Hepatic MR imaging for in vivo differentiation of steatosis, iron deposition and combined storage disorder: Single-ratio in/opposed phase analysis vs. dual-ratio Dixon discrimination

Objective: To assess whether in vivo dual-ratio Dixon discrimination can improve detection of diffuse liver disease, specifically steatosis, iron deposition and combined disease over traditional single-ratio in/opposed phase analysis. Methods: Seventy-one patients with biopsy-proven (17.7 +/- 17.0 days) hepatic steatosis (n = 16), iron deposition (n = 11), combined deposition (n = 3) and neither disease (n = 41) underwent MR examinations. Dual-echo in/opposed-phase MR with Dixon water/fat reconstructions were acquired. Analysis consisted of: (a) single-ratio hepatic region-of-interest (ROI)-based assessment of in/opposed ratios; (b) dual-ratio hepatic ROI assessment of in/opposed and fat/water ratios; (c) computer-aided dual-ratio assessment evaluating all hepatic voxels. Disease-specific thresholds were determined; statistical analyses assessed disease-dependent voxel ratios, based on single-ratio (a) and dual-ratio (b and c) techniques. Results: Single-ratio discrimination succeeded in identifying iron deposition (I/O-Iron threshold <0.88) and steatosis (I/O-Fat threshold>1.15) from normal parenchyma, sensitivity 70.0%; it failed to detect combined disease. Dual-ratio discrimination succeeded in identifying abnormal hepatic parenchyma (F/W-Normal threshold >0.05), sensitivity 96.7%; logarithmic functions for iron deposition (I/O-Iron discriminator <e((0.01-F/WIron)/0.48)) and for steatosis (I/O-Fat discriminator >e((F/WFat-0.01)/0.48)) differentiated combined from isolated diseases, sensitivity 100.0%; computer-aided dual-ratio analysis was comparably sensitive but less specific, 90.2% vs. 97.6%. Conclusion: MR two-point-Dixon imaging using dual-ratio post-processing based on in/opposed and fat/water ratios improved in vivo detection of hepatic steatosis, iron deposition, and combined storage disease beyond traditional in/opposed analysis. (C) 2011 Elsevier Ireland Ltd. All rights reserved.