Rigid motion‐corrected magnetic resonance fingerprinting

摘要: Purpose Develop a method for rigid body motion-corrected magnetic resonance fingerprinting (MRF). Methods MRF has shown some robustness to abrupt motion toward the end of acquisition. Here, we study effects different types during acquisition on and propose novel approach correct this motion. The proposed (MC-MRF) follows 4 steps: (1) sliding window reconstruction is performed produce high-quality auxiliary dynamic images; (2) rotation translation estimated from images by image registration; (3) used acquired k-space data with corresponding rotations phase shifts; (4) are reconstructed low-rank inversion. MC-MRF was validated in standard T1 /T2 phantom 2D vivo brain acquisitions 7 healthy subjects. Additionally, effect through-plane investigated. Results Simulation results show that can introduce artifacts T2 maps, depending when it occurs. improved parametric map quality all experiments in-plane motion, comparable no-motion ground truth. Reduced quality, even after correction, observed particularly smaller structures maps. Conclusion correction proposed, which improves accuracy comparison approaches. Future work will include validation 3D enable also correction.