cDNA Microarray Image Segmentation Using Shape-Adaptive DCT and K-means Clustering

The cDNA microarray technology provides a powerful analytic tool for human genetic research and drug discovery. Image processing plays a crucial role in the extraction and quantitative analysis of the relative abundance of the DNA (Deoxyribonucleic acid) product. The microarray images exhibit variations due to noise impairments. This chapter presents a novel filtering method called Shape-Adaptive DCT (Discrete Cosine Transform), which does well in filtering cDNA microarray images. The experimental result processing by Shape-Adaptive DCT is compared with those obtained from the widely used filtering approaches. Simulation studies reported in this chapter indicate that the proposed filtering method and segmentation using the K-means algorithm yield excellent performance and efficiently suppress noise in cDNA microarray data. The result of the experiment shows its robustness and precision.