Getting the most out of super-resolution using high throughput imaging and incorporating prior knowledge

Single molecule-switching based super-resolution (SMLM) is now almost 20 years old but, despite the availability of increasingly mature commercial instrumentation, it remains a complex and time consuming approach. There is a constant trade-off between signal-to-noise, resolution, and imaging time. I will discuss our recent work in combining rapid single-molecule imaging with smart automation to increase the volume of data that can be captured whilst simultaneously reducing the amount of time a person has to be in front of the microscope. I will also detail our efforts – by using biophysically informed constraints on object structure – to obtain useful quantitative results even when the resulting data might not be as dense (or high-quality) as desired. For example, by constraining the curvature energy of the resulting membrane, we can reconstruct intracellular organelle membranes from SMLM data with much better fidelity than alternative approaches.