Retrospective Motion Correction Algorithms for In Vivo Micro-MRI

University Of Pennsylvania, Philadelphia PA

Investigators

Abstract

0302251 Song Involuntary subject motion often impairs image quality in MRI. The problem is exacerbated in high-resolution imaging due to the long scan times and small voxel sizes. Displacements during data collection even on a sub-millimeter scale can often cause significant degradation of image quality and can prevent the images from being analyzed quantitatively. In order for in vivo micro-MRI to become practical, in which the resolution is in the submillimeter range (100 - 300 um), a robust motion compensation scheme is necessary. The principal objective of the proposed research is to develop improved motion correction techniques for in vivo MR micro-imaging. The work is based on a retrospective data correction technique referred to as "autofocusing," which is an iterative technique that does not require additional data to compensate for motion. Widespread application has been hampered, however, by long data processing times and suboptimal performance, particularly when image signal-to-noise ratio is low. The goal of this project is to develop efficient autofocusing algorithms to detect and compensate for both translational and rotational motion, and to improve its current performance, with particular focus on high-resolution images that are often corrupted by noise. Although the focus of the proposed work is on in vivo micro-MRI, the anticipated result may have broader impact by improving image sharpness and thus achievable spatial resolution not only in high-resolution images, but in any MR imaging application in which subject motion causes image degradation.