Catherine D. G. Hines1, Alex P.
Frydrychowicz2,3, Dana L. Tudorascu4, Gavin Hamilton5,
Karl K. Vigen2, Huanzhou Yu6, Charles A. McKenzie7,
Claude B. Sirlin5, Jean H. Brittain8, Scott B. Reeder1,2
1Biomedical Engineering, University of
Wisconsin-Madison, Madison, WI, United States; 2Radiology,
University of Wisconsin-Madison, Madison, WI, United States; 3Diagnostic
Radiology and Medical Physics, University Hospital Freiburg, Freiburg,
Germany; 4Waisman Laboratory for Brain Imaging and Behavior,
University of Wisconsin-Madison, Madison, WI, United States; 5Radiology,
University of California-San Diego, San Diego, CA, United States; 6Applied
Science Laboratory, GE Healthcare, Menlo Park, CA, United States; 7Medical
Biophysics, University of Western Ontario, London, Ontario, Canada; 8Applied
Science Laboratory, GE Healthcare, Madison, WI, United States
Accurate
quantification of hepatic steatosis is essential for early detection of
non-alcoholic fatty liver disease, which is increasingly common in Western
societies. Quantitative IDEAL provides a means to measure hepatic steatosis
in vivo, although its precision and accuracy are unknown. 40 patients were scanned twice using both
quantitative IDEAL and MRS to assess accuracy and precision. Analysis of Bland-Altman plots, concordance
correlation coefficients, linear regression and confidence intervals indicate
that quantitative IDEAL provides both highly accurate and precise
fat-fractions using MRS as a reference and is a reliable method of in vivo
fat quantification.