Repeatability of magnetic resonance measurements used for estimating Crohn's disease activity

The MR activity indices used for quantification and follow-up of Crohn's disease are composed of a number of subjectively determinable components with equivocal repeatability. The purpose of this article was to assess the repeatability of measurements used for quantitative estimation of Crohn's disease activity in the terminal ileum. In five adults (23-57 y.o.) and 12 children (10-17 y.o.) with active terminal ileitis, the inflamed bowel was divided into 3 cm segments (n = 32 in adults, n = 46 in children), and measurements for the calculation of MaRIA and Clermont scores were performed. Parameters included apparent diffusion coefficients (ADC) for diffusion-weighted imaging (DWI) sequences with selective and non-selective fat suppression, wall signal enhancement before (WSI-preGd) and after (WSI-preGd) gadolinium enhancement, bowel thickness, and presence of ulcers. The measurements were standardised (accurate site-to-site comparison, exact ROI size, where applicable) and repeated by the same researcher after two months. Intra-observer agreement for ADC, WSI-preGd and WSI-postGd, bowel thickness was assessed with a paired t-test, and the significant difference in presence/absence of ulcers was assessed by the Pearson ² test. Absolute difference was not found between the 1^st and 2^nd measurements of ADC, WSI-preGd, WSI-postGd and wall thickness. There was systematic difference in the presence of bowel ulcers. In standardised conditions the repeatability of ADC, WSI-preGd and WSI-postGd is high. Efforts must be made to precisely define the size and appearance of ulcers that may be included in the index calculation.