Bayesian modelling of microarray images

We examine the use of Bayesian signal processing to improve the modelling of microarray images, and ultimately the estimation of gene expression ratios. A novel elliptical spot shape model is presented, with a Bayesian image modelling method. Prior knowledge from neighbouring spots is encompassed in the framework of a Markov random field, potentially enhancing the accuracy and reliability of ratio estimates. The techniques may be particularly beneficial for irregular, overlapping, damaged, saturated, or weakly expressed spots. ©2006 IEEE