Detection of skin cancer from dermoscopic images using image processing and deep learning methods

This study examines the effectiveness of deep learning-based methods in skin cancer detection. Skin cancer is one of the most common types of cancer worldwide, with melanoma having a particularly high mortality rate. In this thesis, deep learning models were used for the automatic classification of skin lesions and melanoma detection. The ISIC 2019 dataset was utilized for this study. Popular architectures such as VGG16, ResNet50, and Vision Transformer were used to compare model performances. The models were evaluated based on metrics such as accuracy, precision, specificity, and F1 score. The results obtained indicate that the proposed models provide high accuracy in skin cancer detection.