3D Registration Based on Normalized Mutual Information: Performance of CPU vs. GPU Implementation
Interactive Graphics Systems Group (GRIS), TU Darmstadt, Germany
Deserno, T.M.: Bildverarbeitung fur die Medizin 2010: Algorithmen – Systeme – Anwendungen; Proceedings des Workshops vom 14. bis 16. Marz 2010 in Aachen. Berlin: Springer, 2010. (Informatik aktuell), pp. 325-329
@article{jung20103d,
title={3D Registration Based on Normalized Mutual Information: Performance of CPU vs. GPU Implementation},
author={Jung, F. and Wesarg, S.},
journal={Bildverarbeitung f{\”u}r die Medizin], Deserno, T. and Hoffmann, J., eds., Informatik aktuell, 5, Springer},
year={2010}
}
Medical image registration is time-consuming but can be sped up employing parallel processing on the GPU. Normalized mutual information (NMI) is a well performing similarity measure for performing multi-modal registration. We present CUDA based solutions for computing NMI on the GPU and compare the results obtained by rigidly registering multi-modal data sets with a CPU based implementation. Our tests with RIRE data sets show a speed-up of factor 5 to 7 for our best GPU implementation.
February 3, 2011 by hgpu
