Reconstructing the image scanning microscopy dataset: an inverse problem

Abstract Confocal laser-scanning microscopy (CLSM) is one of the most popular optical architectures for fluorescence imaging. In CLSM, a focused laser beam excites emission from specific specimen position. Some actuators scan probed region across sample and photodetector collects single intensity value each point, building two-dimensional image pixel-by-pixel. Recently, new fast single-photon array detectors have allowed recording full bi-dimensional transforming CLSM into scanning (ISM). This latter offers significant improvements over traditional imaging but requires an optimal processing tool to extract super-resolved four-dimensional dataset. Here we describe formation process in ISM statistical point view, use Bayesian framework formulate multi-image deconvolution problem. Notably, detector suffers exclusively photon shot noise, enabling development effective likelihood model. We derive iterative maximization algorithm test it on experimental simulated data. Furthermore, demonstrate that dataset redundant, possibility obtaining reconstruction sampled at twice step. Our results prove ISM, under appropriate conditions, Nyquist–Shannon sampling criterium effectively relaxed. finding can be exploited speed up acquisition by factor four, further improving versatility systems.