Table 4.
Predictive power of the ML model and routine clinical models using data of the testing cohort.
| Models | Specificity | Sensitivity | Accuracy | NPV | PPV | Recall | Youden’s index |
|---|---|---|---|---|---|---|---|
| Machine learning model | 64.4% | 87.5% | 69.7% | 94.6% | 42.2% | 87.5% | 1.519 |
| ALBI | 77.8% | 55.0% | 72.6% | 85.4% | 42.3% | 55.0% | 1.328 |
| FIB-4 | 85.2% | 40.0% | 74.9% | 82.7% | 44.4% | 40.0% | 1.252 |
| APRI | 79.3% | 45.0% | 71.4% | 82.9% | 39.1% | 45.0% | 1.243 |
| MELD | 90.4% | 25.0% | 75.4% | 80.3% | 43.5% | 25.0% | 1.154 |
| CTP | 94.8% | 15.0% | 76.6% | 79.0% | 46.2% | 15.0% | 1.098 |
ML, machine learning; NPV, negative predictive value; PPV, positive predictive value; ALBI, albumin-bilirubin grade; FIB-4, fibrosis 4 score; APRI, aspartate transaminase to platelet ratio; MELD, model for end-stage liver disease; CTP, Child-Turcotte-Pugh.