Simultaneous proton density fat-fraction and imaging with water-specific T₁ mapping (PROFIT₁ ): application in liver

To describe and validate a simultaneous proton density fat-fraction (PDFF) imaging and water-specific T₁ mapping (T_1(Water) ) approach for the liver (PROFIT₁ ) with $$ mapping and low sensitivity to $$ calibration or inhomogeneity.

A multiecho gradient-echo sequence, with and without saturation preparation, was designed for simultaneous imaging of liver PDFF, $$ , and T_1(Water) (three slices in ~13 seconds). Chemical-shift-encoded MRI processing yielded fat-water separated images and $$ maps. T_1(Water)  calculation utilized saturation and nonsaturation-recovery water-separated images. Several variable flip angle schemes across k-space (increasing flip angles in sequential RF pulses) were evaluated for minimization of T₁ weighting, to reduce the $$ dependence of T_1(Water)  and PDFF (reduced flip angle dependence). T_1(Water)  accuracy was validated in mixed fat-water phantoms, with various PDFF and T₁ values (3T). In vivo application was illustrated in five volunteers and five patients with nonalcoholic fatty liver disease (PDFF, T_1(Water) , $$ ).

A sin³ (θ) flip angle pattern (0 < θ < π/2 over k-space) yielded the largest PROFIT₁ signal yield with negligible $$ dependence for both T_1(Water) and PDFF. Mixed fat-water phantom experiments illustrated excellent agreement between PROFIT₁ and gold-standard spectroscopic evaluation of PDFF and T_1(Water)  (<1% T₁ error). In vivo PDFF, T_1(Water) , and $$ maps illustrated independence of the PROFIT₁ values from $$ inhomogeneity and significant differences between volunteers and patients with nonalcoholic fatty liver disease for T_1(Water) (927 ± 56 ms vs. 1033 ± 23 ms; P < .05) and PDFF (2.0% ± 0.8% vs. 13.4% ± 5.0%, P < .05).  $$ was similar between groups.

The PROFIT₁ pulse sequence provides fast simultaneous quantification of PDFF, T_1(Water) , and $$ with minimal sensitivity to $$ miscalibration or inhomogeneity.