Meeting Banner
Abstract #0401

Fat and Iron Quantification Using a Multi-Step Adaptive Fitting Approach with Multi-Echo GRE Imaging

Xiaodong Zhong1, Marcel D. Nickel2, Stephan A.R. Kannengiesser2, Brian M. Dale3, Berthold Kiefer2, Mustafa Bashir4

1MR R&D Collaborations, Siemens Healthcare, Atlanta, GA, United States; 2MR Applications Development, Siemens AG, Healthcare Sector, Erlangen, Germany; 3MR R&D Collaborations, Siemens Healthcare, Cary, NC, United States; 4Division of Abdominal Imaging, Duke University Medical Center, Durham, NC, United States

In this work, a multi-step adaptive fitting approach was developed for fat and iron quantification using multi-echo 3D GRE data, which accounts for various factors such as T2* decay, T1 bias, multi-peak fat modeling, and noise bias. Numeric phantoms were created and used to validate the results measured by this approach with the ground truth, and showed that this approach is relatively insensitive to different field strengths, field inhomogeneity, monopolar/bipolar readout, and TE selections. An in vivo patient study showed consistency between the FP results measured with the proposed approach and a spectroscopy-based method.

Abstract PDF Presentation Video

Keywords

abdominal abstracts accordance accounts accurately acquisition adaptive additionally alternatively ambiguity among applications approved audience bias bipolar blue bottom chart chemical circular clinical collaborations complex conditions considered contrast converge convex correct corrected critical cubic custom dale decay defined depicted deposition described develop development deviation directly division dual duke echoes errors examination example excel field finally fitting flow generates gradually great green ground health hepatic hold implemented in vivo included indicates indicators informative inhomogeneity initial insensitive institutes investigated iron known lipid liver local magnitude major maps measured measuring mixing model modeling must negligible nine noise nonlinear numeric optimization overcome patient peak percentage performance phantom phantoms physicists pixel position post priori program proposed quantification radiologists radiology rapid readers readout real reconstruction refer regarding registered relatively remain reporting respectively scripts selection separately sets seven simulated simulation since situation solutions sources spectroscopy speed square step strengths system target truth uniform utilizing validate validated valuable variations vessels vibe water