Methods for automatic microarray image segmentation

Abstract

This paper describes image processing methods for automatic spotted microarray image analysis. Automatic gridding is important to achieve constant data quality and is, therefore, especially interesting for large-scale experiments as well as for integration of microarray expression data from different sources. We propose a Markov random field (MRF) based approach to high-level grid segmentation, which is robust to common problems encountered with array images and does not require calibration. We also propose an active contour method for single-spot segmentation. Active contour models describe objects in images by properties of their boundaries. Both MRFs and active contour models have been used in various other computer vision applications. The traditional active contour model must be generalized for successful application to microarray spot segmentation. Our active contour model is employed for spot detection in the MRF score functions as well as for spot signal segmentation in quantitative array image analysis. An evaluation using several image series from different sources shows the robustness of our methods.

A Comparison of Fuzzy Clustering Approaches for Quantification of Microarray Gene Expression

“…Spot Segmentation-After some necessary preprocessing steps such as the registration of two channel arrays and spot finding, a critical problem in microarray image processing is spot segmentation [14][15][16]. Spots segmentation is an approach to determine which pixels in the target region are due to the actual spot signal, and which pixels are due to background.…”

“…The mean background intensities α will be estimated by the proposed algorithms and then the intensities of test and reference signals will be estimated by (22) From this value, the log ratio is calculated using one of the statistics of Eqs. (15)(16)(17). This model was also used in [12].…”

“…(15)(16)(17) when quantifying the gene expression levels or amount of mRNA. Figure 5 displays the log ratios calculated for a simulated array spot.…”

A novel neural network approach to cDNA microarray image segmentation

“…However, it has been proved in [22] that the segmentation method can significantly influence the identification of gene expression values and subsequent analysis. Furthermore, the time-consuming manual processing of the microarrays has led to the recent interest in using a fully automated procedure to accomplish the task [4,23,24]. Although recognition of spots in either control or experimental channels seems to be straightforward, the task is indeed complicated and challenging.…”

“…Also, the active contour techniques were applied in [27] that are sensitive to noise. Other methods such as the applications of wavelets [21,28] and Markov random fields [24,29,30] showed great promise as well. Lawrence et al [31] presented a Bayesian approach in order to process images produced by these arrays that seek posterior distributions over the size and positions of the spots.…”