Melanoma Analysis using Dermoscopy Images

-Melanoma is the most dangerous type of skin cancer. Uncontrolled growth of cancer cells are mostly occurs in melanocytes. In spite of, to prevent its deadly consequences and for treatment to be efficient, to perform the early diagnostic of the melanoma efficient treatment is necessary. The diagnostic of melanoma, without any support is not possible, therefore dermoscopy was created. Dermoscopy is a device used for examination of skin lesions with a dermatoscope. Dermoscopy images are collected from Dermweb. The standard approach in automatic dermoscopic image analysis has three stages: (i) Image segmentation (ii) Feature extraction and feature selections (iii) Lesion classifications. Skin lesions are classified using SVM (Support Vector Machine) classifier. The skin lesions may be Melanoma, BCC, SK and Nevus. KEYWORDS--melanoma; dermoscopy; SVM classifier; segmentation. I.INTRODUCTION Melanoma is the dangerous skin cancer and accounts for about 75% of deaths associated with skin cancer. To improve the diagnostic performance of melanoma, Dermoscopy technique has been developed. Dermoscopy is a noninvasive skin imaging technique of acquiring a magnified and illuminated image of a region of skin for increased clarity of the spots on the skin. It enhances the visual effect of skin lesion by removing surface reflection of skin. Skin cancer is defined as the uncontrolled growth of cells in the skin due to the spreading of skin cancer cells rapidly and it forms the malignant tumor. The main causes of skin cancer is the over exposure of ultraviolet radiation from sunshine, genetic defects. Skin cancer can be mainly classified as three types such as Basal cell carcinoma (BCC), Melanoma, and Squamous cell carcinoma (SCC). BCC and SCC are called as nonmelanomas. Uncontrolled growth of lesions that rise in basal cells in the skin is called as BCC. Squamous-cell carcinoma (SCC) is a cancer which occurs in squamous cell. It spreads faster to other parts in the body. The virulent form of skin cancer is melanoma which arises in the melanocytes cell. The paper [1], [2] notifies the segmentation based on thresholding. By using a certain threshold value, the lesion will be segmented i.e., 1’s denotes the lesion foreground and 0’s denotes the lesion background by comparing each pixel with certain threshold value. Analyzing dermoscopy images have a high differentiation between lesion and skin of the images. Improved internet based melanoma detection cannot be applied for multimodal images. Several methods are proposed to diagnose the skin lesions for dermoscopy images, like that of the ABCD rule of dermoscopy, 7-point checklist, ELM pattern analysis and CASH algorithm [3]. Previous algorithms are not having the ability to withstand the diagnosis results are difficult. Therefore, an Automatic CAD (Computer Aided Diagnosis) tools are needed, because to achieve the analysis of pigmented skin lesions for dermoscopy images. An automatic method to segment the lesion was proposed by Nikhil Cheerlead al. [4]. Otsu and local binary pattern (LBP) methods are used for segmenting the texture. Neural network classifier was used, which achieves 97% for sensitivity and 93% for specificity. But it doesn’t consider as other type of skin lesions such