Lymph node image segmentation based on Fuzzy c-Means clustering and an improved chan-vese model

The quality of lymph node images is very important for the doctor to do the pathological analysis. For the fuzziness and uncertainty of the edge, the shape and size of lymph nodes, we propose Fuzzy c-Means (FCM) peak clustering which sharpens blurry edges and the improved Chan-Vese (CV) model that enhances detection performances to the noises and fuzzy boundaries. Validation experiments are implemented on mass clinical images. We take the manual segmentation by a medical expert as a standard. Experiment results show that the proposed method can segment the blurry edges of lymph node images quickly and efficiently compared to the traditional CV model.