Related Experiment Video
Updated: Mar 29, 2026
Software-Assisted Quantitative Measurement of Osteoarthritic Subchondral Bone Thickness
Published on: March 18, 2022
A diffusion-based quantification technique for assessment of sacroiliitis in adolescents with enthesitis-related
Kanimozhi Vendhan1, Timothy J P Bray1,2, David Atkinson1
11 Academic Radiology, University College London Centre for Medical Imaging, London, UK.
Objective:
To investigate the use of a quantitative diffusion-weighted imaging (DWI) tool for measuring inflammation of the sacroiliac joints (SIJs) in enthesitis-related arthritis (ERA).
Methods:
A retrospective study was performed with institutional review board approval. Subjects were adolescents who had undergone MRI of the SIJs since January 2010. 10 patients with a clinical diagnosis of ERA and 10 controls with a clinical diagnosis of mechanical back pain were assessed. Axial T1 weighted, short tau inversion recovery (STIR) and DWI (b-values 0, 50, 100, 300 and 600 mm(2) s(-1)) images were acquired. Apparent diffusion coefficient (ADC) maps were generated using a monoexponential fit. On each of four slices, two to three linear regions-of-interest were placed on each joint. Normalized ADC (nADC) values were defined as joint ADC divided by a reference ADC derived from normal sacral bone. STIR images were scored using a modification of an established technique. The correlation between nADC values and STIR scores was evaluated using Spearman's rank correlation.
Results:
Mean nADC values were significantly higher in cases than in controls (p = 0.0015). There was a strong correlation between STIR scores and nADC values (R = 0.85).
Conclusion:
ADC values are significantly increased in inflamed SIJs compared with controls. There is a good correlation between this diffusion-based method and STIR scores of inflammation.
Advances In Knowledge:
We have described and provisionally validated a method for quantifying the severity of inflammation in the SIJs in ERA using ADC measurements. This method is quick, is reproducible and could potentially be automated.
