12 May 2004 Artifact reduction in mutual-information-based CT-MR image registration
Author Affiliations +
Abstract
Abstract Mutual information (MI) is currently the most popular match metric in handling the registration problem for multi modality images. However, interpolation artifacts impose deteriorating effects to the accuracy and robustness of MI-based methods. This paper analyzes the generation mechanism of the artifacts inherent in linear partial volume interpolation (PVI) and shows that the mutual information resulted from PVI is a convex function within each voxel grid. We conclude that the generation of the artifacts is due to two facts: 1) linear interpolation causes the histogram bin values to change at a synchronized pace; 2) entropy computation function Σxlgx is convex. As a remedy we propose to use non-uniform interpolation functions as the interpolation kernels in estimating the joint histogram. Cubic B-splin and Gaussian interpolators are compared and we demonstrate the improvements via experiments on misalignments between CT/MR brain scans.
© (2004) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Mingxiu Wei, Mingxiu Wei, Jundong Liu, Jundong Liu, Junhong Liu, Junhong Liu, "Artifact reduction in mutual-information-based CT-MR image registration", Proc. SPIE 5370, Medical Imaging 2004: Image Processing, (12 May 2004); doi: 10.1117/12.536349; https://doi.org/10.1117/12.536349
PROCEEDINGS
11 PAGES


SHARE
Back to Top