27 March 2009 New GPU optimizations for intensity-based registration
Author Affiliations +
Proceedings Volume 7259, Medical Imaging 2009: Image Processing; 72592O (2009) https://doi.org/10.1117/12.812455
Event: SPIE Medical Imaging, 2009, Lake Buena Vista (Orlando Area), Florida, United States
The task of registering 3D medical images is very computationally expensive. With CPU-based implementations of registration algorithms it is typical to use various approximations, such as subsampling, to maintain reasonable computation times. This may however result in suboptimal alignments. With the constant increase of capabilities and performances of GPUs (Graphics Processing Unit), these highly vectorized processors have become a viable alternative to CPUs for image related computation tasks. This paper describes new strategies to implement on GPU the computation of image similarity metrics for intensity-based registration, using in particular the latest features of NVIDIA's GeForce 8 architecture and the Cg language. Our experimental results show that the computations are many times faster. In this paper, several GPU implementations of two image similarity criteria for both intramodal and multi-modal registration have been compared. In particular, we propose a new efficient and flexible solution based on the geometry shader.
© (2009) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Razik Yousfi, Razik Yousfi, Guillaume Bousquet, Guillaume Bousquet, Christophe Chefd'hotel, Christophe Chefd'hotel, } "New GPU optimizations for intensity-based registration", Proc. SPIE 7259, Medical Imaging 2009: Image Processing, 72592O (27 March 2009); doi: 10.1117/12.812455; https://doi.org/10.1117/12.812455

Back to Top