• OBJECTIVE
    • The aim of this study was to develop a deep learning algorithm for detection of active inflammatory sacroiliitis in short tau inversion recovery (STIR) sequence MRI.
  • METHODS
    • A total of 326 participants with axial SpA, and 63 participants with non-specific back pain (NSBP) were recruited. STIR MRI of the SI joints was performed and clinical data were collected. Region of interests (ROIs) were drawn outlining bone marrow oedema, a reliable marker of active inflammation, which formed the ground truth masks from which 'fake-colour' images were derived. Both the original and fake-colour images were randomly allocated into either the training and validation dataset or the testing dataset. Attention U-net was used for the development of deep learning algorithms. As a comparison, an independent radiologist and rheumatologist, blinded to the ground truth masks, were tasked with identifying bone marrow oedema in the MRI scans.
  • RESULTS
    • Inflammatory sacroiliitis was identified in 1398 MR images from 228 participants. No inflammation was found in 3944 MRI scans from 161 participants. The mean sensitivity of the algorithms derived from the original dataset and fake-colour image dataset were 0.86 (0.02) and 0.90 (0.01), respectively. The mean specificity of the algorithms derived from the original and the fake-colour image datasets were 0.92 (0.02) and 0.93 (0.01), respectively. The mean testing dice coefficients were 0.48 (0.27) for the original dataset and 0.51 (0.25) for the fake-colour image dataset. The area under the curve of the receiver operating characteristic (AUC-ROC) curve of the algorithms using the original dataset and the fake-colour image dataset were 0.92 and 0.96, respectively. The sensitivity and specificity of the algorithms were comparable with the interpretation by a radiologist, but outperformed that of the rheumatologist.
  • CONCLUSION
    • An MRI deep learning algorithm was developed for detection of inflammatory sacroiliitis in axial SpA.