Time

Prog #

Takeshi Yokoo¹, Mark Bydder¹, Gavin Hamilton¹, Michael S. Middleton¹, Jefferey Schwimmer, Joel Lavine, Terek Hassanein, Heather Patton, Nicholas Pinto¹, Richard Znamirowski¹, Lillian Pacheco¹, Claude B. Sirlin¹

¹University of California at San Diego, San Diego, California , USA

Noninvasive hepatic fat quantification is an increasingly important area of application of MR technology.  Several new MR imaging strategies have been proposed in an attempt to account for the T1 and T2* relaxation effects, and to improve the accuracy and reliability of fat quantification.  In this presentation, we describe results from a prospective clinical study in which low flip-angle multi-echo gradient-echo imaging is used to predict the proton-density fat fraction measured by MR spectroscopy.

Catherine Diane Gard¹, Thomas Frances Warner¹, Huanzhou Yu², Ann Shimakawa², Charles A. McKenzie³, Jean H. Brittain⁴, Scott B. Reeder¹

¹University of Wisconsin-Madison, Madison, Wisconsin, USA; ²GE Healthcare, Menlo Park, California , USA; ³University of Western Ontario, London, Canada; ⁴GE Healthcare, Madison, Wisconsin, USA

Early diagnosis of hepatic steatosis with non-invasive quantitative biomarkers is a critical unmet need in the detection and evaluation of non-alcoholic fatty liver disease.  In this work we validate fat quantification of hepatic steatosis in a range of tissue fat percentages in the ob/ob mouse model of hepatic steatosis using IDEAL (Iterative Decomposition with Echo Asymmetry and Least-squares estimation).  We compare fat-fraction measured with IDEAL to histological evaluation and total lipid extraction using the Folch method. Excellent correlation between IDEAL, histology and lipid extraction is observed. These results demonstrate validation of MRI for quantification of hepatic steatosis in vivo.

Maria I. Altbach¹, Zhiqiang Li¹, Scott W. Squire¹, Arthur F. Gmitro¹, Ali Bilgin¹, Christian Graff¹, Eric Outwater¹

¹University of Arizona, Tucson, Arizona , USA

A radial Gradient and Spin Echo (GRASE) method has been developed for liver imaging.  The method allows for the quantitative characterization of liver pathologies  based on T2 values and lipid-water information.

Huanzhou Yu¹, Scott B. Reeder², Ann Shimakawa¹, Charles A. McKenzie³, Jean H. Brittain⁴

¹GE Healthcare, Menlo Park, California , USA; ²University of Wisconsin, Madison, Wisconsin, USA; ³University of Western Ontario, London, Canada; ⁴GE Healthcare, Madison, Wisconsin, USA

Simultaneous estimation of water, fat and T2*, achieved by algorithms such as T2*-IDEAL, is important in the evaluation hepatic fatty infiltration and iron overload. Fat was previously modeled as a single resonant frequency. However, fat has multiple spectral peaks that confound the estimation of fat content and T2*. In particular, T2* may be substantially under-estimated in the presence of fat, which would lead to overestimation of liver iron stores. In this work, we model a more accurate fat spectrum in the T2*-IDEAL algorithm and propose a spectrum self-calibration method that provides improved T2* estimation with no additional data required.

Scott B. Reeder¹, ², Philip Robson², Huanzhou Yu³, Ann Shimakawa³, Catherine D. Gard¹, Charles A. McKenzie⁴, Jean H. Brittain⁵

¹University of Wisconsin, Madison, Wisconsin, USA; ²Beth Israel-Deaconess Medical Center, Boston, Massachusetts, USA; ³GE Healthcare, Menlo Park, California , USA; ⁴University of Western Ontario, London, Canada; ⁵GE Healthcare, Madison, California , USA

Accurate non-invasive quantification of hepatic steatosis is an important unmet need in the diagnosis and management of non-alcoholic fatty liver disease (NAFLD). In this work, we describe the use of IDEAL, a chemical-shift based water-fat separation method, with a modified reconstruction that accounts for the signal from the multiple peaks of fat (“multipeak reconstruction”) and compare this method with MR spectroscopy (PRESS) in 31 patients. Excellent agreement between IDEAL and PRESS with a slope near 1.0 is achieved only when multipeak reconstruction is used, demonstrating the importance of all spectral peaks of fat when quantifying hepatic fat content.

