Motion-Robust ASL Perfusion Imaging with Self-Navigation Using 3D GRASE and CAIPI Sampling

Motivation: Perfusion imaging using Arterial Spin Labeling (ASL) relies on accurate alignment between tag and control images, making it particularly vulnerable to head motion. Goal(s): To correct the inter-shot motion, minimize motion-induced artefacts and enhance perfusion image quality without increasing scan time. Approach: Using pCASL with a segmented 3D GRASE readout and CAIPI sampling, we employ a 3D self-navigator to estimate inter-shot motion. Motion-corrected images are generated through motion-compensated SENSE reconstruction. Results: Motion-corrected perfusion images show reduced artifacts, higher temporal SNR within gray matter, and improved correlation and SSIM between motion-free and motion-corrected images. Impact: We investigate a motion-robust pCASL sequence using a segmented 3D GRASE readout with CAIPI sampling, improving perfusion image quality in the presence of inter-shot motion without increasing scan time. This approach facilitates future clinical applicationwith less cooperative patients.