A.2-Light Microscopy


Light microscopy is the predominant imaging technique in life science due to its flexibility, availability and capacity to image specimens in vivo. Moreover, the development of optical labels, such as green fluorescence protein (GFP), allows targeting specific biological structures and processes for direct observation under the microscope. The list of light microscopy techniques is too broad to make a full description in this brief introduction [1].  Some of the most common ones are confocal light microscopy, multi-photon techniques or super-resolution approaches such as structured illumination and PALM.


One of my main areas of research in bioimaging informatics deals with light sheet microscopy based techniques, and in particular with Digital Scanned Laser Light Sheet Microscopy (DSLM). DSLM is an essential tool for quantitative imaging of cellular dynamics in the embryonic development of complex organisms such as Drosophila and zebrafish. Briefly, it allows three-dimensional (3D) recordings of full embryos in vivo at high-speed time intervals for long periods of time. It achieves that by selectively illuminating one plane of the specimen at a time, thus composing a 3D image by stacking planes along the optical axis. Due to the selective illumination, there is minimal photo-bleaching affecting the sample, so long recordings are possible. Moreover, the illumination and the photon detection is done with different optical arms (Fig. 1), so multi-view image acquisition is possible in order to obtain full coverage of the specimen. The main drawback is the light sheet thickness required to image full embryos, which degrades the resolution along the optical axis with respect to confocal microscopy. However, individual cells and their divisions can still be tracked along the series of images.

Fig.1: DLSM schematic from [2].

A typical experiment for Drosophila development involves 3D stacks every 30 seconds for over 24 hours of development. Multiple views and millions of voxels are needed to cover the large field of view, which implies terabytes (TB) of data recorded in a single experiment. Moreover, each image contains thousands of cells that need to be segmented and tracked in order to extract quantitative cell lineaging information. It is easy to imagine that performing any sort of quantitative analysis in these datasets requires a long pipeline of computational tools. Even the most patient and driven scientists will find impossible to interact with the data without computational aids. 


------------------------------------- References ---------------------------------------------

[1] D. B. Murphy, Fundamentals of Light Microscopy and Electronic Imaging, 1st ed. Wiley-Liss, 2001.

[2] P. J. Keller, A. D. Schmidt, J. Wittbrodt, and E. H. K. Stelzer, “Reconstruction of Zebrafish Early Embryonic Development by Scanned Light Sheet Microscopy,” Science, vol. 322, no. 5904, pp. 1065-1069, Nov. 2008.