• A Maubon, V Berger, P Aubas, J M Ferru, and J P Rouanet.
• Service d'Imagerie Médicale, CMC Beausoleil, Montpellier. a.maubon@cge-ol.fr
• J Radiol. 1999 Mar 1; 80 (3): 291-6.

PurposeTo assess, quantitatively and qualitatively, the diagnostic value of a segmented EPI T1W sequence compared to T1W and T2W TSE sequences.Material And MethodsA prospective analysis of abdominal and pelvic MRI examinations of 70 patients (44 women, 26 men, mean age of 61 years), was performed on a 0.5 T supraconductive magnet with 15 mT/m gradients. The sequences were randomized and compared in a blinded fashion by 3 independent reviewers: TSE T1W (TR/TE = 500/12 ms, NSA = 6, turbo factor 5, 3:49 min), EPI T1W (TR/TE = 500/30 ms, NSA = 6, EPI factor = 7, 2:13 min) and UTSE T2W (TR/TE = 1600-2500/100, NSA = 6, turbo factor = 31, 2:20 min).ResultsQuantitatively, no significant difference was found between T1W sequences for signal to noise ratio. The EPI T1W sequence had lower signal but stronger enhancement after gadolinium injection. Qualitatively, EPI T1W had significantly less flow artefacts (p < 0.001, wilcoxon test), and more chemical shift artifact (p < 0.01). For lesion detection, differences were not statistically significant between T1W sequences or between paired T1W and T2W sequences (sensitivity and specificity 84 and 86% for TSE T1W 76 and 86% for EPI T1W, 78 and 79% for UTSE T2W, 90 and 65% for TSE T1W-UTSE T2W, 88 and 65% for EPI T1W-UTSE T2W). Kappa concordance test (0.686) and Mac Nemar symmetry test (3.55) were high between T1W sequences.ConclusionThe segmented EPI T1W sequence used had equivalent results compared to the TSE T1W sequence, it allows a 40% reduction in acquisition time and this without difference in the diagnostic performances of the reviewers.

32 keywords...

hide…