I Comput Assist Tomogr , Volume 32, Number 5, September/October 2008
Takatoshi Kitamura, MD, Tomoaki Ichikawa, MD, Sukru Mehmet Erturk, MD, Hiroto Nakajima, MD, Hironobu Sou, MD, Tsutomu Araki, MD, Shunichi Okada, MD, and Nobuyuki Enomoto, MD
Purpose: To compare single arterial-phase (SAP) computed tomography (CT) imaging with bolus tracking (BT) with double arterial-phase (DAP) CT imaging for detecting hypervascular hepatocellular carcinoma.
Materials and Methods: The DAP images were obtained at 25 (DAP-early) and 40 seconds (DAP-late) after the start of contrast material injection. All patients underwent SAP-BT imaging where images were obtained 10 seconds after the CT attenuation value of the aorta reached the threshold value of 120 Hounsfield unit (HU) in 29 (group 120-HU), 160 HU in 30 (group 160-HU), and 200 HU in 32 patients (group 200-HU). Attenuation conspicuity with SAP-BT technique was compared with that with DAP technique using repeated-measures analysis of variance. Attenuation conspicuity and mean scan delays with SAP-BT images obtained with different threshold values were compared using analysis of variance. The sensitivities were compared using McNemar and Fisher exact tests.
Results: Within all groups, mean attenuation conspicuity with SAP-BT and DAP-late was significantly higher than that with DAP-early. Regarding SAP-BT, mean attenuation conspicuity in group 200-HU (42 + 18 HU) was significantly higher than those in groups 120-HU (23 ± 11 HU) and 160-HU (25 ± 11 HU). Mean scan delays for SAP-BT were 24.2 seconds in group-120 HU, 26.8 seconds in group-160 HU, and 31.1 seconds in group-200 HU (P < 0.001). The mean sensitivity with SAP-BT technique in group 200-HU (92.7%) was significantly higher than those in groups 120-HU (72.4%) and 160-HU (71.1%).
Conclusions: Single arterial-phase CT scanning with bolus tracking can be effectively used to detect hepatocellular carcinoma when a threshold value of 200 HU is used.