A quantitative phase microscopy method for large phase objects

Abstract

This invention discloses a quantitative phase microscopy imaging method for large-phase objects. It only requires acquiring phase gradient images under asymmetric illumination to achieve high-precision quantitative phase imaging of large-phase objects. Therefore, differential phase-contrast quantitative phase imaging technology based on asymmetric illumination is no longer limited by weak object approximation and can be applied to high-precision phase characterization and measurement of large-phase samples.

Description

Technical Field

[0001] This invention belongs to the field of optical microscopy and quantitative phase imaging technology, specifically a quantitative phase microscopy imaging method for large phase objects. Background Technology

[0002] Differential phase contrast imaging (DPI), as a partially coherent imaging technique, establishes a display relationship between the observed sample and the acquired intensity by modulating the spatial coherence transfer function response. Based on the phase modulation effect of asymmetric illumination, it uses transfer function theory to establish a linear relationship between intensity and phase, thereby achieving demodulation and restoration of the light field information. In 2015, Tian et al. gave an expression for the phase transfer function of differential phase contrast under asymmetric illumination (Tian L, Waller L. Quantitative differential phase contrast imaging in an LED array microscope[J]. OpticsExpress,2015,23(9):11394.), and for the first time realized quantitative phase imaging based on an LED-based differential phase contrast imaging system. This method uses asymmetric illumination and treats the process of recovering (estimating) the phase through intensity distribution as an inverse mathematical problem, recovering the quantitative phase of the sample through a one-step deconvolution.

[0003] As a partially coherent imaging technique, the derivation of the transfer function in differential phase-contrast imaging often requires the introduction of necessary approximations to linearize the light intensity representation and sample phase distribution. However, while these approximations simplify the imaging model, they also impose certain limitations on the algorithm's applicability. This means that the reconstruction accuracy and imaging performance of the phase retrieval algorithm will be limited by the degree of matching between the actual imaging model and the approximation conditions.

[0004] In quantitative phase imaging, the introduction of weak phase approximation effectively separates the modulation effects of the imaging system on different components (absorption and phase) of the sample, establishing amplitude and absorption transfer functions independent of the sample. Based on this, a phase reconstruction algorithm based on deconvolution has been further developed. Under this approximation, the sample phase term needs to be much smaller than 1 rad (to achieve the Taylor expansion of the exponential function) so that the complex transmittance function of the sample (analyzed using a pure phase object as an example) can be approximated as a linear combination of the DC component and the phase term. Summary of the Invention

[0005] To address the aforementioned technical deficiencies in the prior art, this invention proposes a quantitative phase microscopy imaging method for large-phase objects.

[0006] The technical solution to achieve the objective of this invention is as follows: a quantitative phase microscopy imaging method for large-phase objects, the imaging process steps of which are as follows:

[0007] Step 1: Use a serial port to control a high-density programmable LED array, TFT-LCD, or spatial light modulator to generate a semi-circular illumination source and acquire intensity images of large-phase samples.

[0008] Step 2: Perform differential calculation on the acquired intensity image and deconvolve it with the phase transfer function of the optical system to obtain the quantitative phase distribution of the sample as the initial phase value.

[0009] Step 3: Construct a complex transmission function using initial phase value and uniform amplitude, and generate intensity difference images under asymmetric illumination using the Abbe method. At the same time, generate an intensity distribution linearly related to the phase using a weak phase transfer function. Subtract the two intensity distributions to obtain the estimated nonlinear intensity error distribution.

[0010] Step 4: Subtract the estimated nonlinear intensity error from the actual acquired intensity to obtain the linear intensity signal required for deconvolution;

[0011] Step 5: Use the linear intensity estimated in Step 4 to perform deconvolution reconstruction to recover the updated phase signal of the current iteration;

[0012] Step 6: Repeat steps 3 to 5 until the difference between the intensity signal of the acquired intensity differential image under asymmetric illumination generated by the Abbe method and the actual acquired intensity signal is lower than the set value, at which point the iteration terminates.

