W Luboldt1, J F Debatin. 1. Institute of Diagnostic Radiology, University Hospital Zurich, Switzerland.
Abstract
BACKGROUND: To evaluate the potential of magnetic resonance colonography (MRC) in detecting colorectal mass lesions. METHODS: Twenty patients underwent MR imaging (MRI) before colonoscopy. The colon was filled with a gadolinium (0.5 mol):water mixture (1:100) under MRI control, and patients were imaged while breath-holding imaged with a three-dimensional spoiled gradient echo sequence in the prone and supine positions. Images were interactively analyzed based on the combination of multiplanar reconstruction and virtual colonoscopy by a radiologist blinded to colonoscopic findings and the patient's history. MRC interpretations were correlated with colonoscopic results. RESULTS: Polyps smaller than 5 mm could not be identified with MRC. The sensitivity for detecting polyps of 5-10 mm was 70%, whereas mass lesions larger than 10 mm were all detected (sensitivity = 100). The sensitivity, specificity, and accuracy for identifying polyp-positive patients including the three patients with small (<5 mm) polyps were 64%, 89%, and 75%, respectively. CONCLUSION: Virtual colonoscopy based on MRI data is feasible and should be evaluated in a larger sample of patients.
BACKGROUND: To evaluate the potential of magnetic resonance colonography (MRC) in detecting colorectal mass lesions. METHODS: Twenty patients underwent MR imaging (MRI) before colonoscopy. The colon was filled with a gadolinium (0.5 mol):water mixture (1:100) under MRI control, and patients were imaged while breath-holding imaged with a three-dimensional spoiled gradient echo sequence in the prone and supine positions. Images were interactively analyzed based on the combination of multiplanar reconstruction and virtual colonoscopy by a radiologist blinded to colonoscopic findings and the patient's history. MRC interpretations were correlated with colonoscopic results. RESULTS:Polyps smaller than 5 mm could not be identified with MRC. The sensitivity for detecting polyps of 5-10 mm was 70%, whereas mass lesions larger than 10 mm were all detected (sensitivity = 100). The sensitivity, specificity, and accuracy for identifying polyp-positive patients including the three patients with small (<5 mm) polyps were 64%, 89%, and 75%, respectively. CONCLUSION: Virtual colonoscopy based on MRI data is feasible and should be evaluated in a larger sample of patients.
Authors: Andreas G Schreyer; Alois Fürst; Ayman Agha; Ron Kikinis; Karl Scheibl; Jürgen Schölmerich; Stefan Feuerbach; Hans Herfarth; Johannes Seitz Journal: Int J Colorectal Dis Date: 2004-04-15 Impact factor: 2.571