Quantitative analysis of fluorescent image – from descriptive to computational microscopy

Application of fluorescence in microscopy led to a considerable breakthrough in biochemical sciences by facilitating a new way of observing complex biological processes, structures and interactions and locating them in particular parts of the cell. Today, fluorescent imaging is widely applied in published literature to confirm research hypotheses. However, about only one of ten publications provides supporting information for the observations based on fluorescence through a computational analysis of the acquired images. “One picture is worth ten thousand words” was once stated by Frederic R. Bernard, however, images alone may not be enough to provide reliable information. For example, given population diversity, it is very likely that a selected cell ( ‘the cell’ ) will be found that confirms a particular theory, while most others will refute it. In addition the validity of observations conducted on cell cultures may be disputable, as these are usually cancer or immortalized cells and therefore differ significantly from the original tissue. It is known, for example that cultured cancer cells easily evolve into a variety of clones that may not differ morphologically, but when stimulated to undergo differentiation or cell death they tend to react differently to normal cells at the biomolecular level. While cellular analysis requires a great deal of caution, descriptive tissue analysis is even more subjective and interpretations are highly dependent on personal experience of the observer. Baak, in his article [Baak 2002] addressed difficulties in qualitative assessment of microscopic samples in pathology. The reliability was acceptable when an observation was assessed by at least three specialists, but consistency was low especially in more complex analysis. Comparison of the data between different laboratories is also difficult considering the simplicity of observations made and the complexity of the pathological processes.
Modern imaging methods which use charged-coupled devices (CCDs) or photomultipliers (PMTs) provide a possibility to capture the visual observations in a numerical, digitalized form. In this format reliable quantitative data can be extracted which open new ways of understanding and supporting an observation process. In this chapter, various methods of quantitative analysis will be presented and discussed, applicable to studies of a single cell as well as whole populations and tissues.