Table 3.
Performances of 6 Machine Learning Models Using 7 Variables in the Validation Cohort
| Machine Learning Model | Sensitivity | Specificity | Accuracy | AUC |
|---|---|---|---|---|
| Random forest | 67.2 ± 1.6 | 71.3 ± 1.1 | 71.0 ± 0.4 | 0.75 ± 0.01 |
| Light gradient boosting machine | 72.0 ± 1.0 | 65.0 ± 0.6 | 65.9 ± 0.4 | 0.75 ± 0.00 |
| Elastic net | 68.0 ± 1.0 | 71.7 ± 0.5 | 71.0 ± 0.6 | 0.75 ± 0.01 |
| Linear support vector machine | 68.1 ± 1.0 | 71.7 ± 0.5 | 71.1 ± 0.6 | 0.75 ± 0.01 |
| Neural network | 68.9 ± 1.1 | 72.0 ± 0.5 | 71.7 ± 0.5 | 0.75 ± 0.01 |
| Naive Bayes | 35.8 ± 1.0 | 89.0 ± 0.4 | 83.0 ± 0.4 | 0.73 ± 0.01 |
The 7 variables include age, pre-existing heart failure, creatinine clearance, cardiothoracic ratio on chest x-ray, left ventricular ejection fraction, left ventricular end-systolic diameter, and left ventricular asynergy.
AUC = area under the curve.