Takeshi Yokoo1, Masoud Shiehmorteza1, Mark Bydder1, Gavin Hamilton1, Yuko Kono2, Alexander Kuo2, Joel E. Lavine3, Claude B. Sirlin1
1Radiology, UCSD, San Diego, CA, USA; 2Medicine, UCSD, San Diego, CA, USA; 3Pediatrics, UCSD, San Diego, CA, USA
Several groups recently suggested that accurate fat quantification by MR requires correction of the relaxation effects as well as modeling of multi-component fat signal. Using non-T1-weighted multi-echo GRE imaging and 3-peak fat spectral modeling, high fat estimation accuracy was demonstrated at 1.5 T in fatty liver patients. In principle, this general approach would be independent of the field strength and therefore is also applicable at 3.0 T. In this first human study on spectrally-modeled hepatic fat quantification at 3.0 T, we assessed fat estimation accuracy of MR imaging using MR spectroscopy as the reference standard.