Evaluation of dual-phase multi-detector-row CT for detection of intestinal bleeding using an experimental bowel model.

To evaluate dual-phase multi-detector-row computed tomography (MDCT) in the detection of intestinal bleeding using an experimental bowel model and varying bleeding velocities. The model consisted of a high pressure injector tube with a single perforation (1 mm) placed in 10-m-long small bowel of a pig. The bowel was filled with water/contrast solution of 30-40 HU and was incorporated in a phantom model containing vegetable oil to simulate mesenteric fat. Intestinal bleeding in different locations and bleeding velocities varying from zero to 1 ml/min (0.05 ml/min increments, constant bleeding duration of 20 s) was simulated. Nineteen complete datasets in arterial and portal-venous phase using increasing bleeding velocities, and seven negative controls were measured using a 64 MDCT (3-mm slice thickness, 1.5-mm reconstruction increment). Three radiologists blinded to the experimental settings evaluated the datasets in a random order. The likelihood for intestinal bleeding was assessed using a 5-point scale with subsequent ROC analysis. The sensitivity to detect bleeding was 0.44 for a bleeding velocity of 0.10-0.50 ml/min and 0.97 for 0.55-1.00 ml/min. The specificity was 1.00. The area under the curve was calculated to be 0.73, 0.88 and 0.89 for reader 1, 2 and 3, respectively. Dual-phase MDCT provides high sensitivity and specificity in the detection of intestinal bleeding with bleeding velocities of 0.5-1.0 ml/min. Therefore, MDCT should be considered as a primary diagnostic technique in the management of patients with suspected intestinal bleeding.