Comparison of efficacy of MR sequences used for detection of focal liver lesions

Background: The purpose of this study was to compare the sensitivities of different MR sequences (T2 TSE, DW EPI, T1 SGE) in detection of focal hepatic lesions.
Material/Methods: Evaluation included the MR studies (1.5T Unit: Magnetom Avanto) of 81 patients. Pathologic verification was obtained in 29 patients, whereas in remaining 52 patients, final diagnosis was based on other imaging studies, follow-up, and clinical data. In 56 patients, 171 liver lesions were confirmed: 84 malignant (52 metastases, 20 HCCs, 9 hemangioendotheliomas, 2 CCAs, 1 cystadenocarcinoma), and 87 benign (35 cysts, 31 hemangiomas, 13 abscesses, 7 FNHs, 1 cystadenoma). In 25 patients, liver lesions were excluded. Two radiologists identified the number of lesions on images obtained with the use of following sequences: T2WTSE, T2W fat-sat TSE, DW EPI (b=50,400,800 s/mm2), chemical shift T1 SGE (In-Out SGE), 3D T1 spoiled gradient-echo 3DT1SGE, dynamic 3DT1SGE post-Gd-BOPTA injection (with arterial, portal venous, parenchymal, 5 min-delayed phases) and 3DT1SGE 60-90 min post-Gd-BOPTA.
Results: Applied MR techniques yielded following sensitivities for detection of focal hepatic lesions: T2TSE 78.9%, T2 fat-sat 89.5%, DW EPI 95.9%, chemical shift T1SGE 80.7%, 3DT1SGE 81.9%, 3DT1SGE in arterial phase 93.0%, in portal venous phase 92.4%, in parenchymal phase 87.7%, in delayed phase 87.1%, and in hepatobiliary phase (60–90 min post-Gd-BOPTA) 84.8%.
Conclusions: The highest number of focal hepatic lesions was visualized on DW EPI sequence (n=164, sensitivity – 95.9%), followed by 3DT1SGE in arterial phase (n=159, sensitivity – 93.0%), 3DT1SGE in portal venous phase (n=158, sensitivity – 92.4%), and T2TSE fat-sat sequence (n=153, sensitivity – 89.5%).

Keywords: Magnetic Resonance Imaging, Liver Neoplasms, Contrast Media, Gadobenate dimeglumine