High-resolution non-destructive observation method for submicron organelles

Abstract

According to the method, a programmable LED lamp is installed on a microscope condenser, a light source is adjusted to the height of a sample table, then four inclined illumination modes including the upper, the lower, the left and the right are switched, part of coherent light is obliquely incident to a sample, at the moment, emergent light waves carry amplitude and phase information of the sample, and the sample can be observed in a high-resolution non-destructive mode. The sample is collected by the high magnification objective lens, the corresponding intensity image is collected by the CCD camera, and the collected original image is subjected to deconvolution by using the differential phase contrast imaging principle to obtain the biological information such as the thickness and the dry mass of the sample. A programmable annular LED is adopted as an illumination light source to enhance the phase contrast of a low-frequency space and reduce noise interference, and the illumination light source forms an angle of 15 degrees with a sample platform and directly irradiates a sample to generate a uniform illumination light source. The submicron organelle can be dynamically observed in real time through the high numerical aperture objective lens, and organelle edge extraction and 3D form reconstruction are obtained by using a deconvolution reconstruction algorithm.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of computational optical imaging and quantitative phase imaging, and particularly relates to a high-resolution non-invasive observation method for sub-micron organelles. BACKGROUND

[0002] With the increasing demand for microscopic imaging technology in the field of life sciences and biomedical sciences, it is particularly important to improve the imaging resolution to observe the organelle structures in cells, such as mitochondria and lipid droplets, for understanding the physiological state of cells and pathological mechanisms. However, the resolution of bright-field microscopy is limited by the diffraction of light waves and the numerical aperture of the objective lens, thereby limiting the imaging resolution. In recent years, super-resolution imaging technology has made significant breakthroughs in the field of biomedical research. By breaking the optical diffraction limit, super-resolution fluorescence imaging can be achieved using fluorescent molecules, which greatly promotes the development of biomedical sciences. Common super-resolution imaging techniques include stimulated emission depletion microscopy (STED), photoactivated localization microscopy (PALM), stochastic optical reconstruction microscopy (STORM), and structure illumination microscopy (SIM). However, these techniques sacrifice temporal resolution and have special requirements for samples and fluorescent probes. In addition, the photo-toxicity and photobleaching phenomena introduced by fluorescence imaging may cause damage to biological cells, which poses certain challenges and limitations in clinical applications.

[0003] In contrast, quantitative phase imaging methods can achieve high-contrast optical imaging without the need for specific dyes or fluorescent proteins. Off-axis digital holographic microscopy, as the most common interferometric quantitative phase imaging method, quantitatively measures the phase difference caused by the different refractive indices of samples through the interference of a coherent illumination light source. However, this method is easily affected by environmental disturbances and laser speckle, thereby reducing the imaging quality. Fortunately, the emergence of white-light holographic imaging technology has alleviated the impact of coherent noise and mechanical vibrations on imaging, thereby improving the imaging resolution and quality. However, these methods still rely on spatially coherent illumination, thereby limiting the imaging resolution within the range of coherent diffraction limit.

[0004] With the combination of LED lighting and quantitative phase imaging, due to the advantages of low cost, simple control and high luminous flux of LED light source, quantitative phase imaging ushered in a new development. Among them, quantitative differential phase contrast imaging uses LED lamp as the illumination light source, because the LED light source belongs to spatial partially coherent light, which significantly reduces the influence of coherent noise on imaging resolution, and as a non-interferometric quantitative phase imaging technology, it can improve the resistance of the system to mechanical vibration interference, and improve the imaging resolution to twice the coherent diffraction limit.

[0005] The quantitative differential phase contrast imaging method can realize high-resolution observation of biological samples, has potential application value and important significance, and explores the practical effect in clinical application and the potential application in other biomedical fields, improves the resolution of biological cell imaging, and promotes the development of biomedical research. SUMMARY

[0006] The technical problem to be solved by the present application is to provide a sub-micron organelle high-resolution non-invasive observation method, which can break through the resolution limit of traditional optical imaging methods and realize non-invasive, non-invasive, quantitative and high-resolution observation of biological samples.

