Super-resolution SIM-FRET image reconstruction method based on spatial co-localization mask filtering

Abstract

The present invention discloses a super-resolution SIM-FRET image reconstruction method based on spatial co-localization mask filtering. The present invention adopts unified Wiener reconstruction parameters to perform FRET three-channel SR-SIM linear Wiener reconstruction. The reconstruction method can maintain the linear relationship of the fluorescence intensity of the original image to the greatest extent, thereby meeting the fidelity requirement of the fluorescence intensity calculated by FRET; further, the present invention realizes a spatial co-localization mask generation method, which can filter and remove the FRET false positive signal caused by the random noise artifact of the SR-SIM linear Wiener reconstruction in a targeted manner, while retaining the FRET signal with super-resolution fine structure, thereby realizing quantitative super-resolution detection of FERT.

Description

Technical Field

[0001] The invention belongs to the technical field of fluorescence resonance energy transfer detection, and in particular relates to a super-resolution SIM-FRET image reconstruction method based on spatial co-localization mask filtering. Background Art

[0002] Fluorescence resonance energy transfer (FRET) microscopy is currently the only technology that can detect dynamic interactions and spatial distribution of biological macromolecules in real time for a long time in a single living cell. In theory, it can quantitatively analyze protein molecular interactions at a scale of 10 nm in real time in living cells through microscopy. However, FRET microscopy is limited by the resolution constraints of the diffraction limit of the fluorescence microscope, so when observing protein molecular interactions and spatial localization in subcellular regions, only speculation can be made instead of direct observation results. The development of structured light super-resolution (SR-SIM) microscopy makes it possible to observe subcellular spatial structures at a scale of 100 nm in more detail. Its low illumination intensity, fast imaging speed, and multi-color imaging capabilities make it possible to integrate SR-SIM and FRET technologies to achieve dynamic super-resolution quantitative SIM-FRET imaging of living cells.

[0003] Super-resolution imaging achieved by SR-SIM is highly dependent on reconstruction algorithms and is prone to artifacts, while quantitative analysis of FRET requires high fidelity in spatial structure and fluorescence intensity of the input image. Therefore, the generation of SIM artifacts and quantitative FRET microscopy imaging have become a contradiction that hinders the integration of the two. SR-SIM artifacts mainly come from two parts: 1. Spatial structural artifacts of the spatial distribution of the measured sample caused by the error in parameter estimation; 2. Random noise artifacts generated by the noise generated by the detection device are amplified during the reconstruction process. The former can be largely solved by a more robust parameter estimation algorithm, and the fidelity of the spatial structure and the fidelity of the fluorescence grayscale can be retained after the parameter error is corrected, so as not to affect the quantitative analysis of FRET. Although the latter can be removed by increasing the number of photons collected by the detector to improve the signal-to-noise ratio or by introducing a regularization parameter constraint for denoising during the reconstruction process, it is questionable whether the fluorescence intensity of the super-resolution image after iterative reconstruction with the SR-SIM regularization parameter constraint still has a linear quantitative relationship with the original image. Therefore, it is not suitable to use the iteratively reconstructed super-resolution image as the input of FRET calculation. In principle, the linear reconstruction algorithm of SR-SIM can maintain the linear relationship of the fluorescence intensity of the original image, but its disadvantage is the inevitable random noise artifacts. This random noise artifact will produce false positive FRET signals in the FRET calculation and significantly affect the quantitative calculation of FRET efficiency E and Rc. Summary of the invention

[0004] The main purpose of the present invention is to overcome the shortcomings and deficiencies of the prior art and provide a super-resolution SIM-FRET image reconstruction method based on spatial co-localization mask filtering.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] One aspect of the present invention provides a super-resolution SIM-FRET image reconstruction method based on spatial co-localization mask filtering, comprising the following steps:

[0007] Acquire a FRET three-channel raw image data group after structured light modulation; the FRET three-channel includes a donor excitation donor emission channel DD, an acceptor excitation acceptor emission channel DA, and a donor excitation acceptor emission channel AA;

