Meeting Banner
Abstract #0423

Increasing the Resolution of Diffusion-Weighted MRI with Distortion Compensated Orthogonal Acquisitions & Super-Resolution Reconstruction

MAGNA25Benoit Scherrer1, Ali Gholipour1, Simon K. Warfield1

1Radiology, Harvard Medical School, Boston, MA, United States

Increasing the spatial resolution in DW-imaging requires sampling of higher frequencies in k-space which is very challenging with a single shot EPI acquisition. We propose to reduce the spatial encoding burden by employing distortion-compensated orthogonal anisotropic acquisitions, and by achieving super-resolution reconstruction (SRR) of the underlying high-resolution image. We demonstrate that our approach provides better results than acquisition of a single isotropic scan for the same acquisition duration time. This work provides for the first time evidence that SRR, which employs conventional single-shot EPI techniques, may enable resolution enhancement in DWI, and may dramatically impact the way to achieve DW-imaging.

Keywords

achieve achieved achieving acquisition acquisitions additionally aligned aligning anatomy anisotropic appear applying assessment axes axial better beyond blurred bootstrapping brain burden capsule challenging channel chosen clinical coil coils color combining comparable compensate compensated compensation conserved consistently coordinate corpus corrected correction correspond corruption decreases demanding dense descent describes detailed diffusion distortion dual duration effectiveness employing enable enables encoding enhancement ensure estimation evidence evidences except experience exploit field filtering finally fine fines frontal geometric gradient head highly homogeneity impact impacts importantly increasing instead interpolation investigate isotropic larger leads limited location lost making maps material matrix model noise noisy orientation orthogonal outperforms particularly patient poor practice precise prematurely priors problem promising qualitatively radiology real recently reconstruct reconstruction reduce reducing regularized remains reports represent requires resolution respectively sampling scanner school seeding shot simulations slice smoothed space spatial stem stopped streamlines strongly structures subject super tracts trio ultimately uncertainty unclear underlying unprecedented variance variations volume zoom