Digital Image Processing of Live/Dead Staining

Abstract

The quantification of live and dead cells in a substrate is often an essential step in cell biology research. A staining protocol that acts differently on live and on dead cells is applied and the number of cells visible is counted using a microscope. Often this counting is done manually or only evaluated qualitatively. If the number of samples to be analyzed is large, counting live and dead cells will become a labor intensive, and in some cases an unreliable, process. The manual procedure also discards potentially relevant information on the cells beyond their live or dead classification. For example, cell size, shape and distribution cannot be measured manually. Thus, developing a software routine to replace the counting process can result in an increase of both efficiency and quality of the data gathering process. Whether or not the time and/or money spent on creating a dedicated computer algorithm is worthwhile, depends on a large number of factors of which some are specific to the samples and some to the technical expertise available. In a large percentage of cases, creating a computer algorithm may be easier than expected. In order for the reader to correctly asses the difficulty level of his/her specific case, an outline on how to tackle the problem is presented within this chapter. The basic concepts of digital imaging, explained in a possible step by step approach, is offered. It is important to be able to estimate the difficulty level for each specific case. Based on a series of questions the potential of creating a computer algorithm can be offset by the costs to be expected.