[0008] Based on the unified Wiener reconstruction parameters, the FRET three-channel structured light super-resolution linear Wiener reconstruction is performed to obtain the FRET three-channel super-resolution image data;

[0009] Perform data alignment on the FRET three-channel super-resolution images;

[0010] Perform background subtraction on the FRET three-channel super-resolution image;

[0011] Generate a spatial co-localization mask according to the strong correlation characteristics of the FRET area images of the DD and AA channels, and remove FRET false positive signals according to the spatial co-localization mask;

[0012] The super-resolution images after removing the FRET false positive signals were used to measure the FRET efficiency and the donor-acceptor concentration ratio.

[0013] As a preferred technical solution, the acquisition of the FRET three-channel original image data set after structured light modulation is specifically as follows:

[0014] The SR-SIM system acquires three-channel raw image data sets based on channel-sensitized intensity measurement, including the donor channel image set Receptor Channel Image Set And FRET channel image group Wherein r = (x, y) is the spatial coordinate of the fluorescence image pixel, each group of original image data contains several different structured light azimuth angles, and each structured light azimuth angle contains several cosine structured light illumination wide-field fluorescence images with different phase differences.

[0015] As a preferred technical solution, the FRET three-channel structured light super-resolution linear Wiener reconstruction based on unified Wiener reconstruction parameters is specifically as follows:

[0016] The FRET three-channel original image group is subjected to structured light super-resolution linear Wiener reconstruction using unified Wiener reconstruction parameters to obtain the FRET three-channel super-resolution image data.

[0017]

[0018] in, is the Fourier transform of the corresponding image of FRET channel X; k is the coordinate of the fluorescence image pixel converted to the frequency domain, is the initial phase of the FRET channel X corresponding to the structured light direction angle θ, is the modulation depth of the FRET channel X corresponding to the structured light direction angle θ, is the spatial frequency of the FRET channel X corresponding to the structured light direction angle θ, is the optical transfer function of the SR-SIM system, The optical transfer function of the SR-SIM system corresponding to the FRET channel X is expressed in the frequency domain according to the spatial frequency vector k θ and -k θ Spectrum shifting, Respectively The conjugation of the channel X includes the donor excitation donor emission channel DD, the acceptor excitation acceptor emission channel DA and the donor excitation acceptor emission channel AA; w is the reconstruction Wiener parameter, is a Gaussian apodization function, and iFFT is the inverse Fourier transform of the image.

[0019] As a preferred technical solution, the data alignment process for the FRET three-channel super-resolution image is specifically as follows:

[0020] by Super-resolution fluorescence images obtained based on channels Perform affine transformation to achieve pixel-by-pixel alignment of three-channel images.

[0021] As a preferred technical solution, the background subtraction process of the FRET three-channel super-resolution image is specifically as follows:

[0022] Reconstruct super-resolution images for each FRET imaging channel Perform pixel-by-pixel grayscale value statistics, and then take the grayscale value corresponding to the first peak in the grayscale value histogram as the background grayscale value of the field of view After background correction of the image and setting the negative pixel grayscale values ​​to zero, the corrected image of each imaging channel is obtained, specifically:

[0023]

[0024] As a preferred technical solution, the spatial co-localization mask is generated according to the strong correlation characteristics of the FRET area images of the DD and AA channels, specifically:

[0025] according to and Generate Pearson coefficient correlation coefficient weight matrix r map , as follows:

[0026]

[0027] in, is the mean value of the super-resolution fluorescence image after background subtraction in the donor channel, It represents the mean value of super-resolution fluorescence image after background subtraction in receptor channel;

[0028] according to and Generate the Manders colocalization coefficient weight matrix R map , as follows:

[0029]

[0030] r map With R map Multiply them together to get the mixed co-localization weight matrix, and segment the mixed co-localization weight matrix through the adaptive threshold processing algorithm to generate a binary mask plate B colocalmask , as follows:

[0031]

[0032] Where th is the threshold of the adaptive threshold processing algorithm.

