A structured light illumination super-resolution microscopy image reconstruction method and system

Abstract

The purpose of this invention is to address the technical problems of existing technologies, such as long processing time, low imaging efficiency, and image artifacts caused by phase shift errors, which affect researchers' observation and judgment of the detailed structure of biological samples. This invention provides a method and system for reconstructing super-resolution microscopic images under structured illumination. The method uses one wide-field fluorescence image and one structured illumination fluorescence image in each of N different spatial directions, for a total of N+1 original images. Using Hilbert transform, three structured illumination images are generated in each of the N spatial directions: one containing the original structured illumination fluorescence image with 0 phase shift, and two with phase shifts of 2π / 3 and 4π / 3, for a total of 3N structured illumination images. Finally, the final super-resolution image is obtained using a structured illumination super-resolution image reconstruction algorithm. This method effectively avoids the artifact problem caused by phase shift errors, while significantly reducing the number of original images acquired and improving image reconstruction efficiency.

Description

Technical Field

[0001] This invention relates to microscopic imaging methods and systems, specifically to a method for reconstructing super-resolution microscopic images under structured light illumination and a system for implementing the method. Background Technology

[0002] Optical microscopy is a core research tool in the biomedical field, but its spatial resolution is limited by the diffraction limit, which has become a major technological bottleneck. The spatial resolution limit of ordinary optical microscopes is about 200 nm, which cannot meet the high-resolution imaging requirements for studying the fine internal structures of cells. Therefore, super-resolution microscopy is a research hotspot in the field of optical microscopy.

[0003] Currently, with the continuous development of light field modulation and fluorescent probe technology, a number of novel super-resolution microscopy imaging methods have emerged. Single-molecule localization microscopy can achieve nanometer-level resolution, but reconstructing a single super-resolution image requires acquiring tens of thousands of original images, which is too time-consuming and inefficient. Stimulated emission depletion microscopy can also achieve nanometer-level resolution, but the optical path design of this approach is complex, and the high-intensity laser used can affect the activity of biological samples, making it difficult to apply to dynamic imaging of living samples.

[0004] Structured Illumination Microscopy (SIM) has become the preferred technique for super-resolution dynamic imaging of living biological tissues due to its advantages such as high spatial resolution, fast imaging speed, low phototoxicity, and outstanding three-dimensional imaging capabilities. Its core principle is to utilize low-frequency moiré fringes generated by the superposition of high-frequency fringes, and to extract high-frequency information from the sample through decoding algorithms, thereby overcoming the diffraction limit to achieve super-resolution imaging. However, traditional structured illumination microscopy typically requires the acquisition of three images with a phase shift difference of [missing information - likely a value or value] in two spatial directions: 0° and 90°. The cosine structured light illumination image, or three images with phase shift differences of 0°, 120° and 240° were acquired respectively in three spatial directions. For cosine structured light illumination images, the number of original images acquired needs to reach 3N (N is the number of illumination directions, N≥2). This process is time-consuming and has low imaging efficiency. At the same time, phase shift errors introduced by vibrations in the experimental environment can cause artifacts in the super-resolution reconstructed images, directly affecting researchers' observation and judgment of the detailed structure of biological samples. Summary of the Invention

[0005] The purpose of this invention is to solve the technical problems of long time consumption, low imaging efficiency, and image artifacts caused by phase shift errors, which affect researchers' observation and judgment of the detailed structure of biological samples. The invention provides a method and system for reconstructing super-resolution microscopic images under structured light illumination.

[0006] To achieve the above objectives, the technical solution provided by the present invention is as follows:

[0007] A method for reconstructing super-resolution microscopic images under structured light illumination, characterized by:

[0008] S1. A wide-field illumination light field with uniform light intensity distribution is generated by using a light source module and a spatial light modulation module, and the sample is illuminated by the wide-field illumination light field to acquire a wide-field fluorescence image.

[0009] S2. A structured light illumination field with light intensity satisfying a one-dimensional cosine function distribution is generated by time-division modulation using a light source module and a spatial light modulation module. The sample is then illuminated sequentially through the structured light illumination field with light intensity satisfying a one-dimensional cosine function distribution to acquire N structured light illumination fluorescence images, where N≥2.

[0010] S3. For the wide-field fluorescence image, subtract it from N structured illumination fluorescence images respectively, and perform Hilbert transform on each image to generate structured illumination images with phase shifts of 2π / 3 and 4π / 3, so that 3 structured illumination images are obtained in each of the N spatial directions, for a total of 3N structured illumination images; each of the 3 structured illumination images in each spatial direction contains the original structured illumination image with 0 phase shift, and structured illumination images with phase shifts of 2π / 3 and 4π / 3.

[0011] S4. Based on the 3N structured light illumination images, the final super-resolution image is obtained according to the structured light illumination super-resolution image reconstruction algorithm.

