[18F]FDG PET/MR enterography for the assessment of inflammatory activity in Crohn’s disease: comparison of different MRI and PET parameters


Purpose: To evaluate the diagnostic performance of integrated whole-body positron emission tomography (PET)/magnetic resonance (MR) enterography in patients with Crohn’s disease (CD). Methods: Fifty patients with known CD and recurrent symptoms underwent ileocolonoscopy (reference standard) as well as PET/MR enterography. Seven ileocolonic segments were endoscopically analysed using the Simplified Endoscopic Activity Score for Crohn’s Disease (SES-CD) and additionally classified into three categories of inflammation (none, mild to moderate and severe ulcerative inflammation). A total of 14 PET/MR parameters were applied for the assessment of inflamed segments. Contingency tables and the chi-squared test were used for the analysis of qualitative parameters, and the Mann-Whitney U test and receiver operating characteristic (ROC) curve for the analysis of quantitative parameters. The PET/MR parameters were ranked according to their diagnostic value by random forest classification. Correlations between PET/MR parameters and the severity of inflammation on endoscopy and SES-CD were tested using Spearman’s rank correlation test. Results: A total of 309 segments could be analysed. Based on multivariate regression analysis, wall thickness and the comb sign were the most important parameters for predicting segments with active inflammation of any type. SUVmax ratio of the bowel segment (relative to SUVmax of the liver) was the most important parameter for detecting segments with severe ulcerative inflammation. Wall thickness was the only parameter that moderately correlated with inflammation severity on endoscopy as well as with SES-CD (ρ = 0.56 and 0.589, both p < 0.001). Conclusion: PET/MR enterography is an excellent noninvasive diagnostic method, and both MR parameters and PET findings provided high accuracy in detecting inflamed segments.

Click the Cite button above to demo the feature to enablevisitors to import publication metadata into their reference management software.
Felix Nensa
Felix Nensa

My research interests include medical digitalization, computer vision and radiology.