Model-free pixelation correction in SOFI imaging

Abstract

Super-resolution optical fluctuation imaging (SOFI) delivers an enhanced spatial resolution in fluorescence imaging by analyzing spontaneous fluctuations in the fluorophore emission. SOFI images are usually obtained by calculating cross-cumulants between adjacent pixels on the detector, which provides the increased pixel densities required to sample the improved spatial resolution, but can result in pixelation artifacts. In this contribution, we describe a simple, model-free, and computationally efficient algorithm to correct such artifacts by matching the means and variances of the different cross-cumulants. We show that this strategy not only results in pixels that are essentially free of artifacts, but can also correct for detector imperfections such as the spurious correlations present in electron-multiplied CCD cameras.