[0012] Furthermore, in step S2, the intensity distribution of the structured light illumination field in N different spatial directions all satisfy the following formula:

[0013]

[0014] In the formula, The light intensity of the structured light illumination field in the i-th direction is represented by , i = 1, 2, ..., N; r represents the coordinate of the sample plane along the r-direction; I0 represents the average light intensity of the wide-field illumination field; m represents the modulation degree; p represents the spatial frequency. Indicates the initial phase.

[0015] Furthermore, step S2 specifically includes:

[0016] The light intensity is modulated by a light source module and a spatial light modulation module to generate two structured light illumination fields in two different spatial directions with light intensity satisfying a one-dimensional cosine function distribution. The sample is then illuminated sequentially by the two structured light illumination fields in different spatial directions to acquire two structured light illumination fluorescence images.

[0017] The angle between the adjacent directions of the two different spatial directions is 90°, and the two different spatial directions are 0° and 90° respectively.

[0018] Furthermore, step S2 specifically includes:

[0019] The light intensity of the structured light illumination field in three different spatial directions is generated by modulating the light field with a light source module and a spatial light modulation module to satisfy a one-dimensional cosine function distribution. The sample is then illuminated sequentially by the structured light illumination field in the three different spatial directions to acquire three structured light illumination fluorescence images.

[0020] The angle between adjacent directions of the three different spatial directions is 120°, and the three different spatial directions are 0°, 120° and 240° respectively.

[0021] Furthermore, in step S2, all N structured light illumination fluorescence images satisfy the following formula:

[0022]

[0023] In the formula, This represents the original 0-phase-shift structured illumination fluorescence image in the i-th direction, i=1,2…N; Indicates the fluorescence intensity of the sample; The point spread function (PSF) of an imaging system; Let represent the light intensity of the structured light illumination field along the i-th direction of the sample plane along the r direction, where i = 1, 2, ..., N; Indicates the concentration of labeled fluorescent molecules in the sample; Represents the convolution operation; where and , Proportional.

[0024] Furthermore, step S3 specifically includes:

[0025] S3.1. Subtract the wide-field fluorescence image from N structured light illumination fluorescence images to obtain N difference images;

[0026] S3.2. Perform Hilbert transform on the N difference images respectively to generate structured light illumination images with phase shifts of 2π / 3 and 4π / 3, so that 3 structured light illumination images are obtained in each of the N spatial directions, for a total of 3N structured light illumination images; each of the 3 structured light illumination images in each spatial direction contains the original structured light illumination fluorescence image with 0 phase shift and the structured light illumination images with phase shifts of 2π / 3 and 4π / 3.

[0027] Further, in step S3.2, the expressions for the structured light illumination images with phase shifts of 2π / 3 and 4π / 3 are respectively:

[0028] ;

[0029] ;

[0030] In the formula, Let i represent the structured light illumination image with a phase shift of 2π / 3 in the i-th direction, where i = 1, 2, ..., N; Let i represent the structured light illumination image with a phase shift of 4π / 3 in the i-th direction, where i = 1, 2, ..., N; This represents a wide-field fluorescence image, i.e., the fundamental frequency component; This represents the original 0-phase-shift structured illumination fluorescence image in the i-th direction, i=1,2…N; Indicates the in-plane edge of the sample The direction-modulated cosine image is subjected to Hilbert transform.

[0031] Furthermore, in step S4:

[0032] The structured light illumination super-resolution image reconstruction algorithm employs at least one of the fair-SIM algorithm, HiFi-SIM algorithm, and Direct-SIM algorithm.

[0033] Meanwhile, the present invention also provides a structured light illumination super-resolution microscopic image reconstruction system for implementing the above-mentioned structured light illumination super-resolution microscopic image reconstruction method; its special feature is that it includes:

[0034] The light source module is used to provide illumination light;

[0035] The light field modulation module is coupled to the light source of the light source module and is used to modulate the illumination light provided by the light source module into a wide-field illumination light field with uniform light intensity distribution, and a structured light illumination light field with N different spatial directions whose light intensity satisfies a one-dimensional cosine function distribution, where N≥2;

[0036] The microscopic imaging module is used to illuminate the sample with a wide-field illumination light field and a structured light illumination light field to achieve fluorescence imaging, obtaining one wide-field fluorescence image and N structured light illumination fluorescence images;

[0037] The image acquisition module, electrically connected to the microscopic imaging module, is used to acquire one wide-field fluorescence image and N structured light illumination fluorescence images of the sample.

[0038] The data processing module, electrically connected to the image acquisition module, is used to process the one wide-field fluorescence image and N structured light illumination fluorescence images to obtain the final super-resolution image.

[0039] Furthermore, the light field modulation module employs a liquid crystal spatial light modulator (SLM), a digital micromirror device (DMD), a one-dimensional physical grating, or a metasurface device.