[0007] To solve the above technical problems, one technical solution adopted by the present application is to provide a sub-micron organelle high-resolution non-invasive observation method, comprising the following steps:

[0008] Step one, using an ultra-inclined imaging light path including a programmable LED ring lamp, an infinite distance correction optical microscopic system for imaging and a CCD camera for image acquisition to image the sub-micron organelle, adjusting the installation position of the programmable LED ring lamp on the microscope to make the center of the microscope optical axis coincide with the center of the LED ring lamp, and ensuring that the LED generates uniform partially coherent intensity light;

[0009] Step two, selecting a high magnification objective lens, adjusting the programmable LED ring lamp illumination aperture angle to match the numerical aperture of the objective lens by changing the height of the illumination light source from the sample plane;

[0010] Step three, by controlling the programmable ring LED lamp to realize different illumination modes, four asymmetrically illuminated images are taken, the camera sequentially collects four semi-ring LED oblique illumination images, named 、 、 、 And save to PC end;

[0011] Step four, according to the principle of differential phase contrast imaging, the intensity of the transmitted light contains the phase information of the sample, that is, the optical path difference through the sample is changed, and the non-labeled high-contrast imaging of the weakly absorbing biological sample is carried out, the phase information of the sample is restored through deconvolution, and the thickness and dry mass of the sample are calculated according to the phase value, and the non-invasive and non-labeled high-contrast observation of sub-micron cells is completed.

[0012] In a preferred embodiment of the present application, the infinity-corrected optical microscope system comprises an objective lens and a lens, and the sample is located on the front focal plane of the objective lens, and the CCD camera is located in the back focal plane of the lens barrel lens.

[0013] In a preferred embodiment of the present application, the programmable ring LED lamp comprises a ring structure and a plurality of LEDs, the LEDs are installed in the ring structure in a uniform and inclined ring array, the incident angle of all the LEDs is determined by the height of the LED light source from the sample plane, so that each LED is inclined to directly enter the sample at an incident angle of 15°, and the hollow cone illumination is generated by the ring LED array.

[0014] In a preferred embodiment of the present application, the programmable ring LED lamp establishes a serial signal with an external PC end through a single-chip microcomputer, controls the on-off of the lamp, and switches to realize ring bright field, oblique illumination and quantitative differential phase contrast imaging.

[0015] In a preferred embodiment of the present application, in step two, the illumination numerical value of the programmable ring LED lamp is calculated by the following formula:

[0016]

[0017] Wherein The refractive index is constant, that is, 1, The illumination aperture angle is The illumination distance of the illumination light source from the sample plane is The radius of the programmable ring LED illumination array is

[0018] In a preferred embodiment of the present application, the specific steps of step four include:

[0019] S401: using the transfer cross coefficient theory, the transmission function of the sample is The intensity is linearly related to the absorption and phase by using the weak object traditional function:

[0020] (1)

[0021] Wherein, The transmission function of the sample is The absorption function of the sample is The phase delay caused by the sample is ;

[0022] (2)

[0023] (3)

[0024] At this point, the light intensity distribution in the frequency domain is linearly divided into three terms. For background items, For the amplitude term, For phase terms; where, Intensity distribution Fourier transform, Spatial frequency coordinates; Background intensity under uniform lighting; For Dirac function, The spatial frequency distribution function of the lighting source; For the pupil function of the imaging system, Its conjugate; and Absorption functions and phase delay Fourier transform; This is a variable representing the direction of illumination in the spatial frequency domain;

[0025] S402: In order to obtain the transfer function for quantitative phase imaging and thus generate a phase difference image along a specific phase gradient direction, a uniformly distributed, semi-circular incoherent light source is used.

[0026] (4)

[0027] In the frequency domain, the background and amplitude terms are eliminated by differentiating the numerator, retaining only the desired phase term. The phase term is eliminated by adding the denominators, retaining the amplitude term. At this point, the intensity image is further linearized and represented as:

[0028] (5)

[0029] in, and These represent intensity images captured under the upper and lower semi-circular LED lighting, respectively. For space Fourier transform operators, This represents the differential phase-contrast image in the frequency domain. and These represent the phase transfer functions corresponding to the upward and downward lighting directions, respectively; The phase transfer function for quantitative differential phase-contrast imaging;

[0030] S403: Quantitative phase information is obtained by re-inverting the sample intensity image, and the difference between the measured image and the image to be tested is minimized by differentiating the following formula:

[0031] (6)

