Table 4 Performance of segmentation CNNs (implemented using TL).
From: COVID-19 infection analysis framework using novel boosted CNNs and radiological images
Model | Region | DSC | Acc | IoU | BF | Global-Acc | Mean-Acc | Mean-IoU | Weighted-IoU | Mean-BF |
|---|---|---|---|---|---|---|---|---|---|---|
Deeplabv3 | Infected | 95.35 | 98.76 | 98.03 | 97.71 | 99.23 | 99.15 | 98.40 | 98.48 | 97.19 |
Background | 98.65 | 99.53 | 98.76 | 96.66 | ||||||
U-SegNet | Infected | 95.20 | 98.41 | 97.52 | 97.46 | 98.93 | 98.81 | 98.10 | 98.25 | 96.94 |
Background | 98.60 | 99.67 | 98.61 | 96.32 | ||||||
SegNet | Infected | 95.10 | 98.29 | 97.70 | 97.41 | 99.10 | 98.95 | 98.09 | 98.22 | 96.82 |
Background | 98.10 | 99.62 | 98.55 | 96.23 | ||||||
U-Net | Infected | 94.90 | 98.74 | 97.62 | 98.19 | 99.22 | 99.06 | 98.06 | 98.15 | 96.63 |
Background | 98.20 | 99.07 | 98.49 | 95.86 | ||||||
VGG-16 | Infected | 94.80 | 98.29 | 97.07 | 97.11 | 98.89 | 98.79 | 97.61 | 97.74 | 96.18 |
Background | 97.90 | 99.23 | 98.24 | 95.29 | ||||||
FCN-8 | Infected | 94.70 | 98.29 | 97.04 | 97.08 | 98.87 | 98.76 | 97.59 | 97.71 | 96.16 |
Background | 98.00 | 99.21 | 98.15 | 95.27 |
