Detecting melanoma in dermoscopy images using scale adaptive local binary patterns

Recent advances in the area of computer vision has led to the development of various assisted diagnostics systems for the detection of melanoma in the patients. Texture and color are considered as two fundamental visual characteristics which are vital for the detection of melanoma. This paper proposes the use of a combination of texture and color features for the classification of dermoscopy images. The texture features consist of a variation of local binary pattern (LBP) in which the strength of the LBPs is used to extract scale adaptive patterns at each pixel, followed by the construction of a histogram. For color feature extraction, we used standard HSV histograms. The extracted features are concatenated to form a feature vector for an image, followed by classification using support vector machines. Experiments show that the proposed feature set exhibits good classification performance comparing favorably to other state-of-the-art alternatives.

I. Introduction

Melanoma is considered one of the deadliest forms of skin cancer having a very high mortality rate [1]. The incidence rate for melanoma has been increasing over the past decade, making it one of the domains that has been attracting the attention of the cancer research community. It has been found that the early detection of melanoma in the patients can significantly increase the survival rates for the patients [2]. Dermoscopy is a clinical procedure, that is used by the dermatologists to diagnose the skin pathologies. It is a noninvasive procedure that is used for the in vivo observation of the skin lesions. In a dermoscopy procedure, the physicians apply gel to the skin of the patients and inspect the skin with a magnification instrument, known as dermoscope allowing the inspection of surface and sub-surface structures which are not visible to the naked eye [3]. The diagnosis of dermoscopy images is carried out using an assessment of these structures for which several medical rules have been defined, the most common of which are the ABCD rule, Menzies' method and the seven point check list [3]. Studies have shown that the application of these rules has resulted in an effective early detection of melanoma.