[0032] Phase value is obtained from image intensity. The phase value is then calculated by deconvolving the target using the least squares method, and a penalty coefficient is added. To optimize noise in the signal and improve the quality of the inverted image:

[0033] (7)

[0034] in, express Norm, This is a regularization penalty coefficient used to suppress noise and stabilize the inversion process. This represents four DPC images with different lighting directions; in formula (7), For inverse Fourier transform operators, For the first DPC phase transfer function under each illumination direction The complex conjugate, This represents the total number of DPC images involved in the reconstruction. The beneficial effects of this invention are:

[0035] (1) The method described in this invention installs a programmable LED lamp on the microscope condenser lens, adjusts the height of the light source to the sample stage, and switches between four oblique illumination modes (up, down, left, right) so that some coherent light is obliquely incident on the sample. At this time, the outgoing light wave carries the amplitude and phase information of the sample, which is collected by a high magnification objective lens and the corresponding intensity image is acquired by a camera. The original image is deconvolved using the differential phase contrast imaging principle to obtain biological information such as the thickness and dry mass of the sample.

[0036] (2) Traditional commercial LED arrays cannot provide large-angle tilt illumination to meet the illumination numerical aperture matching conditions of high numerical aperture objectives, which is crucial for accurate phase recovery in quantitative phase imaging. In this invention, a programmable ring LED is used as the illumination source to enhance the phase contrast in the low-frequency space and reduce noise interference. It is positioned at a 15-degree angle to the sample platform to directly illuminate the sample, generating a uniform illumination source. The light source is easy to control and install. Submicron organelles can be dynamically observed in real time using a high numerical aperture objective, and the deconvolution reconstruction algorithm can be used to obtain organelle edge extraction and 3D morphology reconstruction.

[0037] (3) The imaging system has a lateral resolution of less than 200 nm and can achieve label-free, non-damaging, quantitative and high-resolution dynamic imaging of biological cells. Attached Figure Description

[0038] Figure 1 This is a schematic flowchart of the high-resolution non-destructive observation method for submicron organelles of the present invention;

[0039] Figure 2 This is a schematic diagram of the super-tilt imaging;

[0040] Figure 3 This is a schematic diagram of a 500nm polystyrene microsphere super-tilted imaging under a 60x objective lens;

[0041] Figure 4 This is a schematic diagram of human umbilical vein endothelial cell imaging under an ultra-tilted device;

[0042] Figure 5 This is a schematic diagram of an ultra-tilted image of human lung adenocarcinoma cells (a549). Detailed Implementation

[0043] The preferred embodiments of the present invention will now be described in detail with reference to the accompanying drawings, so that the advantages and features of the present invention can be more easily understood by those skilled in the art, thereby providing a clearer and more explicit definition of the scope of protection of the present invention.

[0044] Please see Figure 1 and Figure 2 The embodiments of the present invention include:

[0045] A method for high-resolution, non-destructive observation of submicron organelles includes the following steps:

[0046] Step 1: Imaging submicron organelles is performed using an ultra-tilted imaging optical path that includes a programmable LED ring lamp, an infinity-corrected optical microscope system for imaging, and a CCD camera for image acquisition. By adjusting the mounting position of the programmable LED ring lamp on the microscope, the center of the microscope optical axis coincides with the ring center of the LED ring lamp, ensuring that the LED produces uniform partially coherent intensity light.

[0047] In this example, Figure 2 (a) Schematic diagram of ring LED illumination; (b) Ring LED array: each LED illuminates the sample from a different azimuth angle; (c) Showing each LED illuminating the sample at a 15° incident angle; (d) Modified ultra-tilted quantitative differential phase-contrast microscope based on a commercial microscope (Olympus IX73). Combined with... Figure 2 (d) Imaging image: Based on the characteristics of quantitative differential phase-contrast imaging, an imaging optical path was designed, and a modified ultra-tilt quantitative differential phase-contrast microscope based on a commercial microscope (Olympus IX73) was built. The infinity-corrected optical microscopy system includes an objective lens and a lens. The sample is located on the front focal plane of the objective lens, and the CCD camera is located in the rear focal plane of the lens.

[0048] Step 2: In order to observe submicron organelles, a high-magnification objective lens is selected. By changing the height of the illumination source from the sample plane, the illumination aperture angle of the programmable LED ring lamp is adjusted to match the numerical aperture of the objective lens.

