An Automatic and Robust Method for Microarray Image Analysis and the Related Information Retrieval for Microarray Databases

DNA microarray is an increasingly important technique that allows biologists to monitor expression levels of thousands of genes in parallel. This technique is widely used in biological research for studying genomic aberrations in cancer and other genetic diseases, and therefore, it has great potential for clinical diagnostics in the future. However, there are still several critical problems with this technology. In this study, we focus on two important issues. The first issue is related to automatic gridding and spot segmentation for microarray images. It has been reported that the quality of spot segmentation significantly influences data precision in the subsequent data analysis. However, nowadays microarray image analysis software still requires users´ fine tuning to obtain acceptable results. Another important issue is how to automatically collect related information regarding all genes on the microarray slide for subsequent data analysis and data mining. To relieve researchers from manually correcting image processing results and manually collecting the related information for genes, in this paper, we proposed an automatic and robust method for microarray image analysis and the related information retrieval module which is integrated with the proposed database schema for microarray image data.