fusion of dermoscopic datasets: a method to promote performance ofdeep learning in melanoma detection

Skin cancer is one of the diseases that has spread a lot in recent years. the abnormal growth ofskin cells causes this cancer. if this disease is diagnosed in the early stages, it can be easily treatedand prevent the possible death of the patient. therefore, artificial intelligence experts have madegreat efforts to identify and diagnose this disease with the help of computer-aided detection (cad)systems. based on this, machine learning-based approaches have become an efficient way toclassify skin lesions. this paper presents a cad system for classifying melanoma and nevus. weused the combination of two isic 2019 and isic 2020 datasets to overcome the data imbalance.a pre-trained model with extra layers is used to feature extraction and classify skin lesions. theproposed method achieves 94% sensitivity, 97% specificity, and 95% accuracy on test data.