[0049] In this example, the objective lens used is Olympus (UPLSAPO60XW, 60x, 1.2NA). The illumination value can be calculated using the formula:

[0050] (1)

[0051] in The refractive index is 1, and since the environment is atmospheric, the refractive index is always 1. For the illumination aperture angle, For lighting distance, Let be the radius of the LED lighting array. As shown in the formula, by changing the height of the lighting source from the sample plane... To match the numerical aperture of the objective lens, the illumination aperture angle is adjusted. As the height decreases, the numerical aperture gradually increases, and because the medium is air with a refractive index of 1.00, the illumination numerical aperture is less than 1. When an Olympus immersion objective lens with 60x magnification and a numerical aperture of 1.2 is used in the imaging optical path, it consistently meets the requirements. That is, it meets the conditions for bright-field imaging.

[0052] Step 3: By controlling the programmable ring LED light to achieve different lighting modes, four asymmetrical lighting images are captured. The camera sequentially acquires four semi-circular LED oblique lighting images (top, bottom, left, and right), named... , , , And save it to the PC;

[0053] In this example, the programmable ring LED light includes a ring structure and 64 LEDs. The LEDs are uniformly and obliquely mounted in a ring array inside the ring structure. All LEDs are incident at a 15° angle, directly illuminating the sample, generating hollow cone illumination through the ring LED array. The LEDs produce high-angle oblique illumination from different azimuth angles, improving sample contrast and lateral resolution. The programmable ring LED light establishes a serial port signal with the PC via an Arduino microcontroller to control the light's on / off state. The microcontroller can output values ​​from 0 to 255 for each LED, sequentially increasing the ring LED brightness. LabVIEW is designed on the PC to control and switch between ring bright-field, oblique illumination, and quantitative differential phase-contrast imaging.

[0054] Using a spatially partially coherent LED illumination array as the illumination source significantly reduces the impact of coherence noise on imaging resolution. Furthermore, as a non-interferometric quantitative phase imaging technique, it enhances the system's resistance to mechanical vibration interference, increasing the imaging resolution to twice the coherence diffraction limit. After acquiring four ultra-tilted illumination images (top, bottom, left, and right), the phase information of the sample is obtained by measuring the change in optical path difference caused by the LED light passing through the sample. A reconstruction algorithm is used to improve image quality, and the three-dimensional morphology and dry mass of the sample are simultaneously retrieved.

[0055] Step four: Based on the principle of differential phase-contrast imaging, transmitted light contains both intensity and phase information of the sample. This means that the optical path difference changes as the light passes through the sample. Label-free, high-contrast imaging is then performed on weakly absorbing biological samples. The sample's phase information is recovered through deconvolution, and the sample thickness and dry mass are calculated based on the phase value. By achieving quantitative differential phase-contrast imaging with a high-magnification objective lens numerical aperture, the lateral resolution is improved, enabling non-invasive, label-free, high-contrast observation of submicron cells.

[0056] Spatial resolution refers to the minimum distance between two independent features that can be distinguished after passing through an optical imaging system. When the spatial resolution constraint is smaller than the pre-observed feature, the imaging resolution is sufficient. However, the spatial resolution of an optical imaging system is limited by light diffraction and the numerical aperture of the imaging system. Theoretically, increasing the numerical aperture of the imaging system and decreasing the wavelength λ of light are simple and effective methods. This can be achieved by using the formula... The definition determines the range of spatial frequency information transmitted by the objective lens, among which... Let be the refractive index of the medium. It is half the objective aperture angle. Due to limitations in refractive index and objective aperture angle, the resolution limit of a conventional optical microscope is... for:

[0057] (2)

[0058] in, Indicates the numerical aperture of the objective lens. Indicates the wavelength of the incident light.

[0059] In the spatial frequency domain, the minimum resolution spot size of an image is typically related to the system's impulse response, which defines the system's frequency response, i.e., the bandwidth transfer function. The transfer function describes the system's response to different spatial frequency components, thus affecting the system's image resolution. Under coherent illumination, the numerical aperture of the objective lens and the wavelength of light determine the coherent transfer function (CTF). In this case, the transfer function is generally circular with a unit amplitude. This means that the CTF has a stronger spot response to certain spatial frequencies and a weaker response to others. This characteristic of the CTF affects the system's image resolution. The CTF cutoff frequency is defined as... Under incoherent illumination, the system's imaging quality depends only on the light field intensity. In this case, the minimum resolution is no longer limited by the CTF, and the system's cutoff frequency can be defined as... .

