From: Automatic segmentation of fat metaplasia on sacroiliac joint MRI using deep learning
Clinical characteristics | All (n = 123) | Training set (n = 103) | Test set (n = 20) | p value |
---|---|---|---|---|
Age (years) | 26.0 (23.0, 32.0) | 27.0 (24.0, 32.0) | 23.0 (19.3, 30.8) | 0.190 |
Disease duration (months) | 24.00 (7.0, 72.0) | 24.00 (8.0, 72.0) | 13 (1.0, 75.0) | 0.038 |
Sex | Â | Â | Â | 0.797 |
 Male | 89 (72.36) | 75 (72.82) | 14 (70.0) |  |
 Female | 34 (27.64) | 28 (27.18) | 6 (30.0) |  |
HLA-B27 | Â | Â | Â | 0.653 |
  ( +) | 80 (65.0) | 70 (67.9) | 10 (50.0) |  |
  ( −) | 25 (20.3) | 21 (20.4) | 4 (20.0) |  |
Missing | 18 (14.7) | 12 (11.7) | 6 (30.0) | Â |
ESR (mm/H) | 20.0 (10.0, 40.0) | 20.0 (9.0, 35.0) | 43.5 (17.3, 62.0) | 0.027 |
CRP (mg/L) | 7.0 (2.7, 22.1) | 6.9 (2.2, 19.1) | 24.7 (4.9, 40.7) | 0.158 |