DNA microarray image processing based on minimum error segmentation and histogram analysis

Abstract

DNA microarray are an experimental technology which consists in arrays of thousands of discrete DNA sequences that are printed on glass microscope slides and allows the monitoring of expressions for tens of thousands of genes simultaneously. Image analysis is an important aspect for microarray experiments that can affect subsequent analysis such as identification of differentially expressed genes. The aim of this step is to extract the gene expression data included in the spots image. Imge processing for microarray images includes three tasks: spot gridding, segmentation and information extraction. In this study, we address the segmentation and information extraction problems, and propose a new segmentation method and a new background and foreground segmentation correction method for accurate information extraction. The initial segmentation is based on minimum error thresholding under the assumption that the probability density distribution of spot image and background image satisfies Gaussian and the final results is obtained through refining initial segmentation by Bayes decision theory. The advantage of our method is that it does not have any restrictions on the spot shape. We compare our experimental results with those obtained from the widely used software GenePix.