[0040] Compared with the prior art, the present invention has the following advantages:

[0041] (1) The present invention provides a structured light illumination super-resolution microscopic image reconstruction method, which can generate structured light illumination images with phase shifts of 2π / 3 and 4π / 3 in each direction by using only one wide-field fluorescence image and one structured light illumination fluorescence image in each of N different directions (N≥2). This significantly reduces the number of images acquired and improves imaging efficiency. At the same time, the introduction of Hilbert transform can obtain super-resolution images of the sample without reducing the quality of the reconstructed image, effectively avoiding artifact problems caused by phase shift errors. It also reduces the complexity of experimental operations, significantly reduces the number of original images acquired, and improves image reconstruction efficiency.

[0042] (2) The present invention provides a structured light illumination super-resolution microscopic image reconstruction method that is compatible with various structured light illumination microscopic imaging systems and applicable to various back-end structured light illumination super-resolution image reconstruction algorithms in the frequency domain and spatial domain, and has good universality. Attached Figure Description

[0043] Figure 1 This is a flowchart of a structured light illumination super-resolution microscopic image reconstruction method according to the present invention;

[0044] Figure 2 This is a flowchart illustrating the working principle of Embodiment 1 of the structured light illumination super-resolution microscopic image reconstruction method of the present invention (steps S1 and S2 are omitted).

[0045] Figure 3 This is a flowchart illustrating the working principle of Embodiment 2 of the structured light illumination super-resolution microscopic image reconstruction method of the present invention (steps S1 and S2 are omitted).

[0046] Figure 4This is a comparison of the ordinary wide-field fluorescence image of bovine pulmonary artery endothelial cells acquired in step S2 of Example 1 of the present invention, the super-resolution image reconstructed using several SIM super-resolution algorithms in step S4, and the super-resolution image reconstructed using several SIM super-resolution algorithms on the same bovine pulmonary artery endothelial cells using conventional structured illumination micro-imaging technology. Specifically, (a) is the wide-field fluorescence image of the bovine pulmonary artery endothelial cell sample acquired in step S2 of Example 1; (b) is the fair-SIM algorithm super-resolution image of the bovine pulmonary artery endothelial cells reconstructed using the fair-SIM algorithm on the conventional structured illumination micro-imaging technology; (c) is the HiFi-SIM algorithm super-resolution image of the bovine pulmonary artery endothelial cells reconstructed using the HiFi-SIM algorithm on the conventional structured illumination micro-imaging technology; and (d) is the super-resolution image of the bovine pulmonary artery endothelial cells reconstructed using the HiFi-SIM algorithm on the same bovine pulmonary artery endothelial cells. (e) is the super-resolution image of the bovine pulmonary artery endothelial cells reconstructed using the HiFi-SIM algorithm in step S4 of Example 1; (f) is the super-resolution image of the bovine pulmonary artery endothelial cells reconstructed using the Direct-SIM algorithm using the conventional structured illumination micro-imaging technique; (a1)-(f1) are magnified views of (a)-(f), respectively; (g) is the residual image of the super-resolution image of the HiFi-SIM algorithm and the super-resolution image of the HiFi-HT-SIM algorithm; (h) is the residual image of the super-resolution image of the Direct-SIM algorithm and the super-resolution image of the Direct-HT-SIM algorithm.

[0047] Figure 5The images show a wide-field fluorescence image of the endoplasmic reticulum (ER) of COS-7 cells used in Example 2 of this invention, a super-resolution image reconstructed using several SIM super-resolution algorithms in step S4, and a comparison with a super-resolution image reconstructed using several SIM super-resolution algorithms on the ER of COS-7 cells by conventional structured illumination micro-imaging technology. Specifically, (a) is a wide-field fluorescence image of the ER of COS-7 cells; (b) is a super-resolution image reconstructed using the fair-SIM algorithm on the ER of COS-7 cells by conventional structured illumination micro-imaging technology; (c) is a super-resolution image reconstructed using the HiFi-SIM algorithm on the ER of COS-7 cells by conventional structured illumination micro-imaging technology; and (d) is a super-resolution image reconstructed using the HiFi-SIM algorithm on the ER of COS-7 cells by conventional structured illumination micro-imaging technology. (e) is a super-resolution image of the HiFi-HT-SIM algorithm reconstructed using the HiFi-SIM algorithm; (f) is a super-resolution image of the endoplasmic reticulum of COS-7 cells reconstructed using the Direct-SIM algorithm using the conventional structured illumination micro-imaging technique; (f) is a super-resolution image of the Direct-HT-SIM algorithm reconstructed using the Direct-SIM algorithm in step S4 of Example 2; (a1)-(f1) are magnified views of (a)-(f), respectively; (g) is a residual image of the HiFi-SIM algorithm super-resolution image and the HiFi-HT-SIM algorithm super-resolution image; (h) is a residual image of the Direct-SIM algorithm super-resolution image and the Direct-HT-SIM algorithm super-resolution image.