[0033] As a preferred technical solution, the measurement of FRET efficiency and donor-acceptor concentration ratio is specifically as follows:

[0034]

[0035]

[0036]

[0037] Among them, E SIM is the FRET efficiency under structured light super-resolution, Rc SIM is the FRET donor-acceptor concentration ratio under structured light super-resolution, Fc SIM is the fluorescence intensity of the acceptor sensitized emission, G is the sensitization quenching conversion factor, K is the donor-acceptor concentration conversion factor, and a, b, c, and d are the system crosstalk coefficients.

[0038] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0039] (1) The present invention adopts a unified Wiener reconstruction parameter to perform FRET three-channel SR-SIM linear Wiener reconstruction, which can maintain the linear relationship of the fluorescence intensity of the original image to the greatest extent and thus meet the fidelity requirements of the fluorescence intensity calculated by quantitative FRET;

[0040] (2) The present invention generates a spatial co-localization mask and uses the mask to specifically remove the FRET false positive signals caused by the random noise artifacts of the SR-SIM linear Wiener reconstruction, while effectively retaining most of the FRET signals with super-resolution fine structures, thereby achieving quantitative super-resolution detection of FERT. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 is a flow chart of a FRET image reconstruction method based on spatial co-localization mask filtering and SR-SIM according to an embodiment of the present invention;

[0042] Figure 2 is a schematic diagram of generating a spatial co-location mask according to an embodiment of the present invention;

[0043] Figure 3 The schematic diagram of the simulation reconstruction result of the present invention; wherein, Figure 3 (a) is the true value result of the simulation model, Figure 3 (b) is the result of wide-field reconstruction of the simulated FRET three-channel original image data set. Figure 3 (c) is the result of simulating the FRET three-channel original image data set after structured light super-resolution linear Wiener reconstruction but without mask filtering. Figure 3 (d) is the result of simulating the FRET three-channel original image data set after structured light super-resolution linear Wiener reconstruction and mask filtering according to the existing method. Figure 3 (e) The simulated FRET three-channel original image data set is reconstructed by structured light super-resolution linear Wiener reconstruction and based on and Results of mask filtering using spatial colocalization;

[0044] Figure 4 The schematic diagram of the effect of the method of the present invention at different signal-to-noise ratios; wherein, Figure 4 (a) is a schematic diagram of the effect of the co-localization mask filtering method of the present invention at different signal-to-noise ratios, Figure 4 (b) Schematic diagram of the effect of conventional FRET mask filtering method at different signal-to-noise ratios. DETAILED DESCRIPTION

[0045] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without making creative work are within the scope of protection of the present application.

[0046] Example

[0047] like Figure 1 As shown, this embodiment provides a structured light illumination super-resolution FRET image reconstruction method using spatial co-localization mask filtering, comprising the following steps:

[0048] S1. Acquire a FRET three-channel original image data set after structured light modulation; the FRET three-channel includes a donor excitation-donor emission channel DD, an acceptor excitation-acceptor emission channel DA, and a donor excitation-acceptor emission channel AA. For different FRET donors and acceptors, the corresponding excitation-emission channels of FRET are also different. In this embodiment, a fluorescent protein pair GFP-mCherry is selected as a FRET donor-acceptor pair, wherein the donor excitation channel uses a 488nm laser as the excitation light of the donor, the donor emission channel uses a 525±15nm channel as the donor emission fluorescence detection channel, the acceptor excitation channel uses a 561nm laser as the excitation light of the acceptor, and the acceptor emission channel uses a 600±15nm channel as the acceptor emission fluorescence detection channel;

[0049] Acquisition of three-channel raw image data sets based on channel-sensitized intensity measurement (E-FRET method) by SR-SIM system, including donor channel image set (Fluorescence image group detected by donor excitation and donor emission channel DD), acceptor channel image group (Fluorescence image group detected by receptor excitation receptor emission channel AA) and FRET channel image group (Fluorescence image group detected by the donor-excited acceptor emission channel DA); wherein r = (x, y) is the spatial coordinate of the fluorescence image pixel, each group of original image data contains several different structured light azimuth angles, and each structured light azimuth angle contains several wide-field fluorescence images modulated by structured light with different phase differences.

