Project Overview
Skin cancer is one of the most severe diseases, and medical imaging is among the main tools for cancer diagnosis. The images provide information on the evolutionary stage, size, and location of tumor lesions. This paper focuses on the classification of skin lesion images considering a framework of four experiments to analyze the classification performance of Convolutional Neural Networks (CNNs) in distinguishing different skin lesions. The CNNs are based on transfer learning, taking advantage of ImageNet weights. Accordingly, in each experiment, different workflow stages are tested, including data augmentation and fine-tuning optimization. Three CNN models based on DenseNet-201, Inception-ResNet-V2, and Inception-V3 are proposed and compared using the HAM10000 dataset. The results obtained by the three models demonstrate accuracies of 98%, 97%, and 96%, respectively. Finally, the best model is tested on the ISIC 2019 dataset showing an accuracy of 93%. The proposed methodology using CNN represents a helpful tool to accurately diagnose skin cancer disease.
Results Highlight
The results presented in this section correspond to the experiments performed in [1] using transfer learning on the DenseNet-201, Inception-ResNet-V2, and Inception-V3 networks. Specifically, the accuracies presented were achieved using Data Augmentation to level out the mismatch in the database and prevent overtraining with the data. For more information, please refer to the scientific paper article linked below.
| Model | Train | Validation | Test |
|---|---|---|---|
| DenseNet-201 | 1.00 | 0.98 | 0.98 |
| Inception-ResNet-V2 | 0.99 | 0.96 | 0.97 |
| Inception-V3 | 0.99 | 0.97 | 0.96 |
Products
[1] J. Villa-Pulgarin et al., “Optimized Convolutional Neural Network Models for Skin Lesion Classification,” Computers, Materials & Continua., vol. 70, pp. 2131-2148, Sep. 2021, doi: 10.32604/cmc.2022.019529.