Deep Learning Can Improve Early Skin Cancer Detection

Skin cancer is the most common form of affecting humans. Melanoma dangerous type skin cancer; and early diagnosis extremely vital in curing disease. So far, human knowledge this field very limited, thus, developing a mechanism capable identifying disease on can save lives, reduce intervention cut unnecessary costs. In paper, researchers developed new learning technique to classify lesions, with purpose observing presence melanoma.  This based convolutional neural network solution multiple configurations; where employed an International Imaging Collaboration (ISIC) dataset. Optimal results are achieved through composed 14 layers. proposed system successfully reliably predict correct classification dermoscopic lesions 97.78% accuracy.