Comparison of fat suppression capability for chest MR imaging with Dixon, SPAIR and STIR techniques at 3 Tesla MR system

Abstract

PurposeTo directly compare fat suppression efficacy of Dixon, spectral attenuated inversion recovery (SPAIR) and short tau inversion recovery (STIR) techniques for a 3T MR system. Materials and methodsNineteen consecutive patients (11 men, 8 women; mean age 67.9years) underwent chest MR imaging. Contrast-noise ratio (CNR) between muscle and fat (CNRFat) was calculated by ROI measurement. Then, two radiologists used a five-point scale for visual assessment of image quality, fat suppression capability and lesion detection. Next, the quantitative calculations obtained with all three techniques were compared by means of Tukey's honest significant difference (HSD) test. Inter-observer agreements were assessed by kappa statistics and χ2 test. Fisher's least significant difference (LSD) test was used for comparison among all scores. ResultsCNRFat of SPAIR were significantly higher than those of Dixon and STIR (p<0.001). Inter-observer agreements were assessed as significantly moderate, substantial or almost perfect (0.51<κ<0.89, p<0.0001). Total image quality and fat suppression capability of Dixon and STIR were significantly higher than those of SPAIR (p<0.001). ConclusionDixon technique has better fat suppression capability on T2-weighted turbo SE imaging than SPAIR technique on chest MR imaging examined with a 3T MR system, although it can't substitute STIR technique on fat suppression and lesion visualization.