Abstract
Rationale and Objectives. Both magnetic resonance (MR) and computed tomographic (CT) colonography are useful for colon examination. With sensitivities close to those for conventional colonoscopy (CC) for polyps, colonography has been proposed as an alternative to diagnostic CC. MR colonography (MRC) with fecal tagging may be a method of gaining further patient acceptance and widespread use, but the method has to be optimized. The aim of our study was to evaluate the quality of a new contrast agent mixture and to validate a new method for evaluating the tagging efficiency of contrast agents. Materials and Methods. Twenty patients referred to CC underwent dark lumen MRC prior to the colonoscopy. Two groups of patients received two different oral contrast agents (barium sulfate and barium sulfate/ferumoxsil) as a laxative-free fecal tagging prior to the MRC. After MRC, the contrast agent was rated qualitatively (with the standard method using contrast-to-wall ratio) and subjectively (using a visual analog scale [VAS]) by three different blinded observers. Results. Evaluated both qualitatively and subjectively, the tagging efficiency of barium sulfate/ferumoxsil was significantly better (P <.05) than barium sulfate alone. The VAS method for evaluating the tagging efficiency of contrast agents showed a high correlation (observer 11, r = 0.91) to the standard method using contrast-to-wall ratio and also a high interclass correlation (observer 11 and III = 0.89/0.85). MRC found I of 22 (5%) polyps < 6 mm, 2 of 3 (67%) polyps 6-10 mm, and 2 of 2 (100%) polyps > 10 mm. Conclusion. MRC with fecal tagging using barium sulfate/ferumoxsil as contrast agent will give better overall assessment of the colon wall compared to barium sulfate alone. Furthermore, the VAS method of evaluating fecal tagging efficiency correlated with the standard method of calculating the contrast-to-wall ratio
Udgivelsesdato: 2008/5
Udgivelsesdato: 2008/5
Original language | English |
---|---|
Journal | Academic Radiology |
Volume | 15 |
Issue number | 5 |
Pages (from-to) | 576-583 |
Number of pages | 7 |
ISSN | 1076-6332 |
Publication status | Published - 2008 |