Fat-water interface on susceptibility-weighted imaging and gradient-echo imaging: comparison of phantoms to intracranial lipomas.

OBJECTIVE In a clinical setting, lipoma can sometime show low signal intensity on susceptibility-weighted imaging (SWI) mimicking hemorrhage. The purpose of this study was to evaluate the fat-water interface chemical-shift artifacts between SWI and T2*-weighted imaging with a phantom study and evaluate SWI in lipoma cases. MATERIALS AND METHODS SWI, magnitude, high-pass filtered phase, and T2*-weighted imaging of a lard-water phantom were evaluated in the in-phase, out-of phase, and standard partially out-of-phase TE settings used for clinical 3-T SWI (19.7, 20.9, and 20.0 ms, respectively) to identify the most prominent fat-water interface low signal. SWI of five cases of CNS lipoma were retrospectively evaluated by two neuroradiologists. RESULTS TE at 19.7 ms (in-phase) showed the minimum fat-water interface low signal in the phase-encoding direction on magnitude, high-pass filtered phase, and SWI. TE at 20.9 ms (out-of-phase) showed the maximum fat-water interface in the phase-encoding direction on magnitude, high-pass filtered phase, and SWI. TE at 20.0 ms (partially out-of-phase) showed more fat-water interface low signal on SWI than on T2*-weighted imaging, especially in the phase-encoding direction. All lipomas in the five patients showed high signal intensity with surrounding peripheral dark rim on SWI. CONCLUSION Fat-water interface is more prominent on the standard TE setting used for clinical SWI (20.0 ms) than that of T2*-weighted imaging and shows a characteristic surrounding peripheral low-signal-intensity rim in lipoma. Knowing the fat-water appearance on SWI is important to avoid misinterpreting intracranial lipomas as hemorrhages.