Sebastian Sugay¹, Mark Bydder², Takeshi Yokoo², Gavin Hamilton², Nick Pinto², Richard Znamirowski², Eli Soumekh², Tanya Wolfson², Lillian Pacheco², Claude Sirlin²

¹John A. Burns School of Medicine, University of Hawaii, Honolulu, Hawaii, USA; ²University of California, San Diego, California , USA

In this abstract, we show that accounting for T2* decay of water and fat using a 6-point imaging protocol produces stable measurements in fat fraction (FF) despite experimental, SPIO-induced perturbations in T2*. As T2* decreased, the 6 point FF measurements remained unchanged, while the 2 point measurements underestimated FF.

Sophie Cavassila¹, Adriana Bucur¹, Hélène Ratiney¹, Cristina Cudalbu¹, Olivier Beuf¹, Frank Pilleul¹, ²

¹CNRS UMR 5220, Inserm U630, INSA-Lyon, Université de Lyon, Université Lyon 1, Villeurbanne, France; ²Hospices Civils de Lyon, Lyon, France

Despite the increasing prevalence of non-alcoholic fatty liver disease, the criteria used to diagnose the disorder remain poorly defined. Non invasive methods such as proton magnetic resonance spectroscopy (MRS), diffusion-weighted imaging, double-echo imaging (DEI) have been developed in order to substitute liver biopsy which is the gold standard. For this study, twelve patients with suspected liver diseases were investigated by MR imaging and spectroscopy of the liver. Both saturated lipids/water and unsaturated lipids/water ratios determined by MRS showed strong correlations with the percentage of liver steatosis determined by DEI. Moreover, the levels (relative to water) of the groups of metabolites at 3.2ppm and 3.5ppm determined by MRS showed also correlations with the percentage of liver steatosis determined by DEI.

Hero K. Hussain¹, Elif A. Oral, Suzan Rohrer, Barbara McKenna, Frank J. Londy, Marko K. Ivancevic, Thomas L. Chenevert

¹University of Michigan, Ann Arbor, Michigan, USA

The dual-flip MRI method at 3T is reliable to measure hepatic fat.  It correlates well with quantitative histopathologic measures and MRS, though unlike MRS, offers full-liver mapping of fat content and overcomes  heterogeneity.

Jochem Reinier van Werven¹, H. A. Marsman¹, A. J. Nederveen¹, F. J. ten Kate¹, T. M. van Gulik¹, J. Stoker¹

¹Academic Medical Center, Amsterdam, Netherlands

Hepatic steatosis is a risk factor in liver surgery. Liver biopsy remains the gold standard. 1H-MRS is a non-invasive alternative. The purpose of this study was to quantify hepatic steatosis with 3.0T 1H-MRS in patients undergoing liver resection.1H-MRS was performed preoperatively in twenty patients undergoing liver resection. Intra-operatively biopsies were taken for histopathological and biochemical analysis. 1H-MRS measurements of hepatic fat correlated with these assessments. 1H-MRS measurements could differentiate between different grades of hepatic steatosis.3.0T 1H-MRS can measure hepatic fat content and correlates with histopathological and biochemical analysis. It is also able to discriminate between grades of steatosis.

Mark Bydder¹, Takeshi Yokoo¹, Gavin Hamilton¹, Michael S. Middleton¹, Claude Sirlin¹

¹University of California San Diego, San Diego, California , USA

MRI is increasingly applied to the non-invasive assessment of steatosis. A linear correlation is often assumed between MRI estimates of fat and a gold standard (e.g. histology). We show that T1 and T2* relaxation can introduce nonlinearity in a typical liver imaging protocol.