[0048] Figure 6 This is a schematic diagram of the structural principle of a structured light illumination super-resolution microscopic image reconstruction system according to the present invention. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0050] like Figure 1 As shown, the present invention provides a method for reconstructing super-resolution microscopic images under structured light illumination, comprising the following steps:

[0051] S1. A wide-field illumination light field with uniform light intensity distribution is generated by using a light source module and a light field modulation module, and the sample is illuminated by the wide-field illumination light field to acquire a wide-field fluorescence image;

[0052] S2. A structured light illumination field with light intensity satisfying a one-dimensional cosine function distribution is generated by using a light source module and a light field modulation module. The sample is then illuminated sequentially by the structured light illumination field with light intensity satisfying a one-dimensional cosine function distribution to acquire N structured light illumination fluorescence images, where N≥2.

[0053] S3. For the wide-field fluorescence image, subtract it from N structured illumination fluorescence images respectively, and perform Hilbert transform on each image to generate structured illumination images with phase shifts of 2π / 3 and 4π / 3, so that 3 structured illumination images are obtained in each of the N spatial directions, for a total of 3N structured illumination images; each of the 3 structured illumination images in each spatial direction contains the original structured illumination fluorescence image with 0 phase shift, and the structured illumination images with phase shifts of 2π / 3 and 4π / 3.

[0054] S4. Based on the 3N structured light illumination images, the final super-resolution image is obtained according to the structured light illumination super-resolution image reconstruction algorithm.

[0055] To achieve the above method, the present invention uses, as follows: Figure 6 The system described herein is a structured light illumination super-resolution microscopic image reconstruction system. The system comprises a light source module providing illumination light, which is modulated by a light field modulation module into a wide-field illumination light field with uniform intensity distribution, and N structured light illumination light fields with intensity satisfying a one-dimensional cosine function distribution. A microscopic imaging module performs fluorescence imaging. An image acquisition module acquires one wide-field fluorescence image and N structured light illumination fluorescence images. A data processing module uses the wide-field fluorescence image and one structured light illumination fluorescence image from each of the N different directions (N≥2) to generate structured light illumination images with phase shifts of 2π / 3 and 4π / 3 in each direction. Compared to the traditional method requiring 3N original images, this application reduces the original image acquisition to N+1 images (6 images reduced to 3, 9 images reduced to 4), avoiding fringe phase shift control, improving imaging speed, and being compatible with various backend structured light illumination super-resolution image reconstruction algorithms, exhibiting good generalization ability and compatibility.

[0056] Example 1

[0057] like Figure 2 As shown, Example 1 includes the following steps:

[0058] S1. A wide-field illumination light field with uniform intensity distribution is generated by modulation, and the sample is illuminated by the wide-field illumination light field to acquire a wide-field fluorescence image. ;

[0059] S2. Time-division modulation generates two structured light illumination fields in two different spatial directions with light intensity satisfying a one-dimensional cosine function distribution. The sample is then illuminated sequentially by these two structured light illumination fields to acquire two structured light illumination fluorescence images. and The angle between adjacent directions of the two structured light illumination fields in different spatial directions. The two different spatial directions are 0° and 90° respectively;

[0060] The light intensity distribution of the structured light illumination field satisfies the following formula:

[0061]

[0062] In the formula, The light intensity of the structured light illumination field in the i-th direction is represented by i=1,2, which represent the two spatial directions of 0° and 90°; r represents the coordinates of the sample plane along the r direction; I0 represents the average light intensity of the wide-field illumination field; m represents the modulation degree; and p represents the spatial frequency. Indicates the initial phase;

[0063] Both structured light illumination fluorescence images satisfy the following formula:

[0064]

[0065] In the formula, The structured illumination fluorescence image represents the original 0-phase-shifted structured light illumination image in the i-th direction, where i=1,2; Indicates the fluorescence intensity of the sample; The point spread function (PSF) of an imaging system; This represents the light intensity of the structured light illumination field along the i-th direction of the sample plane along the r direction; Indicates the concentration of labeled fluorescent molecules in the sample; Represents the convolution operation; where and , Proportional;

[0066] S3. The wide-field fluorescence image is subtracted from the two structured illumination fluorescence images respectively, and Hilbert transform is performed on them respectively to generate structured illumination images with phase shifts of 2π / 3 and 4π / 3, so that three structured illumination images are obtained in the two spatial directions respectively, including the original structured illumination fluorescence image with 0 phase shift and the structured illumination images with phase shifts of 2π / 3 and 4π / 3, for a total of six structured illumination images;

[0067] S3.1. Subtract the wide-field fluorescence image from each of the N structured light illumination fluorescence images to obtain N difference images; the difference images satisfy the following formula:

[0068]

[0069] After applying a Hilbert transform, the sinusoidal structured light illumination image satisfies the following formula:

[0070]

