A high-resolution diffusion-weighted magnetic resonance imaging method using projection data regeneration

Abstract

This thesis analyzes the effect of ECG-triggering on the pulsatile motion artifacts in diffusion-weighted imaging (DWI) with radial trajectories and proposes a method to acquire high-resolution DW images using a projection data regeneration method. In general, ECG-triggering is used in order to reduce the cardiac pulsatile motion artifacts in DWI, in which even small motion results in severe artifacts due to the strong diffusion gradients. Cardiac gating reduces the artifacts resulting from the cardiac pulsatile motion to some extent, although the artifacts cannot be completely removed. For this reason, a projection data regeneration method is proposed through this study in order to improve image quality of DWI. In the proposed method, the projection data is acquired by radial spin-echo DWI (rSE-DWI), a temporary image is generated by an inverse Radon transform (IRT). Degraded data in the acquired projection data are then replaced with the regenerated projection data from the Radon transform (RT) of the temporary image. Using similar approach, high-resolution DW images are also estimated from low-resolution DW images using projection data regeneration and projection data composition. In this thesis, the proposed method can reduce the imaging time to nearly 1/2 of the original imaging time. The proposed method shows the image quality improvement and high-resolution data generation by computer simulations, and then by in vivo images acquired using rSE-DWI.