A fully-automated nuclear segmentation approach for analysis of cell fluorescent images

Abstract

Fluorescent microscopy imaging is an efficient technique allowing estimating about protein expression in cells. A critical issue is image analysis. The complexity of investigated cell systems hampers application of standard image processing methods. We have developed an image analysis approach for the nuclear segmentation, that is (i) fully-automated, (ii) high-throughput, and (iii) has the improved accuracy of segmentation. The developed approach exploits the morphological information about nuclei in the image and pixel intensity values.