Table 3.
Predictive power of the ML model and routine clinical models using data of the validation cohort.
| Model | Specificity | Sensitivity | Accuracy | NPV | PPV | Recall | Youden’s index |
|---|---|---|---|---|---|---|---|
| Machine learning model | 64.4% | 100.0% | 70.5% | 100.0% | 36.6% | 100.0% | 1.644 |
| ALBI | 52.1% | 73.3% | 55.7% | 90.5% | 23.9% | 73.3% | 1.254 |
| FIB-4 | 86.3% | 53.3% | 80.7% | 90.0% | 44.4% | 53.3% | 1.396 |
| APRI | 67.1% | 53.3% | 64.8% | 87.5% | 25.0% | 53.3% | 1.205 |
| MELD | 84.9% | 33.3% | 76.1% | 86.1% | 31.3% | 33.3% | 1.183 |
| CTP | 0.0% | 100.0% | 17.0% | – | 17.0% | 100.0% | 1.000 |
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.