[0013] Step 7: Perform an inverse Fourier transform on the phase spectrum obtained through iteration to obtain the quantitative phase result of the recovered large-phase object.

[0014] Preferably, the acquired intensity image is represented as follows:

[0015] I(u)=Bδ(u)+A(u)ATF(u)+iΦ(u)WPTF(u)

[0016] In the formula, u represents the frequency component in the frequency domain, and B represents the background term, expressed as B=∫∫S(u j )|P(u j )| 2 d 2 u j u j S(u) represents the phase shift of a single angular illumination source in the frequency domain. j P(u) represents the tilted illumination light at a single angle, and P(u) represents the pupil function of the objective lens in the Fourier plane. j The phase shift u represents the phase shift of a single angular illumination source in the frequency domain. jLet A(u) be the pupil function of the objective lens on the Fourier plane, A(u) be the amplitude spectrum distribution of the sample, Φ(u) be the phase spectrum distribution of the sample, and ATF(u) be the amplitude transfer function, expressed as ATF(u)=∫∫S(u) j )P(u j )[P(u+u j )+P * (uu j )]d 2 u j , P(u+u j ) and P(uu j This indicates a phase shift of u in the frequency domain. j and -u j The objective lens has a pupil function in the Fourier plane, WPTF(u) represents the phase transfer function, and δ(u) represents the impulse function in the frequency domain.

[0017] Preferably, the specific steps for performing differential calculations on the acquired intensity image and deconvolving it with the phase transfer function of the optical system to obtain the quantitative phase distribution of the sample as the initial phase value are as follows:

[0018] The acquired intensity images are differentially analyzed to obtain the intensity signal after differential calculation.

[0019] The intensity signal after differential calculation The phase transfer function of the optical system is deconvolved with the phase transfer function of the optical system to obtain the solution.

[0020]

[0021] In the formula, u represents the frequency component in the frequency domain, u j S represents the phase shift of a single angular illumination source in the frequency domain. lr (u j P(u) represents the tilted illumination light under differential illumination. j The phase shift u represents the phase shift of a single angular illumination source in the frequency domain. j The pupil function of the objective lens in the Fourier plane, P(u+u) j ) and P(uu j This indicates a phase shift of u in the frequency domain. j and -u j The pupil function of the objective lens on the Fourier plane.

[0022] The quantitative phase distribution of the sample obtained by deconvolution is as follows:

[0023]

[0024] In the formula, WPTF iDPC (u) represents the phase transfer function, I i DPC (u) represents the light intensity distribution, α represents the regularization parameter, and i represents the asymmetry axis of the illumination.

[0025] Preferably, a complex transmission function is constructed using an initial phase value and a uniform amplitude. First, the Abbe method is used to generate an intensity difference image under asymmetric illumination. Then, a weak phase transfer function is used to generate an intensity distribution linearly related to the phase. Finally, the two intensity distributions are subtracted to obtain the specific formula for the nonlinear intensity error distribution:

[0026]

[0027] In the formula, S(u j ) represents the lighting source, where u j e represents the spatial frequency corresponding to a single illumination angle. jφ(x) Let P(u) be the complex transmittance function of the object, P(u) be the pupil function, WPTF(uj) be the phase transfer function of the weak object, φ(u) be the phase distribution obtained in step two, and φ(x) represent the phase distribution of the sample.

[0028] Preferably, the linear intensity signal required for deconvolution is specifically:

[0029]

[0030] In the formula, This represents the estimated linear intensity signal. This represents the intensity image after intensity difference analysis. As a nonlinear intensity error, this calculation can make the intensity signal approximate the ideal input of the deconvolution reconstruction algorithm.

[0031] Preferably, the specific formula for deconvolution reconstruction using the linear intensity estimated in step four is as follows:

[0032]

[0033] In the formula, WPTF i (u) represents the phase transfer function. Let represent the linear intensity signal, α represent the regularization factor, n represent the number of iterations, and i represent the asymmetry axis of the illumination.

