Data analysis: Extracting high-resolution shape information from microscopy images

Data analysis: Extracting high-resolution shape information from microscopy images

Light microscopy probably is one of the most important techniques in all of the life sciences. The invention of the microscope has made possible discoveries on which much of our current knowledge and capabilities in biology and medicine are based. Imaging of live samples is limited to light microscopy since the higher energies involved in, e.g., electron microscopy would kill the cells. The resolution of light microscopes is, however, limited by the wavelength of the light to a few hundred nanometers. As modern biology and systems biology are striving to understand the working mechanisms of live cells, we wish to observe individual structures inside single cells and quantify their dynamics and interplay. The size of these structures is often comparable to the wavelength of the light, such that they appear heavily blurred and noisy in the images, preventing analysis of their shape and dynamics.

We have recently developed and implemented an image-processing algorithm that can undo much of the blurring introduced by the microscope without directly solving the ill-posed deconvolution problem. The algorithm uses prior knowledge (models) of the microscope and the observed objects, which allows formulating the image-processing task as an optimization problem. This enables accurate and reliable reconstruction of small intracellular shapes, effectively increasing the localization precision of microscopes more than 100-fold. It thus becomes possible to following, e.g., the dynamics of endocytic organelles during internalization, sorting, and trafficking of viruses that infect the cell, enabling observations and testing of biological hypotheses that were impossible before.

The image shows a micrograph of a single cell with intracellular endosomes fluorescently labeled (white). The shape outlines of the endosomes are reconstructed by the present algorithm with nanometer precision (blue lines). The red crosses show the positions of viruses inside the cell as they are sorted and transported through the endosomal network.


J. A. Helmuth, C. J. Burckhardt, U. F. Greber, and I. F. Sbalzarini. Shape reconstruction of subcellular structures from live cell fluorescence microscopy images. J. Struct. Biol. 167, 1-10, 2009.

J. A. Helmuth and I. F. Sbalzarini. Deconvolving active contours for fluorescence microscopy images. In Proc. Intl. Symp. Visual Computing (ISVC), volume 5875 of Lecture Notes in Computer Science, pages 544–553, Las Vegas, USA, November 2009. Springer