[0050] In particular, in this embodiment, each group of original image data sets contains three different structured light direction angles θ1, θ2 and θ3, each structured light azimuth contains three structured light modulated wide-field fluorescence images with different phase differences n1, n2 and n3, and a total of 27 structured light modulated wide-field fluorescence images are collected in DD, DA and AA channels, which is a preferred technical solution.

[0051] S2. Perform FRET three-channel structured light super-resolution linear Wiener reconstruction based on unified Wiener reconstruction parameters to obtain FRET three-channel super-resolution image data

[0052]

[0053] in, is the Fourier transform of the corresponding image of FRET channel X; k is the coordinate of the fluorescence image pixel converted to the frequency domain, is the initial phase of the FRET channel X corresponding to the structured light direction angle θ, is the modulation depth of the FRET channel X corresponding to the structured light direction angle θ, is the spatial frequency of the FRET channel X corresponding to the structured light direction angle θ, is the optical transfer function (OTF) of the SR-SIM system, The OTF of the SR-SIM system corresponding to the FRET channel X is expressed in the frequency domain according to the spatial frequency vector k θ and -k θ Spectrum shifting, Respectively The conjugation of the channel X includes the donor excitation donor emission channel DD, the acceptor excitation acceptor emission channel DA and the donor excitation acceptor emission channel AA; w is the reconstruction Wiener parameter, is a Gaussian apodization function, and iFFT is the inverse Fourier transform of the image.

[0054] In particular, in the above formula for FRET three-channel structured light super-resolution linear Wiener reconstruction, each group of original image data contains three different structured light direction angles θ1, θ2 and θ3, and each structured light azimuth contains three different phase differences n1, n2 and n3. This selection is a preferred method of this embodiment.

[0055] Furthermore, during the super-resolution linear Wiener reconstruction process, changes in the Wiener parameters will affect the grayscale value of the reconstructed super-resolution image. By maintaining the consistency of the Wiener reconstruction parameters during FRET three-channel super-resolution reconstruction, the distortion of the relative fluorescence intensity between the reconstructed image and the original image can be minimized, thereby meeting the fidelity requirements of the fluorescence intensity during FRET calculation. In this embodiment, the Wiener parameter is selected as a constant between 0.1 and 0.5 according to the signal-to-noise ratio of the three-channel image.

[0056] Furthermore, the initial phase of the FRET channel X corresponding to the structured light direction angle θ Modulation Depth Spatial frequency vector The parameters can be estimated from the structured light modulated FRET three-channel raw image data set using the existing technology "Gustafsson MGL, et al. Three-dimensional resolution doubling in wide-field fluorescence microscopy by structured illumination. Biophys J, 2008".

[0057] Furthermore, the optical transfer function (OTF) corresponding to the FRET channel X can be obtained by Fourier transforming the point spread function (PSF) obtained by photographing the fluorescent microspheres. In this embodiment, the FRET DD channel The 80nm fluorescent microspheres with 488nm excitation and 525nm emission were measured and Fourier transformed after averaging multiple images; FRETDA, AA channel The data were obtained by measuring 80nm fluorescent microspheres with 561nm excitation and 600nm emission, averaging multiple images and performing Fourier transform.

[0058] S3, performing data alignment processing on the FRET three-channel super-resolution image, specifically:

[0059] by Super-resolution fluorescence images obtained based on channels Affine transformation is performed to achieve pixel-by-pixel alignment of the three-channel images. The affine transformation includes linear transformation operations such as rotation, scaling, shearing, reflection, and translation of the image.

[0060] S4. Perform background subtraction on the FRET three-channel super-resolution image, specifically:

[0061] Reconstruct super-resolution images for each FRET imaging channel Perform pixel-by-pixel grayscale value statistics, and then take the grayscale value corresponding to the first peak in the grayscale value histogram as the background grayscale value of the field of view The image is background corrected and the negative pixel grayscale values ​​are set to zero to obtain the corrected image of each imaging channel. The calculation formula is as follows.

