Abstract

Image Scanning Microscopy (ISM) is the evolution of Confocal Laser Scanning Microscopy (CLSM). In CLSM, a focused excitation beam scans the sample, and a single-element detector collects the fluorescence light. The presence of a closed pinhole before the detector enables out-of-focus light rejection and a spatial resolution higher than the diffraction limit, but it compromises the Signal-to-Noise Ratio (SNR). In ISM, a detector array collects the fluorescence light, generating a small image for each scan point x_s. At the same time, each element of the detector acts as a small pinhole – located at the detector position x_d – building multiple scanned images in parallel while scanning the sample. Thus, the final dataset i(x_s|x_d) is four-dimensional and can be equivalently seen as a collection of confocal images – identical but shifted and rescaled – or as a set of small wide-field images that we name micro-images. Adaptive pixel reassignment (APR) estimates and inverts the shifts µ(x_d) of the scanned images, which are later summed together [1]. Therefore, the signal from all the detector elements contributes to the reconstruction of a single super-resolved image, maximizing the SNR of the result. However, the lack of a conventional pinhole allows the detection of out-of-focus light, thus compromising the optical sectioning. Here, we present a new approach that fully exploits the information provided by the detector array, enabling background rejection in ISM without sacrificing the SNR.

© 2023 IEEE