[0071] In the formula, Represents the cosine component of the original structured light fluorescence illumination image; Represents the sinusoidal component of the original structured light fluorescence illumination image; This represents the local phase angle at spatial location r; Indicates the in-plane edge of the sample The direction-modulated cosine image is subjected to Hilbert transform; some parameter expressions are as follows:

[0072]

[0073]

[0074]

[0075] S3.2. Perform Hilbert transform on the N difference images respectively to generate structured illumination images with phase shifts of 2π / 3 and 4π / 3, so that structured illumination fluorescence images containing the original 0 phase shift are obtained in the two spatial directions respectively. Structured light illumination images with a phase shift of 2π / 3 and structured light illumination images with a phase shift of 4π / 3 A total of 6 structured light illumination images;

[0076] The structured light illumination image expressions for phase shifts of 2π / 3 and 4π / 3 are as follows:

[0077]

[0078]

[0079] In the formula, Let i represent the structured light illumination image with a phase shift of 2π / 3 in the i-th direction, where i=1,2; Let i represent the structured light illumination image with a phase shift of 4π / 3 in the i-th direction, where i=1,2; This represents a wide-field fluorescence image, i.e., the fundamental frequency component; The structured illumination fluorescence image with original 0 phase shift in the i-th direction is represented by i=1,2; Indicates the in-plane edge of the sample Perform Hilbert transform on the direction-modulated cosine image;

[0080] S4. Based on the 6 structured light illumination images, the final super-resolution image is obtained according to the structured light illumination super-resolution image reconstruction algorithm.

[0081] Based on the above-mentioned structured light illumination super-resolution microscopic image reconstruction method, Example 1 also provides a structured light illumination super-resolution microscopic image reconstruction system, including a light source module, a light field modulation module, a microscopic imaging module, an image acquisition module, and a data processing module; taking an experiment targeting bovine pulmonary artery endothelial cells (BPAE) as an example:

[0082] The light source module uses a laser, and the light field modulation module uses a liquid crystal spatial light modulator (SLM). The laser provides illumination light, which is then time-division modulated by the SLM to generate a wide-field illumination light field with uniform intensity distribution, and two structured light illumination fields in different spatial directions with intensity satisfying a one-dimensional cosine function distribution. The angle between adjacent directions of the two structured light illumination fields in different spatial directions is... The two different spatial directions are 0° and 90° respectively.

[0083] The microscopic imaging module employs an interferometric SIM super-resolution microscopy system. It illuminates bovine pulmonary artery endothelial cells (BPAE) samples using a wide-field illumination field and a structured illumination field, exciting fluorescent molecules of labeled organelles to emit light, thus achieving fluorescence imaging. The image acquisition module uses a CMOS area array camera to acquire wide-field fluorescence images and two structured illumination fluorescence images of the BPAE samples. Finally, the data processing module processes the wide-field fluorescence images and two structured illumination fluorescence images using Hilbert transform, obtaining six structured illumination images. These are then further processed using various SIM super-resolution algorithms in the backend to obtain the final super-resolution images. A comparison is made between the wide-field fluorescence images of the BPAE samples acquired in step S2, the super-resolution images reconstructed by each SIM super-resolution algorithm, and the super-resolution images reconstructed using various SIM super-resolution algorithms for the same BPAE samples using traditional structured illumination microscopy. Figure 4As shown, (a) is a wide-field fluorescence image of the bovine pulmonary artery endothelial cell sample acquired in step S2; (b) is a super-resolution image of the bovine pulmonary artery endothelial cells reconstructed using the fair-SIM algorithm by conventional structured illumination micro-imaging technology; (c) is a super-resolution image of the bovine pulmonary artery endothelial cells reconstructed using the HiFi-SIM algorithm by conventional structured illumination micro-imaging technology; (d) is a super-resolution image of the bovine pulmonary artery endothelial cells reconstructed using the HiFi-SIM algorithm in step S4; (e) is a super-resolution image of the bovine pulmonary artery endothelial cells reconstructed using the Direct-SIM algorithm by conventional structured illumination micro-imaging technology; and (f) is a super-resolution image of the bovine pulmonary artery endothelial cells reconstructed using the Direct-SIM algorithm in step S4. Super-resolution images of the rect-HT-SIM algorithm; (a1)-(f1) are magnified views of (a)-(f), respectively; (g) is the residual image of the HiFi-SIM algorithm super-resolution image and the HiFi-HT-SIM algorithm super-resolution image. The maximum gray value is 0.40, the minimum gray value is 0, and the average gray value is 0.04. The root mean square error (RMSE) is used to measure the deviation between the two algorithms, and the RMSE is 0.05; (h) is the residual image of the Direct-SIM algorithm super-resolution image and the Direct-HT-SIM algorithm super-resolution image. The maximum gray value is 0.39, the minimum gray value is 0, the average gray value is 0.02, and the RMSE is 0.04. The above statistical results show that the reconstruction method of this embodiment 1 can achieve a super-resolution reconstruction effect that is almost the same as that of traditional structured light micro-imaging technology.

