Table 4.
Predictive performance metrics for ML models trained on T1w and T2w for training and validation sets.
| AUC | Accuracy | Sensitivity | Specificity | PPV | NPV | ||
|---|---|---|---|---|---|---|---|
| XGBoost | Training 95% CI |
0.68 (0.60- 0.75) |
0.66 (0.60 - 0.73) |
0.42 (0.31 - 0.53) |
0.84 (0.77 - 0.90) |
0.66 (0.52- 0.78) |
0.67 (0.59 - 0.74) |
| Test 95% CI |
0.72 (0.56 - 0.87) |
0.75 (0.63 - 0.85) |
0.62 (0.38 - 0.83) |
0.80 (0.66 - 0.89) |
0.53 (0.31 - 0.73) |
0.85 0.72 - 0.94) |
|
|
Regularized Logistic
Regression |
Training 95% CI |
0.63 (0.55 - 0.70) |
0.60 (0.54 - 0.67) |
0.57 (0.46 - 0.67) |
0.63 (0.54 - 0.72) |
0.53 (0.42 - 0.63) |
0.67 (0.58 - 0.76) |
| Test 95% CI |
0.52 (0.36 - 0.68) |
0.41 (0.30 - 0.54) |
0.56 (0.33 - 0.78) |
0.36 (0.23 - 0.51) |
0.24 (0.13 - 0.40) |
0.69 (0.49 - 0.85) |
|
|
Random
Forest |
Training 95% CI |
0.99 (0.99 - 1.0) |
0.97 (0.95 - 0.99) |
0.98 (0.94 - 1.0) |
0.97 (0.93 - 0.99) |
0.96 (0.90 - 0.99) |
0.99 (0.96 - 1.0) |
| Test 95% CI |
0.51 (0.33 - 0.70) |
0.38 (0.27 - 0.51) |
0.68 (0.44 - 0.87) |
0.27 (0.16 - 0.42) |
0.25 (0.14 - 0.40) |
0.70 (0.47 - 0.88) |
|
| SVM | Training 95% CI |
0.54 (0.46 - 0.63) |
0.60 (0.54 - 0.67) |
0.21 (0.14 - 0.32) |
0.90 (0.82 - 0.94) |
0.59 (0.41 - 0.75) |
0.61 (0.53 - 0.68) |
| Test 95% CI |
0.61 (0.45 - 0.76) |
0.68 (0.56 - 0.79) |
0.25 (0.09 - 0.49) |
0.84 (0.71 - 0.93) |
0.36 (0.14 - 0.65) |
0.75 (0.62 - 0.86) |
|