Table 9.
Summary of previous studies on skin disease detection with their performance.
| Author and Year | Method | Dataset | Dataset Size | Acc (%) | Sn (%) | Sp (%) | R (%) | P (%) | F1 (%) |
|---|---|---|---|---|---|---|---|---|---|
| Dwived et al. [67] | Fast R-CNN | - | - | 90 | - | - | - | ||
| Alam and Jihan et al. [68] | DL+ Image Processing | - | - | 85.14 | - | - | - | - | - |
| Wan et al. [69] | Fusion Network | HAM10000 | 10,015 | 92.01 | - | 89.53 | - | - | 88.94 |
| Kumar and Vanmathi et al. [70] | U-Net + RP-Net | - | - | 92.5 | 92.1 | 91.7 | - | - | - |
| Suicmez et al. [71] | Hybrid CNN-Gradient Boost Classifier | HAM10000 | 10,015 | 99.4 | 99.4 | - | 99.4 | 99.4 | 99.4 |
| Suicmez et al. [71] | Hybrid CNN-Machine Learning | ISIC 2020 | - | 100 | 100 | - | 100 | 100 | 100 |
| Lembhe et al. [73] | VGG16: ISR | ISIC | 70.17 | 69 | - | - | 68 | 73 | - |
| Lembhe et al. [73] | ResNet: ISR | ISIC | 86.57 | 87 | - | - | 87 | 87 | - |
| Lembhe et al. [73] | Inception V3: ISR | ISIC | 91.26 | 92 | - | - | 89 | 92 | - |
| Priyadharshini et al. [74] | ELM- TLBO | Kaggle and DermIS | 300 | - | - | - | 92.45 | 89.72 | 91.64 |
| Dandu et al. [75] | Ensemble Classifier | SIIM ISIC | - | 86.38 | - | - | 86.50 | 86.16 | - |