Test-retest assessment of non-contrast MRI sequences to characterise and quantify the small bowel wall in healthy participants.

Objective: Quantitative Magnetic Resonance Imaging sequences have been investigated as objective imaging biomarkers of fibrosis and inflammation in Crohn's disease.

Aim: To determine the repeatability and inter- and intra-observer agreement of these measures in the prepared small bowel wall.

Methods: Ten healthy participants were scanned at 3 T on 2 separate occasions using T1 and T2 relaxometry, IVIM-DWI and MT sequences. Test-retest repeatability was assessed using the coefficient of variation (CoV) and intra-class correlation coefficients (ICCs) were used to evaluate the intra- and inter-observer agreement RESULTS: Test-retest repeatability in the bowel wall was excellent for apparent diffusion coefficient (ADC), magnetisation transfer ratio (MTR), T1, and diffusion coefficient D (CoV 5%, 7%, 8%, and 10%, respectively), good for perfusion fraction (PF) (CoV 20%) and acceptable for T2 (CoV 21%). Inter-observer agreement was good for the T2, D and ADC (ICC = 0.89, 0.86, 0.76, respectively) and moderate for T1 (ICC = 0.55). Intra-observer agreement was similar to inter-observer agreement.

Discussion: This study showed variable results between the different parameters measured. Test-retest repeatability was at least acceptable for all parameters except pseudo-diffusion coefficient D*. Good inter- and intra-observer agreement was obtained for T2, ADC and D, with these parameters performing best in this technical validation study.