[0060] Diffraction-limited phase imaging (DPC) is a quantitative phase imaging technique that uses a partially coherent light source. Using asymmetric illumination, the spectrum of the sample is shifted in the Fourier plane by adjusting the angle and position of the light source to reveal the phase information of thin samples. In diffraction-limited systems, the maximum spatial frequency of a DPC system corresponds to twice the numerical aperture of the objective lens. Its minimum resolution limit This can be expressed by the following formula:

[0061] (3)

[0062] in This represents the numerical aperture of the illumination source, which is modulated by parameters such as the shape, position, and angle of the illumination source. Therefore, using high numerical aperture objectives and ring LED illumination sources in DPC imaging can significantly improve imaging resolution, achieving resolutions below 200 nm.

[0063] The specific imaging principle of differential phase contrast imaging is as follows:

[0064] Quantitative differential phase-contrast imaging (DPC) is a partially coherent quantitative phase imaging technique that uses asymmetric illumination to shift the sample spectrum in Fourier space, revealing the object's phase information. Since LED light sources are spatially partially coherent, the forward model of this system is nonlinear, making its mathematical model difficult to establish and solve. To simplify the solution process for partially coherent systems, a weak target approximation is typically used. This approximates the sample as thin and transparent, neglecting higher-order nonlinear terms in the complex-field Taylor expansion to linearize the intensity image in the forward model of DPC. However, phase and intensity are difficult to separate through direct inversion; therefore, the transfer cross-coefficient theory is employed. In this case, the sample's transmission function... :

[0065] (4)

[0066] The absorption function of the light source after passing through the sample is: Phase delay is The coordinates of the sample in the spatial domain are Using the Weak Object Traditional Function (WOTF), a linear correlation between intensity and absorption and phase can be derived:

[0067] (5)

[0068] (6)

[0069] At this point, the light intensity distribution in the frequency domain is linearly divided into three terms. For background items, Amplitude Transfer Function (ATF) and For phase transfer function (PTF), For the amplitude term, This is the phase term. Intensity distribution Fourier transform, Spatial frequency coordinates; Background intensity under uniform lighting; For Dirac function, The spatial frequency distribution function of the lighting source; For the pupil function of the imaging system, Its conjugate; and Absorption functions and phase delay Fourier transform; This is a variable representing the direction of illumination in the spatial frequency domain.

[0070] (7)

[0071] To obtain the transfer function of the DPC and thus generate a phase difference image along a specific phase gradient direction, a uniformly distributed, semi-circular incoherent light source is used. In an aberration-free system, since the illumination images are symmetrical along the same axial direction (e.g., DPC images acquired vertically), and due to the phase transfer function... As this is an antisymmetric function, in the frequency domain, the numerator is differentiated to eliminate the background and amplitude terms, retaining only the desired phase term. The denominators are then summed to eliminate the phase term, retaining only the amplitude term. The intensity image can then be further linearized as follows:

[0072] (8)

[0073] in, and These represent intensity images captured under the upper and lower semi-circular LED lighting, respectively. For space Fourier transform operators, This represents the differential phase-contrast image in the frequency domain. and These represent the phase transfer functions corresponding to the upward and downward lighting directions, respectively; This is the phase transfer function for quantitative differential phase-contrast imaging.

[0074] At this point, the phase information of the sample is obtained, and its quantitative phase information can be obtained by inversion from the sample intensity image.

[0075] The inverse problem of quantitative differential phase-contrast imaging aims to minimize the difference between the image being measured and the image under test.

[0076] (9)

[0077] At this point, equation (9) is differentiable and can be directly differentiated, and the phase value can be obtained from the image intensity. Generally, the least squares method is used to deconvolve the target and then solve for the phase value, with a penalty coefficient added. This is to optimize noise in the signal and improve the quality of the inverted image.

[0078] (10)

[0079] in, express Norm, This is a regularization penalty coefficient used to suppress noise and stabilize the inversion process. This represents four DPC images with different lighting directions. In formula (10), For inverse Fourier transform operators, For the first DPC phase transfer function under each illumination direction The complex conjugate, The total number of DPC images involved in the reconstruction.

