Meeting Banner
Abstract #4043

Clinical Comparison of Single-R2* and Dual-R2* Correction for Accurate Fat Quantification in the Liver

MAGNA25Debra E. Horng1, 2, Diego Hernando1, Catherine D.G. Hines3, Scott B. Reeder1, 2

1Radiology, University of Wisconsin-Madison, Madison, WI, United States; 2Medical Physics, University of Wisconsin-Madison, Madison, WI, United States; 3Merck Research Laboratories, West Point, PA, United States

We compared the performance of fat-fraction quantification using chemical shift encoded MRI using single-R2* and dual-R2* correction methods, using MR spectroscopy (MRS) as the reference standard. From a group of 97 patients, 20 patients with fat fraction > 10% were identified, and chemical shift encoded fat-water imaging was performed. Fat-fraction was measured from the imaging data using both single-R2* and dual-R2* correction. Fat-fraction estimates were compared to co-registered spectroscopy measurements: single-R2* correction demonstrated higher agreement. Single-R2* modeling is more accurate than dual-R2* modeling for hepatic fat quantification in patients, even in those with high hepatic fat concentrations.

Abstract PDF E-Poster Video

Keywords

accelerator accuracy accurate accurately acknowledge acquisition additional agreement approval assess assuming avoid axial better bias chemical clinical closely common complexity consent correction correlation decay determine deviation displayed dramatically dual echoes eddy encoded errors estimation exhibited fatty fits fitted fitting foundation fraction generated good gratefully identified impact independent informed intercept intracellular iron laboratories leads limitation linear linearly liver localized magnitude maps mathematical matrix measure measured medical methylene mimic minimal model models need noise none noticeably originating overall overload particle particles patient patients peak peaks performance phantom phantoms physics physiologically plots plotted poor post processing program quantification radiology ranging rather reeder regression requires respectively separating setting severe similarly simulated simulation slope spectroscopy studies subjects subsequently subtracted support theoretically true truth typical underlying unknown unstable vacuoles warrants water west whether whose worse worst