Edge-based Active Contours for Microarray Spot Segmentation

The paper proposes a novel approach for image segmentation in case of bio-medical applications. The medical images considered for our analysis are drawn from microarray image databases and they are generally used for the estimation of gene expression levels, which correspond to the average intensities of the microarray spots represented as circular shapes within the image under analysis. The novelty of the proposed image analysis approach consists of a segmentation procedure which uses an active contour model (ACM) based on edge information. Basically, a predefined curve is evolved towards the object edges (i.e. microarray spots). For a precise determination of image object, a cellular neural network approach (CNN) is employed for edge detection of each microarray spot. This is used further on in the curve evolution process for a more accurate segmentation of microarray spots. Specific measures for the characterization of microarray spots features are used to compare the obtained results with the ones delivered by GenePix microarray software platform. (C) 2021 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (https://crativecommons.org/licenses/by-nc-nd/4.0) Peer-review under responsibility of the scientific committee of KES International.