Comparison between signal-to-noise ratio, liver-to-muscle ratio, and 1/T2 for the noninvasive assessment of liver iron content by MRI

Purpose To compare different MRI-derived parameters, i.e., liver signal-to-noise ratio (LSNR), liver-to-muscle ratio (LMR) and liver transversal relaxation rate (R2), in terms of their correlation with the ex vivo determined iron content in an experimental model of liver iron overload. Materials and Methods Multi-echo spin echo (SE) images of the liver were acquired at 4.7 T from a group of 33 male wistar rats subjected to a high iron content diet for feeding periods ranging from 2 to 50 days. Liver transversal relaxation time, liver signal-to-noise ratio, and liver-to-muscle ratio were measured over the same region of interest in order to get a direct comparison between these parameters. After MRI experiments, the rats were sacrificed and the liver iron content was measured ex vivo by atomic absorption spectroscopy. Results The iron content is better correlated to the LSNR than to the other parameters (LMR, R2). Conclusion The finding that liver signal-to-noise ratio is better correlated to the iron content than the liver T2 relaxation rate is relevant for clinical applications of MRI because a T2 determination is more time-consuming, both for acquisition and postprocessing of images, than a simple SNR determination. J. Magn. Reson. Imaging 2003;17:589–592. © 2003 Wiley-Liss, Inc.