Cardiovascular magnetic resonance imaging has been increasingly applied in clinical examinations [[1], [2], [3], [4]]. In particular, quantitative myocardial T1 mapping has raised significant attention in clinical practice [5]. Alterations of myocardial T1 values are associated with tissue characteristics, and therefore quantitative T1 analysis of myocardium has enabled further insight into cardiomyopathies [6].

Numerous cardiac T1 mapping approaches have been proposed, and each of them provides a different profile of advantages. The modified Look-Locker inversion recovery (MOLLI) sequence is commonly used for T1 mapping of the heart [3,4]. The original MOLLI sequence was carried out with three blocks of serial image acquisitions, each following a non-selective inversion pulse. The three blocks incorporate three, three, and five electrocardiograph (ECG)-triggered images with single-shot readouts to sample the inversion recovery (IR) signals. Two recovery periods of at least three heartbeats are needed between each of the two inversion pulses in order to satisfy the condition of sufficient magnetization recovery [7]. A later variant of MOLLI, MOLLI5(3 s)3 [8], allows sufficient T1 recovery during the five heartbeats of the first IR block plus the three-second waiting period. This acquisition scheme effectively reduces the heart-rate dependency of the initial longitudinal magnetization of the second IR block and thus lowers the heart-rate influences on the obtained T1 map [8]. The MOLLI5(3 s)3 is denoted as MOLLI53 hereafter to simplify the presentation in this article.

In addition to IR approaches, T1 mapping methods based on saturation recovery (SR) were also reported [[9], [10], [11]]. The SR-based T1 mapping methods improve imaging efficiency because they do not require extra waiting time for longitudinal relaxation. For example, saturation recovery single-shot acquisition (SASHA) sequence [9] applying a saturation preparation before each readout exhibited advantages in that the T1 measurement was minimally influenced by the prior residual longitudinal magnetization. Therefore, the SASHA sequence produced high T1 accuracy. However, the SR-based method samples the T1 recovery curve with a reduced dynamic range compared with MOLLI. As a result, SASHA suffered from reduced reproducibility in accessing T1 values due to its low dynamic range.

Numerous approaches for T1 mapping have been proposed using distinct acquisition schemes to sample the T1 relaxation signal [4,9,10,12,13]. There is no current standard instruction for clinical cardiovascular magnetic resonance protocols, although MOLLI53 has been commercialized by at least one manufacturer as a clinical package to provide diagnostic information on a variety of cardiomyopathies. In this study, a hybrid MOLLI (hbMOLLI) method that integrated saturation recovery with the inversion recovery sequence was proposed for quantitative T1 mapping in the myocardium within one single breath-hold. By replacing the second inversion pulse of the MOLLI53 technique with a saturation pulse, the long recovery time could be alleviated in hbMOLLI, thereby either reducing total acquisition time for ease of breath-hold or allowing more images to be sampled from the T1 relaxation curve. Although the combinations of SR with IR for cardiac T1 mapping have been documented in the literature [12,13], the primary goal to shorten breath-hold period in our study differed from these previous methods to some extent, as will be discussed in a later section. Furthermore, due to the elimination of the T1 recovery period, the heart rate dependency of the estimated T1 could be potentially reduced. In this study, we designed a novel hbMOLLI sequence sampling both of the IR and SR recovery curves and implemented a fitting algorithm to minimize the fitting error of both curves simultaneously. Possible impact on T1 fitting precision due to reduced dynamic range from the replacement of IR by SR was investigated via computer simulations. Phantom and in-vivo studies were performed to assess the efficacy of the proposed sequence.