[0034] Compared with the prior art, the significant advantages of this invention are as follows: This invention can relax the applicability conditions of weak object approximation of deconvolution solvers, making it applicable to any large phase sample; the iterative algorithm of this invention does not require additional data acquisition, and accurate phase reconstruction can be achieved through continuous deconvolution solvers; this invention is no longer limited by weak object approximation and can be applied to high-precision phase characterization and measurement of large phase samples.

[0035] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings. Attached Figure Description

[0036] The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Throughout the drawings, the same reference numerals denote the same parts.

[0037] Figure 1 This is a flowchart of an algorithm for a quantitative phase microscopy imaging method for large-phase objects.

[0038] Figure 2 This is a schematic diagram of the morphological examination of human breast cancer MCF-4 cells. Detailed Implementation

[0039] It is readily understood that, based on the technical solution of this invention, various embodiments of the invention can be conceived by those skilled in the art without altering the essential spirit of the invention. Therefore, the following detailed embodiments and accompanying drawings are merely illustrative examples of the technical solution of this invention and should not be considered as the entirety of the invention or as limitations or restrictions on the technical solution of this invention. Rather, these embodiments are provided to enable those skilled in the art to gain a more thorough understanding of the invention. Preferred embodiments of the invention are described below in conjunction with the accompanying drawings, which form part of this application and, together with the embodiments of the invention, serve to illustrate the innovative concept of the invention.

[0040] This invention presents a quantitative phase microscopy imaging method for large-phase objects. It acquires intensity information from a sample by modulating a programmable LED array, then reconstructs this intensity information using deconvolution to obtain the quantitative phase distribution of the sample. An estimated intensity difference image is then generated by combining the quantitative phase with a uniform amplitude. Simultaneously, the estimated phase is convolved with a weak phase transfer function to obtain an intensity distribution linearly related to the phase. The estimated nonlinear intensity error is then subtracted from the actual acquired intensity to obtain a linear intensity signal, which is then deconvolved to reconstruct the updated phase signal. The difference between the estimated intensity and the acquired image is used to determine whether to continue iterating to reduce this difference, ultimately obtaining a precise quantitative phase distribution of the sample. This method frees asymmetric illumination-based differential phase-contrast quantitative phase imaging technique from the limitations of weak object approximation, enabling high-precision phase characterization and measurement of large-phase samples.

[0041] A quantitative phase microscopy imaging method for large-phase objects, the specific steps of which are as follows:

[0042] Step 1: Acquisition of sample intensity under differential phase contrast imaging with asymmetric illumination modulation: A computer is used to control a high-density programmable LED array, TFT-LCD or spatial light modulator via serial port to display four semi-circular illumination sources in orthogonal directions, and four intensity images of the large-phase object are acquired.

[0043] Step 2, Quantitative Phase Initialization: The intensity information map obtained in the previous step is placed in the forward model of partial coherence imaging for processing. Linearization is performed according to the weak phase approximation condition to separate the sample amplitude and phase. Then, the phase transfer function is calculated based on the system parameters (including the illumination function and the system pupil function) and deconvolved with the obtained phase gradient spectrum to obtain the quantitative phase distribution of the sample, which is used as the initial phase value for the subsequent iteration operation.

[0044] This section details the phase deconvolution algorithm based on weak phase. First, we analyze the imaging algorithm model under single-angle illumination. The complex transmittance function under single-angle oblique illumination is t(x) = e -α(x)+iφ(x) For a thin, pure-phase sample, the tilted illumination light at a single angle is represented as S(u j ), where x=(x,y) represents the coordinates of the sample plane, and α(x) and φ(x) represent the intensity and phase of the sample, respectively. To separate the sample phase and intensity to simplify the problem, a weak phase approximation is used for the complex amplitude distribution of the sample, resulting in t(x)≈1-α(x)+iφ(x). Therefore, the intensity spectrum distribution on the camera plane is:

[0045] I j (u)=S(u j )δ(u)|P(u)| 2 -S(u j )A(u)[P * (u j )P(u+u j )+P(u j )P * (u+u j )]+iS(u j )Φ(u)[P * (u j )P(u+u j )-P(u j )P * (uu j (1)

[0046] Here, u represents the frequency component in the frequency domain. jThis represents the phase shift of a single angular illumination source in the frequency domain. P(u) represents the pupil function of the objective lens in the Fourier plane (assuming P(u) is a cutoff frequency of NA). obj An ideal low-pass filter ( / λ) allows only frequency components within the objective lens's numerical aperture to pass through. obj (Represents the numerical aperture of the objective lens). The intensity distribution in the frequency domain under partially coherent illumination is calculated. Ignoring higher-order convolution terms in the calculation process, the intensity spectrum distribution of the sample can be divided into three terms, expressed as follows:

[0047] I(u)=Bδ(u)+A(u)ATF(u)+iΦ(u)WPTF(u) (2)

[0048] B represents the background term, expressed as B=∫∫S(u j )|P(u j )| 2 d 2 u j A(u) represents the amplitude spectrum distribution of the sample, Φ(u) represents the phase spectrum distribution of the sample, ATF(u) represents the amplitude transfer function, and WPTF(u) represents the phase transfer function.

[0049] ATF(u)=∫∫S(u j )P(u j )[P(u+u j )+P * (uu j )]d 2 u j

[0050] WPTF(u)=∫∫S(u j )P(u j )[P(u+u j )-P * (uu j )]d 2 u j (3)

[0051] Equation (3) is a general expression for the amplitude transfer function and phase transfer function under partially coherent illumination, applicable to all illumination distributions S(u) and pupil functions P(u). Under asymmetric illumination, the spatial frequencies corresponding to the illumination angles are symmetrically distributed in the spectrum, i.e., u j and -u j This will result in a phase transfer function with an opposite distribution.

[0052] In typical differential phase-contrast imaging systems, the illumination pattern is often designed as asymmetrical, uniform semi-circular illumination. Along each shear direction, two illuminations sequentially illuminate the sample, acquiring two phase-contrast images. A simple differential calculation process is used to obtain the phase gradient image of the sample along that shear direction. Accordingly, based on this calculation model, the phase transfer function expression for differential phase-contrast imaging is given as:

[0053]

[0054] This explanation uses the left-right shearing direction as an example. In the above formula, S... lr (u j ) represents the intensity distribution of the lighting source in the left-right direction, WPTF lr (u) represents the phase transfer function along the left and right axes. The formula reveals that the phase transfer function is jointly determined by the illumination function and the pupil function, while the pupil function is determined by the wavelengths of the objective lens and illumination. In other words, given a fixed system parameter (P(u) is determined), the phase transfer function of the differential phase-contrast imaging system is determined by the illumination function. Therefore, it can be concluded that optimizing the illumination function through illumination spatial coherence modulation is the most direct and effective way to optimize the performance of differential phase-contrast imaging.

[0055] The phase transfer function establishes the connection between the differential phase-contrast image and the quantitative phase of the sample, characterizing how the quantitative phase of the sample is converted into an intensity image. Based on this principle, the quantitative phase of the sample can be calculated through a one-step deconvolution. The most common method is to use the Tikhonov criterion to deconvolve the spectrum of the phase gradient image and the phase transfer function to obtain the quantitative phase result of the sample. The calculation process is as follows:

[0056]

[0057] In the above formula, α represents the regularization parameter, which is usually set to 0.01 to avoid excessive amplification of noise.