[0062]

[0063] S5, generating a spatial co-localization mask according to the strong correlation characteristics of the FRET area images of the DD and AA channels, and removing the FRET false positive signal caused by the random artifacts of the SR-SIM linear Wiener reconstruction according to the spatial co-localization mask;

[0064] Colocalization analysis is a fluorescence analysis method that determines whether two fluorescent markers are in the same area by analyzing the overlap of different fluorescent markers in space. Its resolution is limited by the fluorescence resolution limit of about 200nm. In the study of the FRET physical process, when the donor and the acceptor have a FRET effect, the typical distance scale of the donor-acceptor pair is 10nm. It can be seen that the typical distance for FRET interaction is much smaller than the resolution limit of the fluorescence microscope. Therefore, it can be considered a priori that the necessary condition for the generation of FRET signals is that the donor and the acceptor have extremely strong spatial colocalization.

[0065] On the other hand, the random noise artifacts generated by SR-SIM linear Wiener reconstruction come from the noise of the imaging system. This noise usually has no correlation. Therefore, the present invention proposes to use the co-localization correlation between fluorescence channels to screen and filter out FRET false positive signals, such as Figure 2 As shown, the specific steps are:

[0066] S5.1. According to and Generate Pearson's coefficient correlation coefficient (PCC) weight matrix r map , as follows:

[0067]

[0068] in, is the mean value of the super-resolution fluorescence image after background subtraction in the donor channel, It represents the mean value of super-resolution fluorescence image after background subtraction in receptor channel;

[0069] S5.2. According to and Generate the Manders colocalization coefficient (MCC) weight matrix R map , as follows:

[0070]

[0071] S5.3, r map With R map Multiply them together to get the mixed co-localization weight matrix, and segment the mixed co-localization weight matrix through the adaptive threshold processing algorithm to generate a binary mask plate B colocalmask , as follows:

[0072]

[0073] Wherein, th is the threshold of the adaptive threshold processing algorithm. The adaptive threshold processing algorithm is a prior art "Wellner PD. Adaptive Thresholding for the Digital Desk [J]. Xerox, 1993." and will not be described in detail here.

[0074] S6. Measure the FRET efficiency and the donor-acceptor concentration ratio, specifically:

[0075]

[0076]

[0077]

[0078] Among them, E SIM is the FRET efficiency under structured light super-resolution, Rc SIM is the FRET donor-acceptor concentration ratio under structured light super-resolution, Fc SIM is the fluorescence intensity of the acceptor sensitized emission, G is the sensitization quenching conversion factor, K is the donor-acceptor concentration conversion factor, and a, b, c, and d are the system crosstalk coefficients.

[0079] Furthermore, the sensitization-quenching conversion factor G, the donor-acceptor concentration conversion factor k, and the system crosstalk coefficients a, b, c, and d can be measured by preparing two standard plasmid samples with different fixed FRET efficiencies and a donor-acceptor concentration ratio of 1:1, as well as separately transfected donor and acceptor plasmid samples. For different FRET donors and acceptors and FRET measurement systems, the system parameters such as a, b, c, d, G, and K are also different. In this embodiment, a=0.010135, b=0.001037, c=0.001641, d=0.111086G=0.621361, and k=0.292009 were obtained through pre-measurement.

[0080] In order to further demonstrate the effect of spatial co-localization mask filtering in super-resolution SIM-FRET and compare the results with those of traditional wide-field FRET, super-resolution SIM-FRET without mask filtering and conventional mask filtering, we use GFP-mCherry as the donor-acceptor pair and simulate the actual system parameters in the above embodiment to construct a simulation model. The model assumes that the donor-acceptor with a fixed FRET efficiency of 30% and a donor-acceptor concentration ratio of 1:1 are all distributed in the white area of ​​the radial star pattern, and the FRET efficiency of other black areas is 0. The radial star pattern can well simulate the sparse and dense distribution characteristics of the donor-acceptor in actual biological samples through the spatial distribution from dense in the center to sparse at the edge. In order to be close to the actual imaging process of microscopic imaging, the parameters consistent with the actual system are used in sequence: (1) simulate the structured light illumination modulation FRET sample; (2) PSF convolve the true value result image of the structured light illumination modulation; (3) add %5 intensity Gaussian and Poisson mixed additive noise; (4) repeat (1-3) according to the FRET three-channel parameters to simulate the FRET three-channel original image data set.