[0080] Similarly, the phase values ​​of the DPC images acquired from the left and right sides are obtained using the same method, which will not be elaborated here.

[0081] The effects of the method described in this invention will be illustrated below through examples.

[0082] Polystyrene microspheres are widely used in microscopic imaging. Their standard size provides important reference value for quantitative phase imaging analysis.

[0083] like Figure 3 As shown in (a), the figure presents the original intensity images of super-tilted illumination obtained from the four semi-ring illumination directions (top, bottom, left, and right) using a programmable ring LED array. In these images, the polystyrene microspheres generate different phase gradient information under different illumination directions, enabling the extraction of their phase changes and improvement of lateral resolution in subsequent differential operations. Figure 3 (b) Demonstrates that super-tilted illumination breaks the optical diffraction limit, extending the PTF spectrum to twice that of coherent illumination. Figure 3Images (c1)-(c5) show the results of processing 500nm polystyrene microspheres using different qDPC reconstruction algorithms. The PD-qDPC algorithm (Pupil-Driven Quantitative Differential Phase Contrast) improves the quality of qDPC phase reconstruction by using a modified L0 norm to represent the edge sparsity of the pupil drive and achieving automatic background removal by replicating the qDPC convolution operator in the data fidelity term. An iterative weighted soft thresholding algorithm based on the split Bregman method was developed to address the modified L0 norm problem. Figure 3 As shown in (d1)-(d5), pd-qDPC outperforms other state-of-the-art qDPC methods in terms of phase reconstruction quality and implementation efficiency. This algorithm achieves high-quality qDPC reconstruction without any modifications to the optical system and simplifies system complexity. Figure 3 As shown in (c6), the pd-qDPC algorithm extracts the edges of 500nm polystyrene microspheres, indicating that the algorithm can be used for pattern recognition tasks such as cell segmentation and PTF learning based on edge filtering characteristics. Figure 3 (d6) shows Figure 3 Results of the region method in (c6).

[0084] Figure 4 To conduct biological cell imaging experiments using a super-tilted quantitative differential phase-contrast imaging device, human umbilical vein endothelial cells (HUVECs) were used as the sample. Figure 4 As shown in (a1)-(a2), due to the weak absorption characteristics of the sample, the low contrast under bright-field imaging makes it difficult to observe the cell nucleus and the structural information of the cells within it. The contrast can be significantly improved by using an ultra-tilted device, such as... Figure 4 As shown in (b1)-(b2), spherical lipid droplets, resembling organelles, were observed in HUVEC cells. Figure 4 (c1)-(c2) show the distribution of lipid droplets around the cell nucleus. Lipid droplets are usually detected by staining. By using super-tilt imaging, subcellular organelles and the cell nucleus can be observed, providing a new approach for real-time, non-invasive quantitative observation of cells.

[0085] Figure 5Building upon the above, a super-tilt imaging device was used to dynamically observe cellular lipid droplets. The sample used was human lung adenocarcinoma cells (a549). Figure (a) shows the cell morphology under 60x objective Kohler illumination, and (b) shows the cell morphology under super-tilt quantitative differential phase-contrast imaging, where organelles are clearly visible. By setting the acquisition program, dynamic observation of the cells was possible. The migration of lipid droplets between two cells was observed. Figure 5 (d) shows the extraction of cell edges using the pd-qDPC algorithm. Figure 5 (d1)-(d3) represent cell phase images obtained at different imaging times, showing that the spatial position of lipid droplets within the cell changes over time.

