Title :
MRI-based attenuation correction and medical image registration on GPU
Author :
Cheng, Wei-Hung ; Liu, Yufan ; Lu, Cheng-Chang
Author_Institution :
Comput. Sci. Dept., Kent State Univ., Kent, OH, USA
Abstract :
In this paper, we propose to implement mutual information algorithms with functional data-parallelism to speed up the medical image registration process. Experimental results showed that the GPU implementation improves the registration computational performance with a speedup factor of 18.36×. Additional speedup can be achieved with diligent data profiling. In addition, we suggest the use of the proposed implementation in attenuation correction of MRI/PET image pairs.
Keywords :
biomedical MRI; coprocessors; image registration; medical image processing; parallel processing; positron emission tomography; GPU; MRI based attenuation correction; MRI-PET image pairs; data profiling; functional data parallelism; medical image registration; mutual information algorithms; Graphics processing unit; Image registration; Interpolation; Medical diagnostic imaging; Mutual information; Positron emission tomography;
Conference_Titel :
Biomedical Engineering and Informatics (BMEI), 2011 4th International Conference on
Conference_Location :
Shanghai
Print_ISBN :
978-1-4244-9351-7
DOI :
10.1109/BMEI.2011.6098333
