An Automatic and Power Spectra-based Rotate Correcting Algorithm for Microarray Image

Microarray image analysis, an important aspect of microarray technology, faces vast amount of data processing. At present, the speed of microarray image analysis is quite limited by excessive manual intervention. The geometric structure of microarray determines that, while being analyzed, microarray image should be collimated in the scanning vertical orientation. If rotation or tilt happens in microarray image, the analysis result may be incorrect. Although some automatic image analysis algorithms are used for microarray, still few methods are reported to calibrate the microarray image rotation problem. In this paper, an automatic rotate correcting algorithm is presented which aims at the deflective problem of microarray image. This method is based on image power spectra. Examined by hundreds of samples of clinical data, the algorithm is proved to achieve high precision. As a result, adopting this algorithm, the overall procedure automation in microarray image analysis can be realized