期刊
NEUROIMAGE
卷 53, 期 2, 页码 565-575出版社
ACADEMIC PRESS INC ELSEVIER SCIENCE
DOI: 10.1016/j.neuroimage.2010.06.014
关键词
Diffusion MRI; Diffusion tensor; M-estimation; Robust estimation; Artefact; Physiologic noise; Cardiac pulsation; Head motion; Autism
Motion artefacts are an important but often disregarded problem in diffusion-weighted imaging, which can readily lead to corrupt diffusion model estimations The new processing method proposed in this paper uses robust tensor estimation that is spatially informed to efficiently detect the most frequently occurring artefacts, namely those that result from head and cardiac motion Simulations demonstrate that the method is more robust and accurate than previous methods The tensor estimates are more accurate in motion artefact-free conditions, less sensitive to increases in artefact magnitude and more resistant to increasing artefact frequency. Evaluation with real diffusion-weighted (DW) imaging data shows that the method works excellently, even for datasets with a high degree of motion that otherwise need to be discarded The method is not limited to diffusion tensor imaging but also yields objective artefact reflecting weights that can be used to inform subsequent processing or estimation of higher-order diffusion models (C) 2010 Elsevier Inc All rights reserved
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据