Malignant Melanoma is a type of cancerous cell which occurs in the skin and it is known as the deadliest form of all skin cancers. They sometimes develop from a mole with increase in size, irregular edges,change in color, itchiness or skin breakdown. If detected early, malignant melanoma can be treated successfully. In this paper, detection of the malignant melanoma is done by artificial neural network trained with particle swarm optimization algorithm. The features of the image are extracted using SIFT (Scale invariant feature transform) and SURF (Speeded Up Robust Features) techniques. Once the dominant characteristics are identified it is fed into the ANN(Artificial Neural Networks) for classification of the tumor images as malignant or benign. Simulation results indicate that the neural network trained with particle swarm optimization algorithm provides a good accuracy in the classification.
