Table 2 Performance of the three optimised supervised models, asterisks indicate histological features.
From: Classification of Paediatric Inflammatory Bowel Disease using Machine Learning
Input | Accuracy % (AUC) | Precision | Recall | F1-score | (#) Features |
|---|---|---|---|---|---|
Endoscopy | 71.0% (0.78) | 0.89 | 0.68 | 0.75 | (5) Duodenum, Ileum, D-Colon, Rectum, Perianal |
Histology | 76.9% (0. 82) | 0.81 | 0.86 | 0.83 | (1) Ileum |
Combined (E + H) | 82.7% (0.87) | 0.91 | 0.83 | 0.87 | (8) Duodenum, Ileum, D-Colon, Rectum, Perianal, Oesophagus*, Ileum*, A-Colon* |