[0084] Example 2

[0085] like Figure 3 As shown, Example 2 includes the following steps:

[0086] S1. A wide-field illumination light field with uniform intensity distribution is generated by modulation, and the sample is illuminated by the wide-field illumination light field to acquire a wide-field fluorescence image. ;

[0087] S2. Time-division modulation generates structured light illumination fields in three different spatial directions with light intensity satisfying a one-dimensional cosine function distribution. The sample is then illuminated sequentially through these three structured light illumination fields to acquire three structured light illumination fluorescence images. and The angle between adjacent directions of the structured light illumination fields in the three different spatial directions; The three different spatial directions are 0°, 120° and 240° respectively;

[0088] The light intensity distribution of the structured light illumination field satisfies the following formula:

[0089]

[0090] In the formula, The light intensity of the structured light illumination field in the i-th direction is represented by i = 1, 2, 3, which represent the three spatial directions of 0°, 120°, and 240°; r represents the coordinates of the sample plane along the r-direction; I0 represents the average light intensity of the wide-field structured light field; m represents the modulation degree; and p represents the spatial frequency. Indicates the initial phase;

[0091] The three structured light illumination fluorescence images all satisfy the following formula:

[0092]

[0093] In the formula, The structured illumination fluorescence image represents the original 0-phase-shift image in the i-th direction, where i = 1, 2, 3; Indicates the fluorescence intensity of the sample; The point spread function (PSF) of an imaging system; This represents the light intensity of the structured light illumination field along the i-th direction of the sample plane along the r direction; Indicates the concentration of labeled fluorescent molecules in the sample; Represents the convolution operation; where and , Proportional;

[0094] S3. The wide-field fluorescence image is subtracted from the three structured illumination fluorescence images respectively, and Hilbert transform is performed on each image to generate structured illumination images with phase shifts of 2π / 3 and 4π / 3. This results in three structured illumination images in three spatial directions, each containing the original structured illumination fluorescence image with a phase shift of 0 and the structured illumination images with phase shifts of 2π / 3 and 4π / 3, for a total of nine structured illumination images.

[0095] S3.1. Subtract the wide-field fluorescence image from each of the N structured light illumination fluorescence images to obtain N difference images; the difference images satisfy the following formula:

[0096] ;

[0097] After applying a Hilbert transform, the sinusoidal structured light illumination image satisfies the following formula:

[0098] ;

[0099] In the formula, Represents the cosine component of the original structured light fluorescence illumination image; Represents the sinusoidal component of the original structured light fluorescence illumination image; This represents the local phase angle at spatial location r; Indicates the in-plane edge of the sample The direction-modulated cosine image is subjected to Hilbert transform; some parameter expressions are as follows:

[0100]

[0101]

[0102]

[0103] S3.2. Perform Hilbert transform on the N difference images respectively to generate structured illumination images with phase shifts of 2π / 3 and 4π / 3, so that structured illumination fluorescence images containing the original 0 phase shift are obtained in the three spatial directions respectively. Structured light illumination images with a phase shift of 2π / 3 and structured light illumination images with a phase shift of 4π / 3 A total of 9 structured light illumination images;

[0104] The structured light illumination image expressions for phase shifts of 2π / 3 and 4π / 3 are as follows:

[0105]

[0106]

[0107] In the formula, Let i represent the structured light illumination image with a phase shift of 2π / 3 in the i-th direction, where i = 1, 2, 3; Let i represent the structured light illumination image with a phase shift of 4π / 3 in the i-th direction, where i = 1, 2, 3; This represents a wide-field fluorescence image, i.e., the fundamental frequency component; The structured illumination fluorescence image with original 0 phase shift in the i-th direction is represented by i=1,2,3; Indicates the in-plane edge of the sample Perform Hilbert transform on the direction-modulated cosine image;

[0108] S4. Based on the nine structured light illumination images, the final super-resolution image is obtained according to the structured light illumination super-resolution image reconstruction algorithm.

[0109] Based on the above-mentioned structured light illumination super-resolution microscopic image reconstruction method, Embodiment 2 also provides a structured light illumination super-resolution microscopic image reconstruction system, including a light source module, a light field modulation module, a microscopic imaging module, an image acquisition module, and a data processing module.

[0110] The light source module uses light-emitting diodes (LEDs), and the light field modulation module uses a digital micromirror device (DMD). Illumination light is provided by the LEDs, and the DMD modulates this illumination light in a time-division manner to generate a wide-field illumination light field with uniform intensity distribution, as well as structured light illumination fields in three different spatial directions whose light intensity satisfies a one-dimensional cosine function distribution; the angle between adjacent directions of the three different spatial directions of the structured light illumination fields... The three different spatial directions are 0°, 120° and 240°.

