Skin Cancer Classification Using Convolutional Neural Network with DWT Features

The universal climate change and the ultraviolet radiation from the sun damage the skin cells. The damaged cells that could not be repaired will develop into skin cancer. The worldwide skin cancer resides in the top three positions of leading death cancer. Over the recent decades the death ratio of skin cancer is increasing. Thus, prediction and classification of the disease at early stages is mandatory for the patients survival. The artificial intelligence techniques used to automate the system that detects the skin lesions. The deep learning based convolutional neural network (CNN) approaches evidence on an automatic recognition of skin cancer more accurately. In this proposed method, the Discrete Wavelet Transformation (DWT) based features extraction techniques employed with CNN for skin cancer classification. The DWT method is used to extract the principal features from the skin cancer images after those optimal features are fed into CNN with softmax function, used to classify different types of skin cancers. This proposed method classifies different types of skin cancer more accurately. To analyze and evaluate the performance of the proposed model is tested with HAM 10000 dataset, which produces higher accuracy then the existing machine learning models.