Boris Guiu, Presenter: Nothing to Disclose

Jean Michel Petit, Abstract Co-Author: Nothing to Disclose

Serge Ludwig Aho-Glélé, Abstract Co-Author: Nothing to Disclose

Douraied Ben Salem MD, Abstract Co-Author: Nothing to Disclose

David Masson, Abstract Co-Author: Nothing to Disclose

Périne Manzoni, Abstract Co-Author: Nothing to Disclose

Patrick Hillon, Abstract Co-Author: Nothing to Disclose

Denis Krause MD, Abstract Co-Author: Nothing to Disclose

Bruno Verges, Abstract Co-Author: Nothing to Disclose

Jean Pierre Cercueil MD, Abstract Co-Author: Nothing to Disclose

To assess systematic errors in liver fat fraction (FF) resulting from T1 and T2* decays when using dual- and triple-echo gradient-recalled-echo sequences and to test the robustness of these sequences after iron overloading.

The appropriate ethics committee approved this prospective study and the 40 type-2 diabetic patients (mean age, 60 years) gave written informed consent before undergoing liver FF measurement by 3.0T 1H magnetic resonance spectroscopy (corrected for T1 and T2 decays) as the reference standard and by four gradient-recalled-echo sequences, namely, dual- and triple-echo imaging with 20° and 60° flip angle. The same four sequences were repeated in 8 patients after ferumoxide injection. Systematic errors in liver FF resulting from T1 and T2* decays were determined from the linear regressions of spectroscopy FF values over FF values by each gradient-recalled-echo sequence. Robustness was tested using Wilcoxon’s signed-rank test by comparing spectroscopy FF values and gradient-recalled-echo FF values corrected using the linear regression equations.

T1 decay produced 15.4%-30% relative systematic error and T2* decay 3.5%-5.3% absolute systematic error in FF. After ferumoxide, dual-echo imaging with 20° and 60° flip angle, even when corrected, showed absolute differences of 5.1% (range, 1.6–7.7) (P=0.012) and 6.8% (range, 1.8–9.7) (P=0.012) versus spectroscopy, respectively; compared to only 1.1% (range, 0.1–2.7) (P=0.036) and 1.1% (range, 0.1–2.3) (P=0.208) with triple-echo imaging and 20° and 60° flip angle, respectively.

Triple-echo imaging is considerably more robust than dual-echo imaging. Applying a systematic correction to dual-echo results is not reliable.

In unselected patients with type 2 diabetes, T1 decay produced a 15.4%-30% relative systematic error and T2* decay a 3.5% to 5.3% absolute systematic error in measured liver fat content.

Guiu, B, Petit, J, Aho-Glélé, S, Ben Salem, D, Masson, D, Manzoni, P, Hillon, P, Krause, D, Verges, B, Cercueil, J, Liver Fat Quantification by Gradient-Echo Imaging at 3.0 T: Systematic Errors Due to T1 and T2* Decays and Robustness of Dual- and Triple-Echo Methods.  Radiological Society of North America 2010 Scientific Assembly and Annual Meeting, November 28 - December 3, 2010 ,Chicago IL. http://archive.rsna.org/2010/9009880.html