Super-resolution microscopy with phase masks is a promising technique for 3D imaging and tracking. Due to the complexity and flexibility of the 3D point spread functions, well-accepted recovery algorithms independent of point spread functions are still missing. We introduce a general 3D super-resolution recovery algorithm that works for arbitrary phase masks generating 3D point spread functions. This algorithm uses an advanced fast deconvolution algorithm to generate initial guesses, and further refines the positions by least square fitting and suppresses overfittings using a machine learning determined threshold. This algorithm will also serve as a standard criterion for future phase mask design.  Due to heavy computation in 3D, we further accelerate the algorithm through parallel computation on graphics processing units and demonstrate that through desktop parallelization, the ultimate limit of real-time processing is possible.

Super-resolution algorithm for 3D single-molecule microscopy imaging

251st National Meeting (2016)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.