PURPOSE: To evaluate the performance of magnetic resonance (MR) colonography in the detection of colorectal mass lesions. MATERIALS AND METHODS: Twenty-three patients underwent MR colonography preceding colonoscopy. The colon was filled with a gadolinium-water mixture (1:100) with MR imaging guidance, and the patient was imaged prone and supine with a breath-hold three-dimensional spoiled gradient-recalled sequence. In addition, two-dimensional spoiled gradient-recalled images were acquired before and after intravenous administration of gadopentetate dimeglumine. Images were interactively analyzed on the basis of multiplanar reconstruction by two radiologists. For regions that were not conclusively assessable with multiplanar reconstruction, virtual intraluminal endoscopic images of the colon were reconstructed. MR findings were correlated with colonoscopic results. RESULTS: Two patients were excluded from the analysis. Findings in eight of 11 patients were correctly assessed as normal and in six of 10 as mass-positive. In the four patients with false-negative findings, one had two 8-mm polyps and the other three had polyps smaller than 5 mm. All nine mass lesions larger than 10 mm, as well as four of the 10 polyps ranging between 5 and 10 mm, were detected, but all polyps smaller than 5 mm were missed. In contrast to the polyps less than 5 mm, the four missed polyps (5-10 mm) could be identified retrospectively on virtual intraluminal endoscopic images. Contrast enhancement was documented in 13 polyps. CONCLUSION: Three-dimensional MR colonography provided virtual colonoscopic viewing and helped detection of colonic polyps.
PURPOSE: To evaluate the performance of magnetic resonance (MR) colonography in the detection of colorectal mass lesions. MATERIALS AND METHODS: Twenty-three patients underwent MR colonography preceding colonoscopy. The colon was filled with a gadolinium-water mixture (1:100) with MR imaging guidance, and the patient was imaged prone and supine with a breath-hold three-dimensional spoiled gradient-recalled sequence. In addition, two-dimensional spoiled gradient-recalled images were acquired before and after intravenous administration of gadopentetate dimeglumine. Images were interactively analyzed on the basis of multiplanar reconstruction by two radiologists. For regions that were not conclusively assessable with multiplanar reconstruction, virtual intraluminal endoscopic images of the colon were reconstructed. MR findings were correlated with colonoscopic results. RESULTS: Two patients were excluded from the analysis. Findings in eight of 11 patients were correctly assessed as normal and in six of 10 as mass-positive. In the four patients with false-negative findings, one had two 8-mm polyps and the other three had polyps smaller than 5 mm. All nine mass lesions larger than 10 mm, as well as four of the 10 polyps ranging between 5 and 10 mm, were detected, but all polyps smaller than 5 mm were missed. In contrast to the polyps less than 5 mm, the four missed polyps (5-10 mm) could be identified retrospectively on virtual intraluminal endoscopic images. Contrast enhancement was documented in 13 polyps. CONCLUSION: Three-dimensional MR colonography provided virtual colonoscopic viewing and helped detection of colonic polyps.
Authors: C Savoye-Collet; J B Roset; E Koning; C Charpentier; S Hommel; E Lerebours; J N Dacher; G Savoye Journal: Eur Radiol Date: 2012-04-27 Impact factor: 5.315