Table 1 Literature review of Melanoma classification using deep learning
From: Acral melanoma detection using dermoscopic images and convolutional neural networks
References | Dataset | Methods | Classification | Accuracy |
|---|---|---|---|---|
Antony et al. [18] | Self-collected | ANN | Melanoma | 86.66% |
Alquran et al. [6] | ISIC | SVM | Melanoma | 92.1% |
Yang et al. [4] | Self-collected | Ridge and furrow pattern | AM | 99% |
Esteva et al. [22] | Self-collected | Inception V3 | Melanoma | 72.1% |
Yu et al. [10] | Yonsei University | VGG16 | AM | 80% |
Hosny et al. [9] | PH2 | AlexNet | Melanoma | 98.33% |
Praveenkumar and Dharmalingam [19] | Self-collected | ANN | Melanoma | 97% |
Li et al. [23] | ISIC | CNNs | Melanoma | 82% |
Salido and Ruiz [24] | PH2 | Deep CNN | Melanoma | 93% |
Hosny et al. [15] | ISIC, DERM | AlexNet and augmentation | Melanoma | 87% |
Sherif et al. [25] | ISIC | Deep CNN | Melanoma | 96.57% |
Murugan et al. [8] | ISIC | SVM, kNN | Melanoma | 89.43%, 76.87% |
Kassani SH and Kassani PH [20] | ISIC | Deep CNNs | Melanoma | 92% |
Brinker et al. [21] | ISIC | CNNs | Melanoma | Specificity 86.5%, sensitivity 74.1% |