Jeong-Sik Yu1, Joo Hee Kim, Jae-Joon Chung, Ki Whang Kim. 1. Department of Radiology and the Research Institute of Radiological Science, Yonsei University College of Medicine, Gangnam Severance Hospital, Gangnam-Gu, Seoul, Republic of Korea. yjsrad97@yuhs.ac
Abstract
PURPOSE: To investigate the additional value of diffusion-weighted imaging (DWI) in the MRI assessment of perilesionally recurrent hepatocellular carcinomas (HCCs) after transcatheter arterial chemoembolization (TACE). MATERIALS AND METHODS: For gadopentetate dimeglumine-enhanced multiphasic dynamic images with precontrast T1- and T2-weighted images of 23 patients treated with TACE, two radiologists independently sorted the confidence levels for the presence of a viable tumor in the vicinity of the treated lesions into five grades. In another session, images from DWI (b factor = 50, 400, and 800 s/mm(2)) were added to the previously reviewed images and the same radiologists sorted the confidence levels. RESULTS: A total of 26 lesions (0.7-3.5 cm) from 16 patients were confirmed to be perilesional recurrences of HCC. The area under the receiver operating characteristic curve (A(z)) for the second interpretation session (0.826) was not significantly different (P = 0.299) from that of the first session (0.859). The overall sensitivity was increased from 85% to 92%, but the specificity decreased from 65% to 50% after adding DWI. CONCLUSION: The addition of DWI has the potential to improve sensitivity, but not the overall diagnostic accuracy, in the assessment of perilesional recurrence of HCCs after chemoembolization. (c) 2009 Wiley-Liss, Inc.
PURPOSE: To investigate the additional value of diffusion-weighted imaging (DWI) in the MRI assessment of perilesionally recurrent hepatocellular carcinomas (HCCs) after transcatheter arterial chemoembolization (TACE). MATERIALS AND METHODS: For gadopentetate dimeglumine-enhanced multiphasic dynamic images with precontrast T1- and T2-weighted images of 23 patients treated with TACE, two radiologists independently sorted the confidence levels for the presence of a viable tumor in the vicinity of the treated lesions into five grades. In another session, images from DWI (b factor = 50, 400, and 800 s/mm(2)) were added to the previously reviewed images and the same radiologists sorted the confidence levels. RESULTS: A total of 26 lesions (0.7-3.5 cm) from 16 patients were confirmed to be perilesional recurrences of HCC. The area under the receiver operating characteristic curve (A(z)) for the second interpretation session (0.826) was not significantly different (P = 0.299) from that of the first session (0.859). The overall sensitivity was increased from 85% to 92%, but the specificity decreased from 65% to 50% after adding DWI. CONCLUSION: The addition of DWI has the potential to improve sensitivity, but not the overall diagnostic accuracy, in the assessment of perilesional recurrence of HCCs after chemoembolization. (c) 2009 Wiley-Liss, Inc.
Authors: Ajaykumar C Morani; Khaled M Elsayes; Peter S Liu; William J Weadock; Janio Szklaruk; Jonathan Russell Dillman; Asra Khan; Thomas L Chenevert; Hero K Hussain Journal: World J Radiol Date: 2013-03-28