Impact of hepatic steatosis and MOLLI acquisition parameters on liver T1 estimates

To evaluate the influence of hepatic fat content and sequence acquisition parameters on Modified Look-Locker Inversion Recovery (MOLLI)-based liver T1 estimates, comparing balanced steady-state free precession (bSSFP) and spoiled gradient recalled echo (SPGR) readouts under in-phase (IP), opposed-phase (OP), and fat-suppressed (FS) conditions.

Twenty participants (5 controls, 11 with hepatic steatosis, 4 with other chronic liver diseases; median age 15 years, range 10-28) underwent abdominal MRI at 1.5T. Six single-slice MOLLI acquisitions (bSSFP and SPGR; IP, OP, FS) were performed to obtain hepatic T1 maps. A 3D multi-echo confounder-corrected chemical shift encoded (mDIXON Quant^®) sequence was performed to obtain proton density fat fraction (PDFF), and T2* maps. ROI-based measurements were made by three blinded readers. Agreement and mean difference among MOLLI protocols were assessed using intraclass correlation coefficients (ICC), Bland-Altman analysis, and repeated-measures ANOVA. Correlations between PDFF and T1 differences were evaluated using weighted linear regression. Numerical simulations were implemented to model fat-related effects on MOLLI T1 estimates.

Liver PDFF ranged from 1 to 43% (median 4.6%). Agreement between bSSFP and SPGR T1 estimates was excellent (ICC ≥ 0.90), though SPGR estimated higher T1 values than bSSFP by 30-60 ms, with greater variability. T1-OP differed significantly from T1-IP and T1-FS (both p < 0.0001) and showed significant PDFF-related mean difference, in line with simulation results, while T1-FS and T1-IP estimates showed excellent agreement (ICC ≥ 0.89). Regression analysis showed that bSSFP T1-OP correlated very strongly with PDFF (r = 0.81; +7.8 ms/%PDFF), whereas T1-IP showed a strong negative correlation (r = - 0.79; - 5.0 ms/%PDFF), and T1-FS demonstrated only a moderate negative correlation (r=-0.48; -3.1ms/%PDFF). SPGR based T1 estimates exhibited similar correlations with PDFF. In participants without steatosis, T1 values were consistent across all acquisitions.

Liver MOLLI T1 estimates are strongly influenced by hepatic fat content and acquisition parameters, particularly for OP acquisitions. IP and FS MOLLI acquisitions provide more reliable estimates across fat fractions. These findings support prioritization of IP or FS strategies in clinical and research liver T1 mapping protocols and underscore the need for fat-corrected or water-specific approaches.