Table 2 Performance Metrics for External Validation.
From: A Custom Annotated Dataset for Segmentation of Pulmonary Veins, Arteries, and Airways
Dataset | Class | Count | Dice | Recall | Precision | BDR | DLR |
|---|---|---|---|---|---|---|---|
ATM’22 | Airway Lumen | 289 | 0.821 ± 0.044 (0.522–0.904) | 0.977 ± 0.012 (0.930–0.998) | 0.711 ± 0.064 (0.360–0.845) | 0.987 ± 0.021 (0.750–1.000) | 0.988 ± 0.011 (0.909–1.000) |
Parse2022 | Pulmonary Arteries | 100 | 0.802 ± 0.035 (0.685–0.879) | 0.907 ± 0.053 (0.755–0.985) | 0.726 ± 0.071 (0.525–0.909) | — | — |
HiPas | Pulmonary Arteries | 249 | 0.816 ± 0.041 (0.664–0.892) | 0.825 ± 0.072 (0.547–0.919) | 0.815 ± 0.058 (0.662–0.938) | — | — |
HiPas | Pulmonary Veins | 249 | 0.789 ± 0.042 (0.620–0.858) | 0.794 ± 0.096 (0.508–0.951) | 0.805 ± 0.099 (0.574–0.969) | — | — |
