Temporal registration: a new approach to manage the incomplete recovery of the longitudinal magnetization in the Modified Look-Locker Inversion Recovery sequence (MOLLI) for T1 mapping of the heart.

ABSTRACT

PURPOSE: To correct with post-processing effects of incomplete recovery of the longitudinal magnetization before a new inversion pulse in the Modified Look-Locker Inversion recovery sequence (MOLLI) sequence. THEORY AND METHODS: We model such effects as a temporal shift ([Formula see text]) of the signal of the Look-Locker block following next inversion pulses. After using the following equation [Formula see text], a temporal registration of [Formula see text] is applied to the signal of the affected block to adjust the sampling time of the recovery signal and correct the underlying effect on quantitative T1. To test our approach, simulations, phantoms, and five volunteers' data were used while applying different MOLLI sampling schemes at different heart rates and compared to the reference three-parameter fit. RESULTS: The temporal registration of the affected signals allows to reach higher accuracy on long T1 when compared to the reference three parameters fit (10.15 vs 22.12% for T1 = 1785 ms; 8.22 vs 14.65% for T1 = 1278 ms), and lower average variation in case of rest-period deletion (62 vs 231 ms). CONCLUSION: The proposed approach leads to more accurate T1 in case of incomplete recovery. It is less sensitive to parameters affecting the recovery such as the rest period or the sampling scheme; and, therefore, supports multi-center studies with different MOLLI protocols.