Fast myocardial T(1) mapping using shortened inversion recovery based schemes

BACKGROUND: Myocardial T(1) mapping shows promise for assessment of cardiomyopathies. Most myocardial T(1) mapping techniques, such as modified Look–Locker inversion recovery (MOLLI), generate one T(1) map per breath‐held acquisition (9–17 heartbeats), which prolongs multislice protocols and may be unsuitable for patients with breath‐holding difficulties. PURPOSE: To develop and characterize novel shortened inversion recovery based T(1) mapping schemes of 2–5 heartbeats. STUDY TYPE: Prospective. POPULATION/PHANTOM: Numerical simulations, agarose/NiCl(2) phantom, 16 healthy volunteers, and 24 patients. FIELD STRENGTH/SEQUENCE: 1.5T/MOLLI. ASSESSMENT: All shortened T(1) mapping schemes were characterized and compared with a conventional MOLLI scheme (5‐(3)‐3) in terms of accuracy, precision, spatial variability, and repeatability. STATISTICAL TESTS: Kruskal–Wallis, Wilcoxon rank sum tests, analysis of variance, Student's t‐tests, Bland–Altman analysis, and Pearson correlation analysis. RESULTS: All shortened schemes provided limited T(1) time variations (≤2% for T(1) times ≤1200 msec) and limited penalty of precision (by a factor of ~1.4–1.5) when compared with MOLLI in numerical simulations. In phantom, differences between all schemes in terms of accuracy, spatial variability, and repeatability did not reach statistical significance (P > 0.71). In healthy volunteers, there were no statistically significant differences between all schemes in terms of native T(1) times and repeatability for myocardium (P = 0.21 and P = 0.87, respectively) and blood (P = 0.79 and P = 0.41, respectively). All shortened schemes led to a limited increase of spatial variability for native myocardial T(1) mapping with respect to MOLLI (by a factor of 1.2) (P < 0.0001). In both healthy volunteers and patients, the two‐heartbeat scheme and MOLLI led to highly linearly correlated T(1) times (correlation coefficients ≥0.83). DATA CONCLUSION: The proposed two‐heartbeat T(1) mapping scheme yields a 5‐fold acceleration compared with MOLLI, with highly linearly correlated T(1) times, no significant difference of repeatability, and limited spatial variability penalty at 1.5T. This approach may enable myocardial T(1) mapping in patients with severe breath‐holding difficulties and reduce the examination time of multislice protocols. Level of Evidence: 1 Technical Efficacy Stage: 3 J. Magn. Reson. Imaging 2019;50:641–654.