[0111] The microscopic imaging module uses wide-field illumination and structured light illumination to illuminate the sample to achieve fluorescence imaging. The image acquisition module uses a CCD or CMOS area array digital camera to acquire wide-field fluorescence images and three structured light illumination fluorescence images of the sample. Finally, the data processing module processes the wide-field fluorescence images and the three structured light illumination fluorescence images through Hilbert transform to obtain nine structured light illumination images. Then, it calls the backend SIM super-resolution algorithms for further processing to obtain the final super-resolution image.

[0112] Using wide-field fluorescence images and structured light illumination fluorescence images of the endoplasmic reticulum (ER) of COS-7 cells provided by the open-source biological image database BioSR (DOI:10.1038 / s41592-020-01048-5) as raw data, the aforementioned data processing module invoked various backend SIM super-resolution algorithms to reconstruct super-resolution images. Traditional structured light illumination micro-imaging techniques were also used to reconstruct super-resolution images of the ER of COS-7 cells. The wide-field fluorescence images of the ER of COS-7 cells, the super-resolution images reconstructed using several SIM super-resolution algorithms in step S4, and the super-resolution images of the ER of COS-7 cells reconstructed using several SIM super-resolution algorithms under traditional structured light illumination micro-imaging techniques were compared. Figure 5As shown, (a) is a wide-field fluorescence image of the endoplasmic reticulum (ER) of COS-7 cells; (b) is a super-resolution image of the ER of COS-7 cells reconstructed using the fair-SIM algorithm by conventional structured illumination micro-imaging technique; (c) is a super-resolution image of the ER of COS-7 cells reconstructed using the HiFi-SIM algorithm by conventional structured illumination micro-imaging technique; (d) is a super-resolution image of the ER of COS-7 cells reconstructed using the HiFi-SIM algorithm in step S4; (e) is a super-resolution image of the ER of COS-7 cells reconstructed using the Direct-SIM algorithm by conventional structured illumination micro-imaging technique; and (f) is a super-resolution image of the ER of COS-7 cells reconstructed using the Direct-SIM algorithm in step S4. The super-resolution image of the Direct-HT-SIM algorithm reconstructed by the algorithm; (a1)-(f1) are local magnified images of (a)-(f) respectively; (g) is the residual image of the HiFi-SIM algorithm super-resolution image and the HiFi-HT-SIM algorithm super-resolution image. The statistical results show that its maximum gray value is 0.31, minimum gray value is 0, average gray value is 0.01, and root mean square error is 0.02; similarly, (h) is the residual image of the Direct-SIM algorithm super-resolution image and the Direct-HT-SIM algorithm super-resolution image. The statistical results show that its maximum gray value is 0.36, minimum gray value is 0, average gray value is 0.007, and root mean square error is 0.01; the above statistical results show that the reconstruction method of this embodiment 2 can achieve a super-resolution reconstruction effect that is almost the same as that of traditional structured light micro-imaging technology.

[0113] The above are merely embodiments of the present invention and are not intended to limit the invention. Various modifications and variations can be made to the invention by those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the scope of the technical solution of the present invention should be included within the protection scope of the present invention. Furthermore, it should be noted that the accompanying drawings are merely illustrative and are not drawn to scale, and should not be construed as limiting the actual scope of protection claimed by the present invention.

Claims

1. A method for reconstructing super-resolution microscopic images under structured light illumination, characterized in that, Includes the following steps: S1. A wide-field illumination light field with uniform light intensity distribution is generated by using a light source module and a light field modulation module, and the sample is illuminated by the wide-field illumination light field to acquire a wide-field fluorescence image; S2. The light source module and the light field modulation module are used to modulate the light intensity in N different spatial directions to generate structured light illumination fields with light intensity satisfying a one-dimensional cosine function distribution. The sample is then illuminated sequentially through the N different spatial directions of the structured light illumination fields to acquire N structured light illumination fluorescence images, where N≥2. S3. For the wide-field fluorescence image, subtract it from N structured illumination fluorescence images respectively, and perform Hilbert transform on each image to generate structured illumination images with phase shifts of 2π / 3 and 4π / 3, so that 3 structured illumination images are obtained in each of the N spatial directions, for a total of 3N structured illumination images; each of the 3 structured illumination images in each spatial direction contains the original structured illumination fluorescence image with 0 phase shift, and the structured illumination images with phase shifts of 2π / 3 and 4π / 3. S4. Based on the 3N structured light illumination images, the final super-resolution image is obtained according to the structured light illumination super-resolution image reconstruction algorithm.

2. The method for reconstructing super-resolution microscopic images under structured light illumination according to claim 1, characterized in that, In step S2, the intensity distribution of the structured light illumination field in N different spatial directions all satisfy the following formula: ； In the formula, The light intensity of the structured light illumination field in the i-th direction is represented by , i = 1, 2, ..., N; r represents the coordinate of the sample plane along the r-direction; I0 represents the average light intensity of the wide-field illumination field; m represents the modulation degree; p represents the spatial frequency. Indicates the initial phase.

