Automated Malignant Melanoma Classification Using Convolutional Neural Networks

This research is proposed a design of architecture for melanoma (a kind skin cancer) recognition by using Convolutional Neural Network (CNN), work that will be useful researchers in future projects areas like biomedicine, machine learning, and others related moving forward with their studies improving this proposal. CNN mostly used computer vision branch artificial intelligence), applied to pattern moles determine the existence malignant melanoma, or not, limited dataset. The classifier designed trained case was built through couple layers convolution pooling stacked form neural network 6 followed fully connected complete an output classifier. database train our largest publicly collection dermoscopic images melanomas other lesions, provided International Skin Imaging Collaboration (ISIC), sponsored Society Digital (ISDIS), international effort improve diagnosis. purpose high level accuracy help professionals medicine diagnosis, case, it possible get up 88.75 %.