Recently, I attended the Quantitative BioImaging (QBI) Conference 2020, served right here in Oxford. Amongst the many methods on the menu were new recipes for spicing up your Cryo-EM images with a bit of CiNNamon with a peppering of Poisson point processes in the inhomogeneous spatial case amongst many others. However, like many of today’s top tier restaurants most of the courses on offer were on the smaller side, nano-scale in fact, serving up the new field of Super Resolution Microscopy!

Like buying your phyllo pastry from the shops instead of making it yourself, Super Resolution Microscopy is a bit of a cheat. The diffraction limit of light is a universal bound on the resolution of an image in any optical system (usually 200-600nm depending on wavelength). However using photoswitching, the property of certain fluorescent bound compounds to turn on-and-off when activated with light, imaging systems can squeeze more information out of a scene by capturing multiple images. Images are combined to produce a distribution over bound locations and, like a masterful meringue, the “peaks” formed in these distributions provide a higher degree of location certainty and thus a higher resolution image (as high as 20nm!).

Techniques such as STORM and PALM were some of the earliest super resolution techniques to be cooked up, however new Bayesian and Machine Learning techniques, such as DeepSTORM and DeepSTORM3D, seek to get all of the zest out of the these multilayer image data sets. To see what’s a boilin’ in the Super Resolution world, sit back, grab a cuppa and join Nobel Prize winner Eric Betzig, as he explores real time imaging techniques in a field brimming with possibilities.