactions on Medical Imaging.

[11]  Torsten Rohlfing,et al.  Nonrigid image registration in shared-memory multiprocessor environments with application to brains, breasts, and bees , 2003, IEEE Transactions on Information Technology in Biomedicine.

[12]  Christos A. Papachristou,et al.  A "brick" caching scheme for 3D medical imaging , 2004, 2004 2nd IEEE International Symposium on Biomedical Imaging: Nano to Macro (IEEE Cat No. 04EX821).

[13]  Michael Gschwind.  The Cell Broadband Engine: Exploiting Multiple Levels of Parallelism in a Chip Multiprocessor , 2007, International Journal of Parallel Programming.

[14]  Paul A. Viola,et al.  Alignment by Maximization of Mutual Information , 1997, International Journal of Computer Vision.

[15]  Hiroshi Inoue,et al.  REAL-TIME MUTUAL-INFORMATION-BASED LINEAR REGISTRATION ON THE CELL BROADBAND ENGINE PROCESSOR , 2007, 2007 4th IEEE International Symposium on Biomedical Imaging: From Nano to Macro.

[16]  R. Kikinis,et al.  Toward Real-Time Image Guided Neurosurgery Using Distributed and Grid Computing , 2006, ACM/IEEE SC 2006 Conference (SC'06).