Zehour AlSabban1, Peter Church2, Rahim Moineddin3, Oscar M Navarro1, Mary-Louise Greer1, Thomas Walters2, Govind B Chavhan4. 1. Department of Diagnostic Imaging, The Hospital For Sick Children and Medical Imaging, University Of Toronto, 555 University Avenue, Toronto, Ontario M5G 1X8, Canada. 2. Division of Gastroenterology, Hepatology and Nutrition, The Hospital For Sick Children and Medical Imaging, University Of Toronto, 555 University Avenue, Toronto, Ontario M5G 1X8, Canada. 3. Department of Family and Community Medicine, University of Toronto, Toronto, Ontario, Canada. 4. Department of Diagnostic Imaging, The Hospital For Sick Children and Medical Imaging, University Of Toronto, 555 University Avenue, Toronto, Ontario M5G 1X8, Canada. Electronic address: govind.chavhan@sickkids.ca.
Abstract
PURPOSE: To determine interobserver agreement (IOA) and accuracy of conventional MR enterography (MRE), qualitative diffusion, and apparent diffusion coefficient (ADC) values for detecting clinically active inflammation. METHODS: MREs in 57 consecutive children with suspected inflammatory bowel disease were retrospectively reviewed. RESULTS: Substantial IOA for conventional MRE (kappa=0.65) and qualitative diffusion (kappa=0.64), but fair to good IOA for ADC, (intra-class coefficient=0.63) were seen. Conventional MRE detected active clinical inflammation well (area under curve [AUC] 0.725), while qualitative diffusion and ADC did not perform well (AUC=0.572 and 0.461, respectively). CONCLUSION: DWI can be helpful in diagnosing inflammatory bowel disease but does not perform well in identifying those with active inflammation.
PURPOSE: To determine interobserver agreement (IOA) and accuracy of conventional MR enterography (MRE), qualitative diffusion, and apparent diffusion coefficient (ADC) values for detecting clinically active inflammation. METHODS: MREs in 57 consecutive children with suspected inflammatory bowel disease were retrospectively reviewed. RESULTS: Substantial IOA for conventional MRE (kappa=0.65) and qualitative diffusion (kappa=0.64), but fair to good IOA for ADC, (intra-class coefficient=0.63) were seen. Conventional MRE detected active clinical inflammation well (area under curve [AUC] 0.725), while qualitative diffusion and ADC did not perform well (AUC=0.572 and 0.461, respectively). CONCLUSION: DWI can be helpful in diagnosing inflammatory bowel disease but does not perform well in identifying those with active inflammation.