Fat suppression at 2D MR imaging of the hands: Dixon method versus CHESS technique and STIR sequence

To compare the effectiveness of fat suppression and the signal-to-noise ratio (SNR) of the Dixon method with those of the CHESS (Chemical Shift-Selective) technique and STIR (Short Tau Inversion Recovery) sequence in hands of normal subjects at 2D MR imaging.

14 healthy volunteers (mean age of 29.4 years) consented to have both hands prospectively imaged with SE T1 Dixon, T1 CHESS, T2 Dixon, T2 CHESS and STIR sequences in a 1.5T MR scanner. Three radiologists scored the effectiveness of fat suppression in bone marrow (EFS^BM) and soft tissues (EFS^ST) in 20 joints per subject. One radiologist measured the SNR in 10 bones per subject. Statistical analysis used two-way ANOVA with random effects, paired t-test and observed agreement to assess differences in effectiveness of fat suppression, differences in SNR and inter-observer agreement.

EFS^BM was statistically significantly higher for T1 Dixon than for T1 CHESS and for T2 Dixon than for T2 CHESS (p<0.0001). EFS^BM was significantly higher for T2 Dixon than for STIR in the coronal plane (p=0.0020). The SNR was significantly higher for T1 Dixon than for T1 CHESS and for T2 Dixon than for STIR (p<0.0001). The SNR was significantly lower for T2 Dixon than for T2 CHESS (p<0.0001).

The Dixon method yields more effective fat suppression and higher SNR than the CHESS technique at 2D T1-weighted MR imaging of the hands. At T2-weighted MR imaging, fat suppression is more effective with the Dixon method while SNR is higher with the CHESS technique.