Background: We evaluated the performance of double-contrast MR colonography in detecting colorectal tumors experimentally induced by chemical substances in the large bowel of the rat.
Material/methods: Eight Sprague-Dawley rats developed 67 chemically induced tumors of the large bowel. The animals were imaged with a double-contrast MR technique at 0.5T, using a head coil for both signal transmission and reception. Prior to MR acquisition, room air was insufflated into the rats' large bowel. Unenhanced and contrast-enhanced T1, proton-density and T2 sequences were obtained in the axial and coronal planes. Autopsy was performed immediately after MR examination. MR images were interpreted by consensus of two observers, and the results were compared with post-mortem data.
Results: Sensitivity (60% vs. 93%), specificity (63% vs. 81%), positive (85% vs. 93%) and negative (30% vs. 81%) predictability, and global diagnostic accuracy (60% vs. 89%) were obtained for all tumors (n=67) and for the group of tumors larger than 1.5 cm (n=43), respectively. No tumors less than 1.5 cm were detected.
Conclusions: Double-contrast MR colonography is a promising modality in detecting colorectal neoplasms larger than 1.5 cm in an animal tumor model.