Microarray image processing based on clustering and morphological analysis
APBC '03 Proceedings of the First Asia-Pacific bioinformatics conference on Bioinformatics 2003 - Volume 19
Hi-index | 0.00 |
In this paper, we propose a new regular gridding and segmentation approach for microarray image. Initially, the microarray images are preprocessed using Stationary Wavelet Transform (SWT), followed by a hard thresholding filtering technique to get a de-noised microarray image. Then, we use autocorrelation to enhance the self-similarity of the image profile to get a regular gridding. Fuzzy Gaussian Mixture Model (FGMM) is used for spot segmentation. This approach has the capabilities of fitting data as generalized GMM but it can reduce about half of their computational time. Comparing probability based GMM with distance based FGMM, the latter outperforms the former in terms of computational efficiency, Due to the nature of the fast computation and nonlinear fitting of the FGMM approach. The proposed approach was evaluated using images from the Stanford Microarray Database (SMD), proved more accurate in intensity computation and more reliable means for estimating gene expression than conventional methods.