Table 2 The metrics of the segmentation model for visual impairment and corneal perforation tasks.
From: Establishment of a corneal ulcer prognostic model based on machine learning
Model | After 1 month | After 1 month | After 1 month | After 1 month | After 3 months | After 3 months | After 3 months | After 3 months |
|---|---|---|---|---|---|---|---|---|
Name | XGBoost | XGBoost | XGBoost | XGBoost | XGBoost | XGBoost | LightGBM | LightGBM |
Task | Corneal perforation-train | Corneal perforation-test | Improvement of vision-train | Improvement of vision-test | Corneal perforation-train | Corneal perforation-test | Improvement of vision-train | Improvement of vision-test |
Accuracy | 0.97 | 0.85 | 0.95 | 0.68 | 0.97 | 0.91 | 0.97 | 0.97 |
AUC | 0.99 | 0.81 | 0.99 | 0.77 | 0.99 | 0.97 | 0.99 | 0.98 |
95%CI | 0.98–1.00 | 0.63–1.00 | 0.99–1.00 | 0.63–0.91 | 0.99–1.00 | 0.92–1.00 | 0.98–1.00 | 0.94–1.00 |
Sensitivity | 1.00 | 0.71 | 0.97 | 0.62 | 1.00 | 1.00 | 0.96 | 0.90 |
Specificity | 0.92 | 0.87 | 0.98 | 0.80 | 0.96 | 0.93 | 1.00 | 1.00 |
PPV | 0.95 | 0.50 | 0.94 | 0.73 | 0.92 | 0.66 | 0.97 | 0.96 |
NPV | 0.97 | 0.86 | 0.97 | 0.63 | 0.98 | 0.94 | 0.95 | 1.00 |
Precision | 0.95 | 0.50 | 0.94 | 0.73 | 0.92 | 0.66 | 0.97 | 0.96 |
Recall | 1.00 | 0.71 | 0.97 | 0.62 | 1.00 | 1.00 | 0.96 | 0.90 |
F1 | 0.97 | 0.58 | 0.95 | 0.67 | 0.96 | 0.80 | 0.97 | 0.93 |