[0081] Figure 3 is a schematic diagram of the reconstruction results of the above model; Figure 3 (a) shows the true value results of the simulation model. Figure 3 (b) shows the result of wide-field reconstruction of the simulated FRET three-channel raw image data set. Figure 3 (c) shows the result of the simulated FRET three-channel raw image data set after structured light super-resolution linear Wiener reconstruction but without mask filtering. Figure 3 (d) shows the results of the simulated FRET three-channel raw image data set after structured light super-resolution linear Wiener reconstruction and mask filtering according to the method previously published by our research group "Automatic background recognition and data screening for online E-FRET quantitative imaging". Figure 3 (e) shows the simulated FRET three-channel original image data set after structured light super-resolution linear Wiener reconstruction and according to and The result of mask filtering using spatial colocalization is Figure 3 (a) and Figure 3 (c) Figure 3 (d) Figure 3 From the comparison of (e), it can be seen that the results of the linear Wiener reconstruction of the FRET three-channel SR-SIM using the unified Wiener reconstruction parameters in the present invention are consistent with the true value; Figure 3 (a) Figure 3 (b) Figure 3(e) It can be seen from the comparison between the SIM-FRET of the present invention that more FRET signals of fine structures beyond the diffraction limit are reconstructed compared with the traditional wide-field FRET calculation, and these fine structure FRET signals are consistent with the true value, further demonstrating that the method of the present invention meets the fidelity requirements of FRET calculation of fluorescence intensity; Figure 3 (c) Figure 3 (d) Figure 3 (e) It can be seen that spatial co-localization mask filtering can effectively remove the FRET false positive signals caused by random artifacts of SR-SIM linear Wiener reconstruction, compared with conventional FRET mask filtering and no mask filtering, and thus achieve quantitative super-resolution detection of FERT.

[0082] Figure 4 Schematic diagram of the effect of the method of the present invention at different signal-to-noise ratios, Figure 4 (a) is a schematic diagram of the effect of the co-localization mask filtering method of the present invention at different signal-to-noise ratios, Figure 4 (b) is a schematic diagram of the effect of the conventional FRET mask filtering method at different signal-to-noise ratios. It can be seen that the colocalization mask filtering method of the present invention has better false positive signal filtering and fine structure retention capabilities under different signal-to-noise ratio conditions compared to the conventional FRET mask filtering method, further demonstrating the robustness of the method of the present invention.

[0083] The above embodiments are preferred implementation modes of the present invention, but the implementation modes of the present invention are not limited to the above embodiments. Any other changes, modifications, substitutions, combinations, and simplifications that do not deviate from the spirit and principles of the present invention should be equivalent replacement methods and are included in the protection scope of the present invention.

