Fat quantification by use of phase change in dual-echo magnetic resonance imaging

Abstract

The demand for measurement of visceral fat in medical checkups is increasing because of attention to central obesity. Computed tomography is widely used for this purpose, but carries risks associated with radiation exposure. Fast imaging with a dual-echo technique acquiring water and fat signals both in-phase and out-of-phase using magnetic resonance imaging is desirable, but not all scanners are capable of this technique. We tried to quantify fat content by using a dual-echo gradient echo pulse sequence with an arbitrary echo time (TE) set. We calculated the phase change during the TE interval and extracted the region of fat by threshold processing. Variations in the precision of counting fat pixels caused by differences between TE1 and TE2 were measured in phantom experiments. We then evaluated the validity of this method by examination of a volunteer. The phantom study showed a large error when the TE interval was a multiple of the in-phase time for fat and water, and the error increased according to the prolongation of the TE. In the human study, phase wrapping readily occurred in the phase-change image. However, the region of fat was easily extracted with high-pass filtering. The fat signal can also be extracted by use of scanners that cannot be set simultaneously to in-phase and out-of-phase TE.

Evaluation of magnetic field homogeneity using in-out signal cycle mapping in gradient recalled echo images of a mixed water/oil phantom as a rough indication for daily quality control

“…The phase change caused in the TE increment is calculated by complex division from two sets of real and imaginary data. [20] Although there are other methods to unwrap the phase, these phase calculations cannot be easily performed by a MRI operator. As B 0 homogeneity worsens, the spectral peak becomes lower and the line width becomes broader in both methods at the off-center position.…”