Robust automatic feature extraction for protein microarrays

In this paper, we present a robust methodology for image registration, segmentation, and feature extraction for protein microarrays. Originally designed for application to an Agilent microarray platform, the algorithms used are easily adapted to other platforms. Linear and nonlinear filtering techniques are used to identify protein signals on the array. After signal identification, expression values for each protein are then derived. Emphasis is placed on robustness of feature identification and low computational complexity.