Three-dimensional Super-resolution Imaging Based On Fluorescence Blinking

As a microptic technology detecting and collecting the fluorescence generated by using the molecular to absorb the light with specific wavelength,fluorescence microscope has a history of more than 100 years.Recently,due to the popularization of fluorescence in situ hybridization(Fish)and green fluorescent protein(Gfp)in genetics and protrinology,with auxiliary drive of digital CCD imaging technique,more vigorous vitality is given to this traditional technology.However,as an optical imaging tool,the spatial resolution of traditional fluorescence microscope is always affected by diffraction effect,Limited in 200 nm at lateral direction and 500 nm in axial resolution,which fails to observe finer structures in cells.Although there are multiple methods breaking through the limit of optical diffraction with the progress of science,time resolution and spatial resolution are restricted mutually.In addition,the problem of fluorescence microscopy is not only the resolution limited of diffraction,but also the ability of multicolor fluorescence imaging.Recently,in order to avoid the crosstalk of muticolor spectrum,it is usually to separate the spectrum with filters,prisms into mutichannel,but it is not only inevitably reduced the intensity of fluorescence but also with little work in separate the spectrum when the polychromatic spectrum increase or abut.To breaking the mutual restriction of temporal resolution and spatial resolution,and the difficulty of separation of muticololr fluorescence spectrum,Based on the theory of superresolution optical fluctuation imaging(SOFI),in this paper emphatically explore the technology of muticolor fluorescence separation and high-speed three-dimensional imaging mode.The main work of this paper in shown:1: A technical concept of muticolor fluorescence separationGenerally,the fluorescence spectrum has a bandwidth of tens of nanometers,so it will be limited by the crosstalk between spectral components when muticolor imaging,in order to break this limitation,we can use quantum dots with narrower spectral width and filters with smaller bandpass width,unfortunately there is a physical limit,so the number of fluorescence will not be infinite,based on SOFI algorithm a wavelength fluctuation correlation algorithm has proposed,the key of this method is let the variation of wavelength similar to the variation of fluorescence intensity in SOFI algorithm,Through this method,spectral curve function in multicolor fluorescence imaging can be analogized to the reduction of full width at half maximum of optical point spread function by analogizing the variation of wavelength to the variation of fluorescence intensity in SOFI algorithm.The results of simulation indicate that its reduction ability is consistent with SOFI algorithm,and under the condition of 500 sequence images of two-color fluorescence spectra with the width of 60 nm and center-to-center space of20 nm in second-order spectral algorithm,spectral imaging resolution can be increased by √2times.2: Based on the theory of distorted grating and the second-order spectral algorithm,Stimulated the large depth of field imaging at 4 μ m and the enhancement of spectral resolution at ± 1 diffractionIn order to realize the quick three-dimensional super-resolution imaging,optical imaging system is required to have a certain imaging depth and ability of quick imaging.This paper proposes the method combining SOFI with distorted grating and introduces spectral algorithm of ±1 diffraction order with spectral ability.And the imaging performance of 4 μ m depth of field and the peak separation of the two-color fluorescence spectrum width 60 nm and the center spacing 20 nm at ± 1 diffraction level are verified by simulation.3: Three-dimensional imaging experiment based on SOFI is carried outBased on wide-field fluorescence,dual objective multi-plane imaging system is designed and established,This system reaches the collection speed of acquiring 1200 slice images within2 seconds under the ideal situation that the frame rate of the camera is 100 HZ.Besides,a set of three-dimensional reconstruction algorithm based on Matlab image processing platform is compiled,and super-resolution images reconstructed by six groups of sequence images are used to reconstruct the microtubule structure of cervix carcinoma(Hela cell).