Skin Cancer Diagnosis using Support Vector Machine based on Gray Level Co-occurrence Matrix

Abstract

As today's world is more developed than ever, the way to diagnose the diseases is better and more precise. Cancer has been identified as one of the leading causes of death. In this study, features from dermoscopy pictures were extracted using the Local Binary Pattern (LBP) and Gray Level Co-Occurrence Matrix (GLCM) methods. Melanoma, the most dangerous type of skin cancer, or benign skin tumors were then classified using Support Vector Machine (non-cancerous). Five features, contrast, dissimilarity, homogeneity, energy and correlation were extracted by GLCM. Radial Basis Function (RBF) Kernel of SVM was trained with the features and, then tested the images and classified whether the cancer or not. Performance was evaluated with the confusion matrix by testing accuracy, specificity, sensitivity and precision. Using image processing software, this study proposes a technique for locating skin cancer. In order to identify whether skin cancer is present, the system first gets an image of a skin lesion as input and analyzes it using image processing methods. The contrast, homogeneity, dissimilarity, energy, and correlation analysis are all checked by the Lesion Image analysis tools during the image segmentation and feature stages. The image is categorized as either a benign or malignant cancer lesion using the derived feature parameters.