Research On Partial-frequency-spectrum Reconstruction Algorithm Of Structured Illumination Microscopy

Structured Illumination Microscopy(SIM)is a wide-field,high-efficiency superresolution imaging technology with low phototoxicity,fast imaging speed,and no special fluorescent labeling.It has incomparable advantages in living cell imaging.While many life activities are fleeting,the current imaging speed will miss many important intracellular mechanisms of action.Increasing the temporal resolution plays an important role in the observation of the mechanism of interaction between the organelles and the internal mechanism.The improvement of the reconstruction rate is also beneficial to realize real-time cell imaging.At present,mainstream SIM systems have been fully optimized in terms of hardware performance and system synchronization efficiency,and the space for improvement is limited,and the reconstruction algorithm still requires a large amount of calculation,which affects the real-time performance of imaging.In order to effectively improve the time resolution and reconstruction speed,this thesis studies the structure illumination microscopy reconstruction algorithm,reduces the reconstruction speed and time resolution by reducing the number of original images needed for reconstruction.The main work is as follows:1.A super-resolution image reconstruction algorithm based on partial-frequencyspectrum implementation is proposed.The algorithm utilizes the similarity of the spectrum in the SIM image,estimates the low-frequency information through the widefield fluorescence image,and then subtracts and separates the high-frequency information.Two SIM images with phase shift of ? are acquired in each direction,and a total of 6 raw SIM images in three directions can reconstruct the super-resolution image.2.Numerical simulation of partial-frequency-spectrum reconstruction algorithm by MATLAB.The method of solving parameters such as phase,fringe frequency and deconvolution is optimized to speed up the reconstruction and improve the image quality.3.Compare the images reconstructed by partial-frequency-spectrum algorithm with the effects of other SIM reconstruction algorithms.The simulation analysis results show that the partial-frequency-spectrum algorithm can obtain the image resolution similar to the standard SIM reconstruction algorithm,and the reconstruction rate is increased by about 5 times.4.The super-resolution effect of partial-frequency-spectrum algorithm was verified by collecting bovine lung endothelial cells and fluorescent microspheres.Experiments show that the partial-frequency-spectrum algorithm proposed in this thesis can effectively improve the spatial resolution in the experimental data,and can obtain the resolution equivalent to the standard SIM under better signal-to-noise ratio.