An Early Diagnosis System for Melanoma

A multi-step system was created for early diagnosis of melanoma cancers. Image processing algorithms (edge detection and image segmentation) were used to extract the standard ABCD (Asymmetry, Border, Color, and Diameter) features of a skin mole. The extracted ABCD features were analyzed statistically to understand the impact of each characteristic. The features were then further tested in a machine learning algorithm known as Artificial Neural Networks for a comprehensive diagnosis. These combined steps provided about 80% accuracy and can successfully function as preliminary cancer diagnosis.