Claims

1. A super-resolution SIM-FRET image reconstruction method based on spatial co-localization mask filtering, characterized in that: The steps include: Acquire a FRET three-channel raw image data group after structured light modulation; the FRET three-channel includes a donor excitation donor emission channel DD, an acceptor excitation acceptor emission channel DA, and a donor excitation acceptor emission channel AA; Based on the unified Wiener reconstruction parameters, the FRET three-channel structured light super-resolution linear Wiener reconstruction is performed to obtain the FRET three-channel super-resolution image data, specifically: The FRET three-channel original image group is subjected to structured light super-resolution linear Wiener reconstruction using unified Wiener reconstruction parameters to obtain the FRET three-channel super-resolution image data. : ； in, For FRET channel X Fourier transform of the corresponding image; k is the coordinate of the fluorescence image pixel transformed into the frequency domain, For FRET channel X Corresponding structured light direction angle θ The initial phase of For FRET channel X Corresponding structured light direction angle θ The modulation depth under For FRET channel X Corresponding structured light direction angle θ The spatial frequency of is the optical transfer function of the SR-SIM system, Indicates FRET channel X The optical transfer function of the SR-SIM system is calculated in the frequency domain according to the spatial frequency vector and Spectrum shifting, , Respectively The conjugate, channel X It includes a donor excitation donor emission channel DD, an acceptor excitation acceptor emission channel DA, and a donor excitation acceptor emission channel AA; w To reconstruct the Wiener parameters, is a Gaussian apodization function, iFFT is the inverse Fourier transform of the image; Perform data alignment on the FRET three-channel super-resolution images; Perform background subtraction on the FRET three-channel super-resolution image; Generate a spatial co-localization mask according to the strong correlation characteristics of the FRET area images of the DD and AA channels, and remove FRET false positive signals according to the spatial co-localization mask; The super-resolution images after removing the FRET false positive signals were used to measure the FRET efficiency and the donor-acceptor concentration ratio.

2. The super-resolution SIM-FRET image reconstruction method based on spatial co-localization mask filtering according to claim 1, characterized in that: The method of obtaining the FRET three-channel original image data set after structured light modulation is specifically as follows: The SR-SIM system acquires three-channel raw image data sets based on channel-sensitized intensity measurement, including the donor channel image set , receptor channel image group And FRET channel image group ;in is the spatial coordinate of the fluorescence image pixel, each group of original image data contains several different structured light azimuth angles, and each structured light azimuth angle contains several cosine structured light illumination wide-field fluorescence images with different phase differences.

3. The super-resolution SIM-FRET image reconstruction method based on spatial co-localization mask filtering according to claim 1, characterized in that: The data alignment process for the FRET three-channel super-resolution image is specifically as follows: by Super-resolution fluorescence images obtained based on channels Perform affine transformation to achieve pixel-by-pixel alignment of three-channel images.

4. The super-resolution SIM-FRET image reconstruction method based on spatial co-localization mask filtering according to claim 1, characterized in that: The background subtraction process for the FRET three-channel super-resolution image is specifically as follows: Reconstruct super-resolution images for each FRET imaging channel Perform pixel-by-pixel grayscale value statistics, and then take the grayscale value corresponding to the first peak in the grayscale value histogram as the background grayscale value of the field of view , perform background correction on the image and set the negative pixel grayscale values ​​to zero to obtain the corrected image of each imaging channel, specifically: 。 5. The super-resolution SIM-FRET image reconstruction method based on spatial co-localization mask filtering according to claim 1, characterized in that: The spatial co-localization mask is generated according to the strong correlation characteristics of the FRET area images of the DD and AA channels, specifically: according to and Generate Pearson's coefficient correlation coefficient weight matrix , as follows: ； in, is the mean value of the super-resolution fluorescence image after background subtraction in the donor channel, It represents the mean value of super-resolution fluorescence image after background subtraction in receptor channel; according to and Generate Manders colocalization coefficient weight matrix , as follows: ； Will and The hybrid co-localization weight matrix is ​​obtained by multiplication, and the hybrid co-localization weight matrix is ​​segmented by an adaptive threshold processing algorithm to generate a binary mask. , as follows: ； in, th is the threshold of the adaptive thresholding algorithm.

6. The super-resolution SIM-FRET image reconstruction method based on spatial co-localization mask filtering according to claim 1, characterized in that: The measurement of the FRET efficiency and the donor-acceptor concentration ratio is specifically as follows: ； ； ； in, is the FRET efficiency under structured light super-resolution, is the FRET donor-acceptor concentration ratio under structured light super-resolution, is the fluorescence intensity of receptor sensitization emission, G is the sensitization-quenching conversion factor, K is the donor-acceptor concentration conversion factor, a , b , c , d is the system crosstalk coefficient.

Images