Table 3 Comparison of prediction performance of three machine learning models trained using tumor-related features (Model 1), emphysema-related features (Model 2), and combined image features (Model 3)
From: Developing global image feature analysis models to predict cancer risk and prognosis
Performance index | Model 1 | Model 2 | Model 3 |
|---|---|---|---|
Sensitivity | 50.0% | 57.7% | 69.2% |
Specificity | 82.7% | 76.5% | 87.7% |
Positive predictive value | 48.1% | 44.1% | 64.3% |
Negative predictive value | 83.8% | 84.9% | 89.9% |
Overall accuracy | 74.8% | 72.0% | 83.2% |