Objective: The objective of our study was to evaluate the effect of single breath-hold dynamic subtraction MDCT of the liver on the performance of radiologists in detecting focal enhancement during the hepatic arterial phase.
Subjects and methods: This prospective study included 40 patients: 22 had hypervascular hepatocellular carcinoma (HCC), and 18 were without liver tumors. We obtained four-phase contrast-enhanced scans using a 16-MDCT unit. The section thickness and interval were 2 and 0.5 mm, respectively. Scanning for the first through fourth scans was started 10, 35, 70, and 180 seconds after the inception of contrast injection, respectively. Scanning for the first and second phase was within a single breath-hold. We subtracted the first-phase images from the second-phase images using software developed in-house. We used receiver operating characteristic (ROC) analysis with a continuous rating scale from 1 to 100 to compare observer performance in the detection of focal enhancement on second-phase images. Eight radiologists participated in the observer performance test, and their performances with unenhanced and contrast-enhanced original images were compared with their performances using contrast-enhanced subtracted images.
Results: For the eight observers, the mean area under the best-fit ROC curve (A(z)) values without and with the subtracted images were 0.86 +/- 0.05 (SD) and 0.91 +/- 0.03, respectively. The difference was significant (p < 0.01, two-tailed paired Student's t test).
Conclusion: The display of subtracted images significantly improved the diagnostic performance of radiologists in the detection of focal enhancement during the hepatic arterial phase (p < 0.01).