Added diagnostic value of T2-weighted MR imaging to gadolinium-enhanced three-dimensional dynamic MR imaging for the detection of small hepatocellular carcinomas

• Published in: European journal of radiology 2008-08, Vol.67 (2), p.304-310
• Main Authors: Kim, Young Kon, Lee, Young Hwan, Kim, Chong Soo, Han, Young Min
• Format: Article
• Language: English
• Subjects: Adult; Aged; Biological and medical sciences; Carcinoma, Hepatocellular - diagnosis; cirrhosis; Contrast Media; Female; Gastroenterology. Liver. Pancreas. Abdomen; Humans; Image Interpretation, Computer-Assisted; Investigative techniques, diagnostic techniques (general aspects); Liver; Liver neoplasm; Liver Neoplasms - diagnosis; Liver. Biliary tract. Portal circulation. Exocrine pancreas; Magnetic resonance imaging; Magnetic Resonance Imaging - methods; Male; Medical sciences; Meglumine - analogs & derivatives; Middle Aged; Organometallic Compounds; Other diseases. Semiology; Predictive Value of Tests; Radiology; Retrospective Studies; ROC Curve; Sensitivity and Specificity; Tumors
• ISSN: 0720-048X; 1872-7727
• DOI: 10.1016/j.ejrad.2007.07.001

Abstract Purpose To assess the added value of T2-weighted MRI to gadolinium-enhanced dynamic MRI for detection of HCCs. Materials and methods Two readers retrospectively analyzed MRIs of 115 patients with 131 HCCs (size; 0.6–2.0 cm) that had been diagnosed by histology ( n = 41) or imaging findings ( n = 90). Two separate blind image analyses of the gadolinium set and the combined T2-weighted imaging and gadolinium sets were performed. Diagnostic accuracy was evaluated using the alternative-free response receiver operating characteristic method with four-point scale. Sensitivity and positive predictive value were also calculated. Results For both observers, the Az values and sensitivities with the combined T2-weighed imaging and gadolinium set (mean Az 0.806, sensitivity 84.7) were significantly higher than those with the gadolinium set (mean Az 0.660, sensitivity 59.9) ( p < 0.05). The addition of T2-weighted imaging led to a change in diagnosis for 27 lesions by both observers, which at gadolinium set were assigned a confidence level of 1 or 2 but at additional reading of T2-weighted imaging were assigned a confidence level of 3 or 4. For the positive predictive values, each image set showed a similar value for each observer. Conclusion The addition of T2-weighted imaging to gadolinium-enhanced 3D dynamic imaging could be helpful in the detection of HCC by increasing reader confidence for HCCs with equivocal findings on gadolinium-enhanced MRIs.