Double-exposure optical sectioning structured illumination microscopy based on Hilbert transform reconstruction.

Xing Zhou,Baoli Yao,Yanlong Yang,Tong Peng,Junwei Min,Ming Lei,Jia Qian,Guangde Chen,Shaohui Yan,Dan Dan,Tong Ye,Jonathan A Coles

doi:10.1371/journal.pone.0120892

Abstract

Structured illumination microscopy (SIM) with axially optical sectioning capability has found widespread applications in three-dimensional live cell imaging in recent years, since it combines high sensitivity, short image acquisition time, and high spatial resolution. To obtain one sectioned slice, three raw images with a fixed phase-shift, normally 2π/3, are generally required. In this paper, we report a data processing algorithm based on the one-dimensional Hilbert transform, which needs only two raw images with arbitrary phase-shift for each single slice. The proposed algorithm is different from the previous two-dimensional Hilbert spiral transform algorithm in theory. The presented algorithm has the advantages of simpler data processing procedure, faster computation speed and better reconstructed image quality. The validity of the scheme is verified by imaging biological samples in our developed DMD-based LED-illumination SIM system.

Highlights

Optical sectioning microscopy can produce a clear image of the focal plane deeply within a thick sample
The tested specimen is a mixed pollen grain specimen purchased from Carolina Biological Supply Company (Burlington, USA), which exhibits strong auto-fluorescence under the excitation of 450 nm LED light
In the fast and adaptive bi-dimensional empirical mode decomposition (FABEMD)-Hilbert spiral (HS) algorithm, we decompose the IS into nine bi-dimensional intrinsic mode functions (BIMFs) and selectively reconstruct the optimal band-pass filtered pattern to obtain the sectioned image according to Ref. [14]

Summary

Introduction

Optical sectioning microscopy can produce a clear image of the focal plane deeply within a thick sample. Many different techniques are designed to improve the quality of optical sectioning [1,2,3]. As a wide-field optical microscopy, structured illumination microscopy (SIM) has found widespread applications for investigation of cell structures and for time-series imaging of living specimen due to its optical sectioning capability and high imaging speed [4,5,6]. The basic idea of SIM for optical sectioning is to decode the in-focus information and remove the background of out-of-focus portion. The most commonly used decoding algorithm was proposed by Neil et al [7], in which a sinusoidal fringe is projected on the surface of specimen. By taking the root mean square (RMS) of the differences of the adjacent images, an optically sectioned image can be reconstructed. Neil’s algorithm provides a simple way to PLOS ONE | DOI:10.1371/journal.pone.0120892 March 23, 2015

Methods

Results

Conclusion