[0086] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for high-resolution, non-destructive observation of submicron organelles, characterized in that, Includes the following steps: Step 1: Imaging submicron organelles is performed using an ultra-tilted imaging optical path that includes a programmable LED ring lamp, an infinity-corrected optical microscope system for imaging, and a CCD camera for image acquisition. By adjusting the mounting position of the programmable LED ring lamp on the microscope, the center of the microscope optical axis coincides with the ring center of the LED ring lamp, ensuring that the LED produces uniform partially coherent intensity light. Step 2: Select a high-magnification objective lens and adjust the programmable LED ring lamp illumination aperture angle to match the numerical aperture of the objective lens by changing the height of the illumination source from the sample plane. Step 3: By controlling the programmable ring LED light to achieve different lighting modes, four asymmetrical lighting images are captured. The camera sequentially acquires four semi-circular LED oblique lighting images (top, bottom, left, and right), named... , , , And save it to the PC; Step four: Based on the principle of differential phase contrast imaging, the transmitted light contains the intensity and phase information of the sample. That is, the optical path difference changes as it passes through the sample. Label-free high-contrast imaging is performed on weakly absorbing biological samples. The phase information of the sample is recovered by deconvolution. The sample thickness and dry mass are calculated based on its phase value, thus completing the non-invasive label-free high-contrast observation of submicron cells.

2. The method for high-resolution, non-destructive observation of submicron organelles according to claim 1, characterized in that, The infinity-corrected optical microscopy system includes an objective lens and a lens, with the sample located on the front focal plane of the objective lens and the CCD camera located in the rear focal plane of the lens barrel.

3. The method for high-resolution, non-destructive observation of submicron organelles according to claim 1, characterized in that, The programmable ring LED lamp includes a ring structure and several LEDs. The LEDs are uniformly and obliquely installed in a ring array inside the ring structure. The incident angle of all LEDs is determined by the height of the LED light source from the sample plane, so that each LED is obliquely and directly incident into the sample at an incident angle of 15°, and hollow cone illumination is generated through the ring LED array.

4. The method for high-resolution, non-destructive observation of submicron organelles according to claim 1, characterized in that, The programmable ring LED light establishes a serial port signal with an external PC through a microcontroller to control the light's on / off state, thereby switching between ring bright field, oblique illumination, and quantitative differential phase contrast imaging.

5. The method for high-resolution, non-destructive observation of submicron organelles according to claim 1, characterized in that, In step two, the illumination value of the programmable ring LED is calculated using the following formula: , in The refractive index is always 1. For the illumination aperture angle, The illumination distance between the light source and the sample plane. The radius is the programmable ring LED lighting array.

6. The method for high-resolution, non-destructive observation of submicron organelles according to claim 1, characterized in that, Step four includes the following specific steps: S401: Using the transmission cross-coefficient theory, the sample's transmission function is... Using traditional functions for weak objects, a linear correlation between intensity, absorption, and phase is derived: (1) in, Let be the transmission function of the sample. Let be the absorption function of the sample. The phase delay caused by the sample is given by the coordinates of the sample in the spatial domain. ; (2) （3） At this point, the light intensity distribution in the frequency domain is linearly divided into three terms. For background items, For the amplitude term, For phase terms; where, Intensity distribution Fourier transform, Spatial frequency coordinates; Background intensity under uniform lighting; For Dirac function, The spatial frequency distribution function of the lighting source; For the pupil function of the imaging system, Its conjugate; and Absorption functions and phase delay Fourier transform; This is a variable representing the direction of illumination in the spatial frequency domain; S402: In order to obtain the transfer function for quantitative phase imaging and thus generate a phase difference image along a specific phase gradient direction, a uniformly distributed, semi-circular incoherent light source is used. （4） In the frequency domain, the background and amplitude terms are eliminated by differentiating the numerator, retaining only the desired phase term. The phase term is eliminated by adding the denominators, retaining the amplitude term. At this point, the intensity image is further linearized and represented as: （5） in, and These represent intensity images captured under the upper and lower semi-circular LED lighting, respectively. For space Fourier transform operators, This represents the differential phase-contrast image in the frequency domain. and These represent the phase transfer functions corresponding to the upward and downward lighting directions, respectively; The phase transfer function for quantitative differential phase-contrast imaging; S403: Quantitative phase information is obtained by re-inverting the sample intensity image, and the difference between the measured image and the image to be tested is minimized by differentiating the following formula: （6） Phase value is obtained from image intensity. The phase value is then calculated by deconvolving the target using the least squares method, and a penalty coefficient is added. To optimize noise in the signal and improve the quality of the inverted image: （7） in, express Norm, This is a regularization penalty coefficient used to suppress noise and stabilize the inversion process. This represents four DPC images with different lighting directions; in formula (7), For inverse Fourier transform operators, For the first DPC phase transfer function under each illumination direction The complex conjugate, The total number of DPC images involved in the reconstruction.

Images