Water-Only Look-Locker Inversion recovery (WOLLI) T1 mapping

PurposeModified Look-Locker Inversion recovery (MOLLI) and Shortened MOLLI (ShMOLLI) T1 values can deviate substantially from the water T1 in voxels containing large amounts of water and fat. We introduce Water-Only Look-Locker Inversion recovery (WOLLI) to map water T1 in tissue containing fat. TheoryWOLLI comprises adiabatic water-selective inversion and several balanced steady-state free precession (bSSFP) readouts, followed immediately by adiabatic fat-selective inversion and bSSFP readout(s). Data are fitted pixel-wise and an adapted Deichmann-Haase saturation-correction gives the water T1. MethodsWe compared two WOLLI protocols (WOLLI-w7f1 and WOLLI-w7f2f1), ShMOLLI and single-voxel spectroscopy at 3T in: simulations, fat/water phantoms, and 12 subjects livers. ResultsIn simulations with 0-40% fat fraction, T1 varied by: 100% (MOLLI). In phantoms, the accuracy of the methods (i.e. worst-case T1w difference vs spectroscopy) was: WOLLI-w7f1 (best, maximum difference of 10.2%), WOLLI-w7f2f1 (max. 10.5%), and ShMOLLI (max. 81.5%). In the liver, the root-mean-squared deviations vs spectroscopy were: 374ms (WOLLI-w7f1), 645ms (WOLLI-w7f2f1), and 700ms (ShMOLLI). ConclusionCompared to ShMOLLI, WOLLI is less influenced by fat, but more by B0-inhomogeneity. WOLLI was a clear improvement in phantoms but less so in vivo. WOLLI may be suited to scan organs with high fat content such as the liver.