Feasibility of wide detector three-pass arterial phase liver CT in patients with cirrhosis: timing of hyperenhancing lesion peak conspicuity.

Abstract

To evaluate feasibility of a wide detector liver CT protocol with three acquisitions in the hepatic arterial phase. Forty-one patients with cirrhosis prospectively underwent a wide detector axial liver CT protocol. Three 16cm axial liver acquisitions were obtained during a single breath hold at peak aortic enhancement plus 10, 20, and 25s. Two readers working separately scored overall exam quality, identified hyperenhancing lesions, and subjectively scored and ranked relative lesion conspicuity. Objective lesion enhancement was measured and CNR calculated. Data were analyzed using a generalized linear models and Tukey's post hoc testing. Seventy-one hyperenhancing lesions were identified with average size of 1.8cm (range 0.4-9.6cm). The two readers separately identified 60 and 54 lesions on the 10s arterial acquisition, 70 and 67 on the 20s, and 52 and 51 on the 25s. The readers determined all exams had diagnostic image quality. Subjective ranking of lesion conspicuity was greatest at 20s in 62% of lesions but was greatest at 10 or 25s in 38%. CNR was highest at 20s in 58% of lesions but was highest at 10 or 25s in 42%. Overall, there was no significant difference in mean CNR between the three arterial acquisitions. A wide detector axial liver CT protocol with three acquisitions in the hepatic arterial phase is technologically feasible and results in diagnostic image quality. With this protocol, peak subjective and objective hyperenhancing lesion conspicuity may be earlier or later than 20s in up to 40% of lesions.