3. The method for reconstructing super-resolution microscopic images under structured light illumination according to claim 1, characterized in that, Step S2 is as follows: The light intensity is modulated by a light source module and a spatial light modulation module to generate two structured light illumination fields in two different spatial directions with light intensity satisfying a one-dimensional cosine function distribution. The sample is then illuminated sequentially by the two structured light illumination fields in different spatial directions to acquire two structured light illumination fluorescence images. The angle between the adjacent directions of the two different spatial directions is 90°, and the two different spatial directions are 0° and 90° respectively.

4. The method for reconstructing a super-resolution microscopic image under structured light illumination according to claim 1, characterized in that, Step S2 is as follows: The light intensity of the structured light illumination field in three different spatial directions is generated by modulating the light field with a light source module and a spatial light modulation module to satisfy a one-dimensional cosine function distribution. The sample is then illuminated sequentially by the structured light illumination field in the three different spatial directions to acquire three structured light illumination fluorescence images. The angle between adjacent directions of the three different spatial directions is 120°, and the three different spatial directions are 0°, 120° and 240° respectively.

5. The method for reconstructing super-resolution microscopic images under structured light illumination according to claim 1, characterized in that, In step S2, all N structured light illumination fluorescence images satisfy the following formula: ； In the formula, This represents the original 0-phase-shift structured illumination fluorescence image in the i-th direction, i=1,2…N; Indicates the fluorescence intensity of the sample; The point spread function (PSF) of an imaging system; Let represent the light intensity of the structured light illumination field along the i-th direction of the sample plane along the r direction, where i = 1, 2, ..., N; Indicates the concentration of labeled fluorescent molecules in the sample; Represents the convolution operation; where and , Proportional.

6. The method for reconstructing a super-resolution microscopic image under structured light illumination according to claim 1, characterized in that, Step S3 is as follows: S3.

1. Subtract the wide-field fluorescence image from N structured light illumination fluorescence images to obtain N difference images; S3.

2. Perform Hilbert transform on the N difference images respectively to generate structured light illumination images with phase shifts of 2π / 3 and 4π / 3, so that 3 structured light illumination images are obtained in each of the N spatial directions, for a total of 3N structured light illumination images; each of the 3 structured light illumination images in each spatial direction contains the original structured light illumination fluorescence image with 0 phase shift and the structured light illumination images with phase shifts of 2π / 3 and 4π / 3.

7. The method for reconstructing a super-resolution microscopic image under structured light illumination according to claim 6, characterized in that, In step S3.2, the expressions for the structured light illumination images with phase shifts of 2π / 3 and 4π / 3 are respectively: ； ； In the formula, Let i represent the structured light illumination image with a phase shift of 2π / 3 in the i-th direction, where i = 1, 2, ..., N; Let i represent the structured light illumination image with a phase shift of 4π / 3 in the i-th direction, where i = 1, 2, ..., N; This represents a wide-field fluorescence image, i.e., the fundamental frequency component; This represents the original 0-phase-shift structured illumination fluorescence image in the i-th direction, i=1,2…N; Indicates the in-plane edge of the sample The direction-modulated cosine image is subjected to Hilbert transform.

8. The method for reconstructing super-resolution microscopic images under structured light illumination according to claim 1, characterized in that, In step S4: The structured light illumination super-resolution image reconstruction algorithm employs at least one of the fair-SIM algorithm, HiFi-SIM algorithm, and Direct-SIM algorithm.

9. A structured light illumination super-resolution microscopic image reconstruction system, used to implement the structured light illumination super-resolution microscopic image reconstruction method according to any one of claims 1-8; characterized in that, include: The light source module is used to provide illumination light; The light field modulation module is coupled to the light source of the light source module and is used to modulate the illumination light provided by the light source module into a wide-field illumination light field with uniform light intensity distribution, and a structured light illumination light field with N different spatial directions whose light intensity satisfies a one-dimensional cosine function distribution, where N≥2; The microscopic imaging module is used to illuminate the sample with a wide-field illumination light field and a structured light illumination light field to achieve fluorescence imaging, obtaining one wide-field fluorescence image and N structured light illumination fluorescence images; The image acquisition module, electrically connected to the microscopic imaging module, is used to acquire one wide-field fluorescence image and N structured light illumination fluorescence images of the sample. The data processing module, electrically connected to the image acquisition module, is used to process the one wide-field fluorescence image and N structured light illumination fluorescence images to obtain the final super-resolution image.

10. A structured light illumination super-resolution microscopic image reconstruction system according to claim 9, characterized in that: The light field modulation module employs a liquid crystal spatial light modulator (SLM), a digital micromirror device (DMD), a one-dimensional physical grating, or a metasurface device.

Images