Abstract #2704

Deep learning-based diffusion method alleviates spurious group differences due to head motion

Ting Gong¹, Hongjian He¹, Zhiwei Li², Zhichao Lin², Feng Yu², and Jianhui Zhong^1,3

¹Center for Brain Imaging Science and Technology, Key Laboratory for Biomedical Engineering of Ministry of Education, College of Biomedical Engineering and Instrumental Science, Zhejiang University, Hangzhou, China, ²Department of Instrument Science & Technology, Zhejiang University, Hangzhou, China, ³Department of Imaging Sciences, University of Rochester, Rochester, NY, United States

Head motion occurring during the acquisition of diffusion-weighted (DW) images will cause deterioration in quality of diffusion model reconstruction, which could lead to spurious group differences of DW measures when there is difference in head motion for different groups. We have previously developed a method for robust diffusion kurtosis mapping of motion-contaminated data. In this study, we applied it in a group level, and the results demonstrated its ability in ameliorating spurious group differences due to head motion. The method can be applied to data with different motion level thus improving the utilization and statistic power of some valuable but motion-corrupted DW data.

This abstract and the presentation materials are available to members only; a login is required.

Join Here