Table 1 Significant scholarly works that ML techniques to compare with the performance of APACHE II
From: Improvement of APACHE II score system for disease severity based on XGBoost algorithm
Study | Data source | Condition | Number of patients | Machine learning algorithms | Accuracy | AUC |
|---|---|---|---|---|---|---|
Samaneh Layeghian Javan et al.[19] | MIMIC III | Cardiac arrest | 4611 | Stacking algorithm | 0.76 | 0.82 |
Min Woo Kang et al.[20] | Seoul National University Hospital | Continuous renal replacement therapy | 1571 | Random forest | / | 0.78 |
Meng Hsuen Hsieh et al. [21] | Chi-Mei Medical Center | Patients with unplanned extubation in intensive care units | 341 | Random forest | 0.88 | 0.91 |
Zhongheng Zhang et al.[22] | SAILS study and OMEGA study | Acute respiratory distress syndrome | 1071 | Neural network | / | 0.821 |
Dan Assaf et al.[21] | Sheba Medical Center | Coronavirus disease (COVID-19) | 162 | Random forest | 0.92 | 0.93 |
Grupo de Trabajo Gripe A Grave et al.[23] | GETGAG/SEMICYUC database | Severe influenza | 3959 | Random forest | 0.83 | 0.82 |
Kuo-Ching Yuan et al.[24] | Taipei Medical University Hospital | Sepsis | 434 | XGBoost | 0. 82 | 0.89 |
Scherpf M et al.[39] | MIMIC III | Sepsis | 1050 | Recurrent neural network | / | 0.81 |
Zhang Z et al.[40] | MIMIC III | Acute kidney injury | 6682 | XGBoost | / | 0.86 |
Kong G et al.[41] | MIMIC III | Sepsis | 16,688 | Gradient boosting machine | Â | 0.85 |
Our work | MIMIC III | ICU patients | 24,777 | XGBoost | 0.87 | 0.81 |