IEEE Transactions on Pattern Analysis and Machine Intelligence
Texture Synthesis by Non-Parametric Sampling
ICCV '99 Proceedings of the International Conference on Computer Vision-Volume 2 - Volume 2
Can graph-cutting improve microarray gene expression reconstructions?
Pattern Recognition Letters
Hi-index | 0.00 |
Microarrays allow biologists to determine the gene expressions for tens of thousands of genes simultaneously, however due to biological processes, the resulting microarray slides are permeated with noise. During quantification of the gene expressions, there is a need to remove a gene's noise or background for purposes of precision. This paper presents a novel technique for such a background removal process. The technique uses a gene's neighbour regions as representative background pixels and reconstructs the gene region itself such that the region resembles the local background. With use of this new background image, the gene expressions can be calculated more accurately. Experiments are carried out to test the technique against a mainstream and an alternative microarray analysis method. Our process is shown to reduce variability in the final expression results.