Implications of bulk motion for diffusion-weighted imaging experiments: Effects, mechanisms, and solutions

This review article describes the effect of bulk motion on diffusion-weighted imaging experiments, and examines methods for correcting the resulting artifacts. The emphasis throughout the article is on two-dimensional imaging of the brain. The effects of translational and rotational motion on the MR signal are described, and the literature concerning pulsatile brain motion is examined. Methods for ameliorating motion effects are divided into three generic categories. The first is methods that should be intrinsically insensitive to macroscopic motion. These include motion-compensated diffusion-weighting schemes, single-shot EPI, projection reconstruction, and line scanning. Of these, only single-shot EPI and projection reconstruction methods can obtain high-quality images without compromising on sensitivity. The second category of methods is those that can be made insensitive to bulk motion. The methods examined here are FLASH and RARE. It is shown that for both sequences motion insensitivity is in general attained only at the cost of a 50% reduction in sensitivity. The final set of methods examined are those that correct for motion, primarily navigator echoes. The properties and limitations of the navigator echo approach are presented, as are those of methods which attempt to correct the acquired data by minimizing image artifacts. The review concludes with a short summary in which the current status of diffusion imaging in the presence of bulk motion is examined.