From: A two-stage deep learning architecture for radiographic staging of periodontal bone loss
 | The model | Dentists’ mean (min–max) | ||||||
---|---|---|---|---|---|---|---|---|
I | II | III/IV | Total | I | II | III/IV | Total | |
Accuracy | 0.88 | 0.79 | 0.87 | 0.77 | 0.78 (0.75–0.82) | 0.64 (0.60–0.67) | 0.76 (0.73–0.79) | 0.59 (0.54–0.64) |
Precision | 0.77 | 0.78 | 0.76 | 0.77 | 0.55 (0.49–0.65) | 0.66 (0.62–0.72) | 0.57 (0.52–0.63) | 0.6 (0.55–0.65) |
Sensitivity | 0.76 | 0.75 | 0.81 | 0.77 | 0.57 (0.54–0.59) | 0.46 (0.33–0.52) | 0.82 (0.69–0.88) | 0.6 (0.54–0.66) |
Specificity | 0.92 | 0.82 | 0.90 | 0.88 | 0.85 (0.82–0.90) | 0.79 (0.71–0.84) | 0.74 (0.67–0.83) | 0.79 (0.77–0.82) |
F 1 | 0.77 | 0.77 | 0.78 | 0.77 | 0.56 (0.51–0.62) | 0.53 (0.43–0.6) | 0.66 (0.65–0.69) | 0.59 (0.53–0.64) |