Jiachen Zhuo1,2, Jonathan Simon2, Rao Gullapalli1
1Radiology, University of Maryland School of Medicine, Baltimore, MD, United States; 2Electrical & Computer Engineering, University of Maryland College Park, College Park, MD, United States
Processing for DKI reconstruction originally used nonlinear least squares methods which were inefficient. Faster DKI reconstruction using linear methods have been proposed as a compromise when real time reconstruction is needed. We investigated fitting accuracy of both the linear and nonlinear methods on a clinically relevant short imaging protocol (2 b-values) and compared the generated DTI and DKI parameters to the gold standard (5 b-values). Our results suggest that linear fit for DKI reconstruction with a tensor first approach provides comparable accuracy to nonlinear fit, hence could be the method of choice for faster and more efficient DKI reconstruction.