Unlocking the full power of image scanning microscopy with maximum likelihood reconstruction

Image Scanning Microscopy (ISM) enables super-resolution at an excellent signal-to-noise ratio thanks to a detector array. The microscope collects a confocal-like image for each detector element, generating a large dataset that requires tailored processing tools to be converted into a single super-resolved image. We propose a novel algorithm to fuse the dataset into an image with enhanced optical sectioning and resolution. Our method exploits the information inherently contained in the dataset to reject out-of-focus contributions and reconstruct an image with a smaller pixel size and a better resolution. The proposed method requires minimal user inputs and outperforms existing reconstruction methods.