Automatic fuzzy-neural based segmentation of microscopic cell images

In this paper, we propose a novel, completely automated method for the segmentation of lymphatic cell nuclei represented in microscopic specimen images. Actually, segmenting cell nuclei is the first, necessary step for develop-ing an automated application for the early diagnostics of lymphatic system tu-mors. The proposed method follows a two-step approach to, firstly, find the nu-clei and, then, to refine the segmentation by means of a neural model, able to localize the borders of each nucleus. Experimental results have shown the fea-sibility of the method.