BACKGROUNDType II diabetes mellitus usually related to visceral and other organ (ectopic) fat. The purpose of this study was to detect hepatic and pancreatic fat infiltration in type II diabetes mellitus patients using 3.0T magnetic resonance (MR) and to compare the performance of iterative decomposition of water and fat with echo asymmetry and least-squares estimation (IDEAL-Quant) with single-voxel proton spectroscopy (H(1)-MRS).

METHODSThe study protocol was approved by our Institutional Review Board. Written informed consent was obtained from each subject in this study. We prospectively performed IDEAL-Quant and single-voxel proton spectroscopy with an echo time of 35 ms on 24 type II diabetes patients and 10 healthy volunteers. The hepatic proton density fat fraction (HPDFF) and pancreatic proton density fat fraction (PPDFF) were calculated, compared, and analyzed by t-tests and Spearman's correlation.

RESULTSThe HPDFF and PPDFF measured with IDEAL-Quant were significantly different between the healthy volunteers and type II diabetes patients (th = 9.377, P = 0.000; tp = 2.813, P = 0.008). The HPDFF and PPDFF measured with MRS were also significantly different between the healthy volunteers and type II diabetes patients (th = 5.342, P = 0.000; tp = 2.63, P = 0.013). The HPDFF and PPDFF measured by the two methods were in good agreement (rh = 0.854, P = 0.000; rp = 0.774, P = 0.000). The HPDFF and PPDFF were not significantly correlated with each other (rMRS = 0.203, p = 0.248; rIDEAL-Quant = 0.301, P = 0.084). The PPDFF measured with IDEAL-Quant was associated with body mass index (r = 0.377, P = 0.028).

CONCLUSIONIDEAL-Quant is a nice method for hepatic and pancreatic fat detection, and it can be applied in clinical practice.