Single molecule localization microscopy with autonomous feedback loops for ultrahigh precision

Simao Coelho, Jongho Baek, Matthew S Graus, James M Halstead, Philip R Nicovich, Kristen Feher, Hetvi Gandhi, Katharina Gaus

Preprint posted on 5 December 2018

Self-correcting microscope: real-time drift correction and optical path re-alignment via feedback enable super-super-resolution imaging

Selected by Lars Hubatsch

Categories: biophysics, immunology


Super resolution microscopy has revolutionized imaging of biological samples by improving (lateral) resolution by an order of magnitude, from a few hundred nanometers to tens of nanometers, under optimized conditions, winning its inventors the 2014 Nobel Prize in Chemistry. Even higher resolutions are theoretically possible, however, current approaches suffer from hardware-induced problems.

In particular, drift-induced artefacts are problematic, because to obtain super-resolution, many images have to be captured of one and the same field of view. In every image, only a small sub-population of fluorescently labelled molecules are lit up, enabling precise localization of each molecule, using the knowledge that one and only one molecule is responsible for a given fluorescent spot. By combining the knowledge of all the molecules’ localizations, one obtains the final image, often by combining tens of thousands of sparsely lit frames. Evidently, it is crucial that the absolute position of molecules stays the same between frames. Any drift will have a severe impact on resolution, because one molecule could be mapped to different positions.

Key Technological Improvements

This preprint aims to overcome the current hardware limitations of super-resolution microscopy by introducing three different hardware-based correction mechanisms. In a first step, the authors use a feedback loop between the stage and fiducial markers close to the sample to realign the stage such that the same area of the sample is captured in the field of view at all times. Second, the light path, which also forms a major error source, is kept well-aligned by an independent feedback loop. Here, a piezo-electric mirror is used to repeatedly realign the optical path. Third, by imaging a nano-fabricated array of holes with precisely known positions, systematic errors of the camera and chromatic and optical aberrations can be reduced. Taken together, these three correction mechanisms increase precision and enable longer acquisitions.


Among several use cases highlighted by the preprint, particularly impressive is the ability to perform relatively precise measurements of the spatial separation of molecules on the scale of a few nanometers. According to the authors, similar measurements are not possible using other current technologies such as FRET. As proof of principle the nearest neighbor distance between pCD3ζ and CD45, two molecules involved in initiation of T cell receptor signaling, is measured, answering a so far unresolved question.

One thing missing from a non-expert’s perspective is a concise account of how difficult this setup is to implement in practice. Also, from a practical point of view, making the code available in a public repository and not ’on request’ would be a nice addition. In any case, feedback single-molecule localization microscopy seems like a potential game-changer, possibly being able to replace FRET and allowing imaging at unprecedented spatial resolution.


Posted on: 14 January 2019


Read preprint (No Ratings Yet)

Have your say

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Sign up to customise the site to your preferences and to receive alerts

Register here