Preoperative detection of malignant liver tumors: Comparison of 3D-T2-weighted sequences with T2-weighted turbo spin-echo and single shot T2 at 1.5 T

Publication date: March 2018
Source:European Journal of Radiology, Volume 100
Author(s): Maxime Barat, Philippe Soyer, Raphael Dautry, Marc Pocard, Rea Lo-Dico, Haythem Najah, Clarisse Eveno, Christophe Cassinotto, Anthony Dohan
PurposeTo assess the performances of three-dimensional (3D)-T2-weighted sequences compared to standard T2-weighted turbo spin echo (T2-TSE), T2-half-Fourier acquisition single-shot turbo spin-echo (T2-HASTE), diffusion weighted imaging (DWI) and 3D-T1-weighted VIBE sequences in the preoperative detection of malignant liver tumors.MethodsFrom 2012 to 2015, all patients of our institution undergoing magnetic resonance imaging (MRI) examination for suspected malignant liver tumors were prospectively included. Patients had contrast-enhanced 3D-T1-weighted, DWI, 3D-T2-SPACE, T2-HASTE and T2-TSE sequences. Imaging findings were compared with those obtained at follow-up, surgery and histopathological analysis. Sensitivities for the detection of malignant liver tumors were compared for each sequence using McNemar test. A subgroup analysis was conducted for HCCs. Image artifacts were analyzed and compared using Wilcoxon paired signed rank-test.ResultsThirty-three patients were included: 13 patients had 40 hepatocellular carcinomas (HCC) and 20 had 54 liver metastases. 3D-T2-weighted sequences had a higher sensitivity than T2-weighted TSE sequences for the detection of malignant liver tumors (79.8% versus 68.1%; P < 0.001). The difference did not reach significance for HCC. T1-weighted VIBE and DWI had a higher sensitivity than T2-weighted sequences. 3D-T2-weighted-SPACE sequences showed significantly less artifacts than T2-weitghted TSE.Conclusion3D-T2-weighted sequences show very promising performances for the detection of liver malignant tumors compared to T2-weighted TSE sequences.



from Imaging via alkiviadis.1961 on Inoreader http://ift.tt/2DedHbd