Resolution, accuracy and precision in super-resolved microscopy images using SUPPOSe

Abstract

In this work we carry out a performance analysis of the SUPPOSe (Superposition of virtual point sources) algorithm, a deconvolution method with super-resolution for single images based on the superposition of point sources of equal intensity. By testing the method with synthetic and experimental images, we determine the extent of the accuracy, precision and resolution of the method and how these parameters are modified under different signal/noise conditions. Performing simulations with a numerical aperture NA∼1.3 and an emission wavelength λ=520nm, a SUPPOSe resolution of 75-80nm with an accuracy of 20nm and a precision of 10nm is shown to be possible. We also tested the method with experimental images, biological structures and fluorescent substrates from calibration samples, obtaining results consistent with the improvements obtained in the simulations.