Oral Abstract Session
SCMR 22nd Annual Scientific Sessions
Background: Conventional myocardial T1 mapping techniques such as MOLLI are limited to single T1 map acquisition per breathhold1. Multi-slice or full ventricular coverage may be beneficial for significant variations in left ventricular wall thickness and/or fibrosis2-3, thus resulting in multiple breathholds, increased patient discomfort and prolonged scan time. In this work, we sought to develop, characterize and demonstrate clinical feasibility of a novel FASt single-breathhold multi-slice myocardial T1 mapping (FAST1) sequence at 1.5T.
(1) Sequence and reconstruction: As described in Fig.1, the FAST1 sequence enables the acquisition of five T1 maps of five different slices in one breathhold. A slice-selective inversion pulse is applied and followed by two ECG-triggered single-shots of images of the same slice. This imaging block is repeated five times for different slices within the same breathhold. T1 map reconstruction used a dictionary search with a normalized single-parameter model combined with scaling, inversion factor and heart rate corrections.
(2) Experimental evaluation: All imaging was performed on a 1.5T Siemens Aera scanner. FAST1 was compared to MOLLI in phantom, 9 healthy volunteers for native myocardial T1 mapping, and 17 consecutive patients for both native and post-contrast (13 among them) T1 mapping. Both sequences used the same bSSFP imaging parameters: TR/TE/FA 2.7ms/1.1ms/35°, FOV 360×306mm2, voxel size 1.4×2.1mm2, slice thickness 8mm, GRAPPA factor 2, partial Fourier factor 7/8, bandwidth 1085Hz/px, shortest TI 100ms. 15 contiguous slices providing full left ventricular coverage were acquired in 3 separated breathholds using FAST1, while 3 slices were acquired in 3 additional breathholds using MOLLI. This protocol was performed five times in phantom4, twice for each healthy volunteer and once for each patient.
Results: In phantom, FAST1 and MOLLI yielded relative T1 errors of 8.7±1.5% vs. 6.6±0.9%, precision of 7.5±4.4ms vs. 5.6±3.5ms and repeatability of 1.4±0.6ms vs. 0.8±0.3ms, respectively. Over all 9 healthy volunteers, FAST1 yielded slightly lower T1 measures (952±22ms vs. 988±23ms, p<0.0001), decreased precision by a factor of 1.2 (57±8ms vs. 47±7ms, p=0.0046) and similar repeatability (16±5ms vs. 14±5ms, p=0.43) compared to MOLLI (Fig.2). FAST1 and MOLLI T1 times were highly linearly correlated in patients (>0.92) (Fig.3).
Conclusion: FAST1 enables myocardial T1 mapping with full ventricular coverage in three breathholds with limited precision penalty, highly linearly correlated T1 times, similar repeatability and five-fold increased slice coverage compared to MOLLI.