Murat Aksoy1, Samantha Holdsworth1, Stefan Skare1, Roland Bammer1
1Department of Radiology, Stanford University, Stanford, CA, USA
Due to its prolonged acquisition time, the correction of motion artifacts in high resolution DTI is essential for acceptable image quality. Short-Axis PROPELLER-EPI (SAP-EPI) has been proven to be very effective in eliminating phase and motion artifacts as well as geometric distortions. However, gross patient motion has two effects on the acquired data: pixel misregistration and change in diffusion encoding direction. While the pixel misregistration can be addressed by coregistration of low-resolution images, the change in diffusion encoding direction makes it incorrect to combine different blades to get the diffusion weighted images. In this study, we addressed this issue by combining SAP-EPI with the novel non-linear tensor estimation scheme that estimates the diffusion tensors from the complex k-space data directly. The results show an increased accuracy of main eigenvector orientation compared to the conventional schemes for tensor estimation.