Improved image segmentation with prospective motion correction in MRI

Abstract

Artifacts caused by patient motion during a magnetic resonance imaging (MRI) scan can corrupt the image data, reduce the effective resolution and make further processing difficult or impossible. Long measurements, as required for high resolution imaging, are particularly prone to motion artifacts, as motion is more likely to occur over longer scan periods. These artifacts can lead to problems in the further image processing of the data. Prospective motion correction offers a possibility to correct for patient motion during the measurement and avoid motion artifacts. Image data from comparative brain scans with and without motion correction are presented and analyzed. Segmentation of gray matter (GM) and white matter (WM) was performed on both datasets using the software packages SPM and Freesurfer, which are widely used in neuroscience studies. The results show that motion correction significantly improves the segmentation quality and suggest that it is also useful for other image processing applications performed on MRI data. © Springer-Verlag Berlin Heidelberg 2012.