Automated acquisition of many images of cells, followed by computer analysis, allows quantitative assessment of cellular phenotypes.

Microscopic imaging has historically been a powerful but laborious and descriptive method. When Antonie van Leeuwenhoek first observed microorganisms in the 1600s and Santiago Ramón y Cajal discerned the fine structure of the nervous system more than 200 years later, these scientists spent long hours observing and drawing what they saw. Modern day scientists—with their sophisticated microscopes and CCD cameras—can instantly record more detailed observations, but for most biologists, the main objective is still to obtain descriptive, representative images of the phenomenon under observation.

Today's computers and the sophisticated algorithms they can run are beginning to change the game, though. These developments make it possible for a computer to take control of most of the imaging process and have the power to transform microscopy from a mostly descriptive method to a more quantitative technique that is capable of making observations that human observers could never match.

This gradual revolution is already occurring in several applications. Large-scale automated imaging screens of cells in culture or small model organisms are capable of detecting very rare or subtle phenotypes that would be missed by human observers. Similarly, image analysis algorithms can recognize, categorize and track cellular phenotypes in hundreds or even thousands of cells at once—something that is beyond the abilities of even the most highly trained microscopist. Even applications that are possible without automation, such as measurements of gene expression in large numbers of cells, whether in vivo or in vitro, can be much more efficient and quantitative when they are automated.

Automated microscopy methods do require substantial technical know-how to correctly design and use the necessary algorithms, and care must be taken when evaluating the results. These techniques are gradually becoming more widespread, though, and efforts are underway to make them more accessible to a wider range of users. Caution is warranted to ensure that novices—and even experts—use these methods properly, but the exciting possibilities these methods provide make them well worth watching.