A super-resolution reconstruction algorithm for two-photon fluorescence polarization microscopy

Abstract

Two-photon fluorescence polarization microscopy is widely used to monitor the orientation and structural information of biomolecules labeled by fluorescence dipoles, while suffering from the spatial resolution limitation. In this paper, we propose an effective reconstruction algorithm for two-photon fluorescence polarization microscopy to reconstruct the super-resolution image and obtain the finer orientation information of dipoles at corresponding super-resolution scale. First, a conventional statistic model is employed to characterize the orientation distribution of dipole-clusters. By combining the model with the two-photon microscope imaging theory, then an optimization model is proposed and the reconstruction algorithm is developed based on conjugate gradient least square method and applied to various samples. The reconstruction results demonstrate that the imaging resolution of 90 nm can be reached and the finer orientation distribution information of dipole-clusters at the corresponding resolution can be obtained. Finally, Monte Carlo analysis has been performed, which verifies that the algorithm has high accuracy in reconstructing the parameters of the finer orientation distribution of dipole-clusters. The analysis of sample noise proves that the reconstruction algorithm can be utilized to actual samples within the appropriate noise range.