Table 4 Performance comparison of clinical, radiomics, and fusion models in training and testing cohorts of dual-phase CT.
From: Machine learning models using dual-phase CT radiomics for early detection of PRISm
Cohorts | Models | AUC(95%CI) | ACC | Sensitivity | Specificity | PPV | NPV |
|---|---|---|---|---|---|---|---|
Training cohort | Clinical model | 0.825(0.750–0.900) | 0.741 | 0.607 | 0.857 | 0.810 | 0.703 |
Inspiratory-radiomic model | 0.826(0.753–0.900) | 0.767 | 0.768 | 0.757 | 0.754 | 0.780 | |
Clinical-Inspiratory-radiomic model | 0.890(0.834–0.947) | 0.828 | 0.679 | 0.957 | 0.950 | 0.763 | |
Expiratory-radiomic model | 0.838(0.768–0.908) | 0.776 | 0.589 | 0.950 | 0.917 | 0.712 | |
Clinical-Expiratory-radiomic model | 0.901(0.845–0.956) | 0.836 | 0.750 | 0.917 | 0.894 | 0.797 | |
Dual-phase radiomic model | 0.851(0.783–0.918) | 0.767 | 0.607 | 0.917 | 0.872 | 0.714 | |
Clinical-Dual-phase-radiomic model | 0.901(0.846–0.955) | 0.836 | 0.786 | 0.883 | 0.863 | 0.815 | |
Internal validation cohort | Clinical model | 0.771(0.639–0.903) | 0.681 | 0.812 | 0.613 | 0.520 | 0.864 |
Inspiratory-radiomic model | 0.800(0.642–0.959) | 0.830 | 0.625 | 0.935 | 0.833 | 0.829 | |
Clinical-Inspiratory-radiomic model | 0.810(0.677–0.944) | 0.809 | 0.812 | 0.806 | 0.684 | 0.893 | |
Expiratory-radiomic model | 0.798(0.642–0.955) | 0.809 | 0.625 | 0.903 | 0.769 | 0.824 | |
Clinical-Expiratory-radiomic model | 0.819(0.680–0.957) | 0.830 | 0.750 | 0.871 | 0.750 | 0.871 | |
Dual-phase radiomic model | 0.796(0.638–0.955) | 0.809 | 0.687 | 0.871 | 0.733 | 0.844 | |
Clinical-Dual-phase-radiomic model | 0.821(0.684–0.957) | 0.787 | 0.812 | 0.774 | 0.650 | 0.889 | |
External validation cohort | Clinical model | 0.778(0.653–0.904) | 0.626 | 0.581 | 0.929 | 0.982 | 0.250 |
Inspiratory-radiomic model | 0.757(0.645–0.870) | 0.551 | 0.484 | 1.000 | 1.000 | 0.226 | |
Clinical-Inspiratory-radiomic model | 0.845(0.732–0.958) | 0.888 | 0.925 | 0.643 | 0.945 | 0.562 | |
Expiratory-radiomic model | 0.800(0.702–0.899) | 0.682 | 0.645 | 0.929 | 0.984 | 0.283 | |
Clinical-Expiratory-radiomic model | 0.817(0.695–0.940) | 0.916 | 0.968 | 0.571 | 0.937 | 0.727 | |
Dual-phase radiomic model | 0.810(0.719–0.900) | 0.738 | 0.710 | 0.929 | 0.985 | 0.325 | |
Clinical-Dual-phase-radiomic model | 0.813(0.694–0.932) | 0.626 | 0.581 | 0.929 | 0.982 | 0.250 |
