Study of Melanocytic Nevi using image processing

Melanocytic nevi are benign pigmented skin lesions associated with increased risk of transforming into malignant lesions. Most of the malignant lesions are misclassified as Clark Nevi based on the dermoscopic criteria. Histopathological analysis forms a gold standard. However, this leads to unnecessary biopsies of benign lesions. Thus the identification of benign lesion from malignant lesions plays a prominent role in diagnosis of melanoma. Although significant progress has been made in developing automated diagnostic tools, there exists a need for developing reliable non-invasive diagnostic tools. This can be achieved by incorporating the detection of dermoscopic features based on their histopathological relevance. This study uniquely provides the histopathological correlations of dermoscopic features to facilitate reliable non-invasive diagnosis. The relevance of the dermoscopic features are provided. It is inferred that shape, color and dermoscopic features such as blue-white veil and pigment network would contribute significantly to the field of melanoma diagnosis.