Meeting Banner
Abstract #2566

A framework for intensity-based affine registration of multiparametric prostate MRI via mutual information and genetic algorithms

Ethan Leng1, David Porter2, Andrew Larson1, Xiaoxuan He1, Benjamin Spilseth3, and Gregory J. Metzger1

1Center for Magnetic Resonance Research, University of Minnesota, Minneapolis, MN, United States, 2Minnesota Supercomputing Institute, University of Minnesota, Minneapolis, MN, United States, 3Department of Radiology, University of Minnesota, Minneapolis, MN, United States

An image registration framework was developed to perform 3D, affine, intensity-based co registration of multiparametric MRI series using mutual information as the similarity metric. The proposed methods include corrections to compensate for the effects of an endorectal coil, which is commonly used in prostate MRI. Experiments to characterize the registration method demonstrate that it is theoretically accurate to within 1.0 mm (when estimating the translation component). Qualitatively, significant improvements are seen in the co-localization of parametric maps with the anatomic images. The proposed framework may readily be integrated into a CAD system for prostate cancer detection.

This abstract and the presentation materials are available to members only; a login is required.

Join Here