The effect of changes to MOLLI scheme on T1 mapping and extra cellular volume calculation in healthy volunteers with 3 tesla cardiovascular magnetic resonance imaging

Adam K. McDiarmid, David A. Broadbent, David M. Higgins, Peter P. Swoboda, Ananth Kidambi, David P. Ripley, Bara Erhayiem, Tarique A. Musa, Laura E. Dobson, John P. Greenwood, Sven Plein


Background: Diffuse myocardial fibrosis may be quantified with magnetic resonance (MR) by calculating extracellular volume (ECV) fraction from native and post-contrast T1 values. The ideal modified look-locker inversion recovery (MOLLI) sequence for deriving T1 values has not been determined. This study aims to establish if systematic differences exist between suggested MOLLI schemes.
Materials and methods: Twelve phantom gels were studied with inversion recovery spin echo MR at 3.0 tesla to determine reference T1. Gels were then scanned with six MOLLI sequences (3s)3b(3s)5b; 4b(3s)3b(3s)2b; 5b(3s)3b with flip angles of both 35° and 50° at a range of heart rates (HRs). In 10 healthy volunteers MOLLI studies were performed on two separate occasions. Mid ventricular native and post contrast T1 was measured and ECV (%) calculated.
Results: In phantoms, the co-efficient of variability at simulated HR [40-100] with a flip angle of 35° ranged from 6.77 to 9.55, and at 50° from 7.71 to 11.10. T1 was under-estimated by all MOLLI acquisitions. Error was greatest with longer T1, and increased as HR increased. The 10 volunteers had normal MR studies. Native T1 time was similar for all acquisitions but highest with the 5b(3s)3b 35° scheme (1,189.1±33.46 ms). Interstudy reproducibility was similar for all MOLLIs.
Conclusions: The 5b(3s)3b MOLLI scheme agreed best with reference T1, without statistical difference between the six schemes. The shorter breath-hold time of 5b(3s)3b scheme may be preferable in clinical studies and warrants further investigation.