Native myocardial T1 mapping using rapid 1.5T T1FLASH and T1 MOLLI during breath-hold and free-breathing

Abstract

Quantitative myocardial T1 mapping is an established technique in cardiovascular magnetic resonance (CMR). However, conventional T1 mapping requires relatively long breath-holds (BH), which may lead to breathing artifacts and reduced image quality in patients with limited BH capacity. The T1 fast low angle shot (T1FLASH) sequence enables myocardial T1 mapping with shorter BH durations. In this prospective study, 14 healthy adult volunteers (8 female; mean age 27 ± 4 years) underwent myocardial T1 mapping at 1.5 T using modified Look-Locker inversion recovery with motion correction (T1 MOLLI MOCO) and rapid T1FLASH. Both sequences were acquired during BH and free breathing (FB). Global and segmental myocardial T1 values and image quality were assessed. Interrater, intrarater, and interstudy reproducibility were evaluated. Comfort was assessed using a questionnaire. Statistical analysis included the Shapiro-Wilk test, paired t-test, Pearson correlation, Bland-Altman analysis, intraclass correlation coefficient, Wilcoxon signed-rank test, and Spearman rank correlation (p < 0.05). BH duration was significantly shorter with rapid T1FLASH compared with T1 MOLLI MOCO (3–4 s vs. 12–14 s), and participant comfort was higher with rapid T1FLASH (p < 0.05). During BH, global myocardial T1 values were significantly higher with rapid T1FLASH (1030 ± 35 ms) than with T1 MOLLI MOCO (1000 ± 26 ms; p < 0.01). The slightly increased variability observed with rapid T1FLASH remained below diagnostically relevant thresholds (4% error margin). Both techniques showed moderate-to-good correlation for global T1 values (r = 0.63; p < 0.01). Rapid T1FLASH demonstrated good image quality that was comparable for clinical interpretation during BH, although T1 MOLLI MOCO received slightly higher qualitative image scores. During FB, variability of global and segmental T1 measurements increased significantly, with apical segments being more and AHA segments 2 (basal anteroseptal), 5 (basal inferolateral), 8 (mid anteroseptal), 11 (mid inferolateral) less affected by respiratory artifacts. Interrater, intrarater, and interstudy reproducibility were good to excellent. Rapid T1FLASH demonstrated good agreement with conventional T1 MOLLI MOCO and reproducible myocardial T1 measurements, with substantially reduced BH duration while maintaining good image quality. These findings support further evaluation of rapid T1FLASH as an alternative T1 mapping approach in patients with limited BH capacity. Conventional T1 MOLLI MOCO remains the preferred technique when adequate BH can be achieved.