Three-dimensional radiomics-based machine learning models can stratify women with endometrial carcinoma at preoperative multiparametric MRI based on high-risk histopathologic features and can ident... Background Stratifying high-risk histopathologic features in endometrial carcinoma is important for treatment planning. Radiomics analysis at preoperative MRI holds potential to identify high-risk ...

To determine the utility of computed tomographic (CT) enteroclysis for characterization of the status of the anastomotic site in patients with Crohn disease who had previously undergone ileocolic resection. Written informed consent was prospectively obtained from all patients, and the institutional review board approved the study protocol. CT enteroclysis findings in 40 patients with Crohn disease who had previously undergone ileocolic resection were evaluated independently by two readers. Endoscopic findings, histopathologic findings, and/or the Crohn disease activity index was the reference standard. Interobserver agreement between the two readers was calculated with kappa statistics. Associations between CT enteroclysis findings and anastomotic site status were assessed at univariate analysis. The sensitivity, specificity, and accuracy of CT enteroclysis, with corresponding 95% confidence intervals (CIs), for the diagnosis of normal versus abnormal anastomosis and the diagnosis of anastomotic recurrence versus fibrostenosis were estimated. Interobserver agreement regarding CT enteroclysis criteria was good to perfect (kappa = 0.72-1.00). At univariate analysis, stratification and anastomotic wall thickening were the two most discriminating variables in the differentiation between normal and abnormal anastomoses (P < .001). Stratification (P < .001) and the comb sign (P = .026) were the two most discriminating variables in the differentiation between anastomotic recurrence and fibrostenosis. In the diagnosis of anastomotic recurrence, severe anastomotic stenosis was the most sensitive finding (95% [20 of 21 patients]; 95% CI: 76.18%, 99.88%), both comb sign and stratification had 95% specificity (18 of 19 patients; 95% CI: 73.97%, 99.87%), and stratification was the most accurate finding (92% [37 of 40 patients]; 95% CI: 79.61%, 98.43%). In the diagnosis of fibrostenosis, both severe anastomotic stenosis and anastomotic wall thickening were 100% sensitive (eight of eight patients; 95% CI: 63.06%, 100.00%), and using an association among five categorical variables, including severe anastomotic stenosis, anastomotic wall thickening with normal or mild mucosal enhancement, absence of comb sign, and absence of fistula, yielded 88% sensitivity (seven of eight patients; 95% CI: 47.35%, 99.68%), 97% specificity (31 of 32 patients; 95% CI: 83.78%, 99.92%), and 95% accuracy (38 of 40 patients; 95% CI: 83.08%, 99.39%). CT enteroclysis yields objective and relatively specific morphologic criteria that help differentiate between recurrent disease and fibrostenosis at the anastomotic site after ileocolic resection for Crohn disease. http://radiology.rsna.org/lookup/suppl/doi:10.1148/radiol.09091165/-/DC1.