Computational processing and analysis of dynamic fluorescence image data

Abstract

With the many modes of live cell fluorescence imaging made possible by the rapid advances of fluorescent protein technology, researchers begin to face a new challenge: How to transform the vast amounts of unstructured image data into quantitative information for the discovery of new cell behaviors and the rigorous testing of mechanistic hypotheses? Although manual and semiautomatic computer-assisted image analysis are still used extensively, the demand for more reproducible and complete image measurements of complex cellular dynamics increases the need for fully automatic computational image processing approaches for both mechanistic studies and screening applications in cell biology. This chapter provides an overview of the issues that arise with the use of computational algorithms in live cell imaging studies, with particular emphasis on the close coordination of sample preparation, image acquisition, and computational image analysis. It also aims to introduce the terminology and central concepts of computer vision to facilitate the communication between cell biologists and computer scientists in collaborative imaging projects.