Efficient Lymph Node Segmentation for Skin Cancer Biopsy Images

Abstract

In this work, we present an automated segmentation technique to segment lymph nodes in a biopsy image. The proposed technique consists of three steps. First, we perform a coarse segmentation of the image based on its histogram and identify four regions: lymph node tissue, melanoma, other tissue region and background region. We then calculate a few texture features, based on the neighborhood pixels, for each pixel in the classified image. Finally, Support Vector Machine (SVM) is applied on the extracted features to perform fine segmentation of the lymph nodes. Experimental results on 76 different lymph node images show that the proposed technique can robustly segment the lymph nodes.