Table 3 Effects of data generated by GACDN With baselines
From: GACDN: generative adversarial feature completion and diagnosis network for COVID-19
| Â | ACC (%) | SEN (%) | SPE (%) | AUC (%) |
|---|---|---|---|---|
Radiomic | Â | Â | Â | Â |
LR | 87.74 ± 1.20 | 90.34 ± 1.86 | 83.90 ± 3.07 | 87.12 ± 1.36 |
SVM | 86.59 ± 1.93 | 89.30 ± 2.64 | 82.72 ± 3.57 | 86.01 ± 1.97 |
KNN | 85.76 ± 3.27 | 90.04 ± 2.19 | 79.72 ± 6.96 | 84.88 ± 3.68 |
NN | 87.19 ± 1.25 | 84.23 ± 2.75 | 91.60 ± 2.95 | 87.91 ± 1.22 |
Radiomic + GACDN |  |  |  |  |
LR | 90.83 ± 1.25 | 92.41 ± 1.86 | 88.52 ± 2.63 | 90.46 ± 1.29 |
SVM | 89.13 ± 2.06 | 90.24 ± 2.77 | 87.62 ± 3.12 | 88.93 ± 2.10 |
KNN | 89.61 ± 1.78 | 91.97 ± 2.54 | 86.19 ± 2.09 | 89.08 ± 1.64 |
NN | 90.23 ± 1.43 | 90.44 ± 1.79 | 90.98 ± 2.25 | 91.12 ± 1.36 |