Hui Xue1, Saurabh Shah2, Andreas
Greiser3, Christoph Guetter1, Christophe Chefdhotel1,
Marie-Pierre Jolly1, Sven Zuehlsdorff2, Jens Guehring1,
Peter Kellman4
1Imaging & Visualization,
Siemens Corporate Research, Princeton, NJ, United States; 2CMR
Research & Development, Siemens Medical Solutions USA, Inc., Chicago, IL,
United States; 3Imaging & IT Division, Siemens AG, Healthcare
Sector, Erlangen, Germany; 4National Heart, Lung & Blood
Institute, National Institutes of Health, Bethesda, MD, United States
The state-of-art technique for cardiac T1 mapping is the modified Look-Locker Inversion Recovery (MOLLI) which acquires multiple images across several heart-beats. Its clinical applicability is often limited by frequent myocardial motion because of imperfect breath-hold or varying R-R interval. A fully automated motion correction directly utilizing MOLLI images is highly challenging due to significantly varying image contrast. We therefore propose a novel registration algorithm based on estimating motion-free synthetic images presenting similar contrast to original MOLLI data by solving a variational energy minimization problem. The validation was performed in vivo on a large cohort of patient datasets.