[0058] Step 3, Nonlinear Intensity Error Estimation: A complex transmission function is constructed using the phase and uniform amplitude obtained in Step 2, generating an estimated complex amplitude distribution. Then, the Abbe method is used to generate an intensity difference image acquired under asymmetric illumination. This image, obtained under partially coherent imaging, is represented as a linear superposition of the intensities obtained from independently illuminating an object with a complex distribution using multiple point light sources. Here's a detailed explanation of the Abbe method:

[0059] The intensity of light follows Abbe diffraction theory, meaning that the mutual incoherence of illumination points leads to an incoherent superposition of intensities induced by each point. The cross-propagation and superposition at each point typically result in a nonlinear dependence of the partially coherent field on the properties of the sample.

[0060]

[0061] In the formula, x represents the two-dimensional (2D) coordinates of the camera plane, and u represents the spectral coordinates. The sample is expressed in the Fourier spectral form of its complex transmission function: (where φ(x) is the phase of the sample), T(u1)T * (u2) is its cross spectrum. S(u) represents a partially coherent illumination source, which is typically an extended source where the illumination NA is less than or equal to the objective lens NA. H(u) represents the coherence transfer function (CTF), which can be expressed as... (P(u) is the pupil function of the objective lens, while) It is the defocusing factor of the light field propagating on different planes. The ideal intensity distribution of the image is calculated using equation (6), which is the Abbe method.

[0062] Next, the phase calculated in step two is convolved with the weak phase transfer function (WPTF) to obtain the intensity distribution linearly related to the phase. Nonlinear intensity error. The intensity residual between the forward generation intensity and the phase linear correlation intensity of partially coherent imaging is obtained by calculating:

[0063]

[0064] Here S(u) j ) represents the lighting source, where uj is the spatial frequency corresponding to a single lighting angle, and e jφ(x) Let P(u) be the complex transmittance function of the object, and P(u) be the pupil function (which is a cutoff frequency of NA). obj / λ is a circular distribution, NA obj λ is the numerical aperture of the objective lens, and λ is the illumination wavelength. WPTF(u j The phase transfer function of a weak object is given by φ(u), which is the phase distribution obtained in the previous step. Formula (6) gives the expression for estimating the nonlinear error, where the theoretical intensity is obtained by the classical image forward generation model based on Abbe theory, and the intensity linearly related to the phase is generated by the transfer function theory.

[0065] This step estimates the nonlinear error in the intensity by using the intensity generated by the Abbe method and the linear intensity generated by the weak object phase transfer function. The intensity generated by the Abbe method matches the actual acquired intensity, while the intensity generated by the weak object transfer function convolution is the ideal intensity obtained from deconvolution. Subtracting the two yields the difference between the intensity required by the actual acquired image and the intensity obtained from the deconvolution solution.

[0066] Step 4, Linear Intensity Estimation: Subtract the estimated nonlinear intensity error from the actual acquired intensity. This eliminates the nonlinear intensity signal affecting the deconvolution algorithm, leaving only the linear intensity signal required for deconvolution.

[0067]

[0068] Here This represents the estimated linear intensity signal. This represents the intensity image after intensity difference is obtained from the actual acquisition.

[0069] This step approximates the ideal input of the deconvolution reconstruction algorithm by subtracting the nonlinear errors estimated by the two intensity generation models from the actual acquired intensity.

[0070] Step 5, Linear Intensity Deconvolution Reconstruction: Deconvolution reconstruction is performed using the estimated linear intensities obtained in the previous step to recover the updated quantitative phase distribution of the current iteration.

[0071] During the repetition of steps three through five, the nonlinear intensity error will be continuously corrected, which will improve the estimated linear intensity. Near-ideal linear intensity distribution The quantitative phase imaging solution is formulated as the following optimization problem:

[0072]

[0073] Then, the phase of the sample is updated by deconvolution iterative reconstruction based on the following formula (9). The deconvolution iterative reconstruction process can be expressed as:

[0074]

[0075] Here, α represents the regularization factor, used to eliminate ill-conditioned inversion and avoid excessive noise amplification caused by division by zero. In the formula, n represents the number of iterations, and i represents the asymmetry axis of the illumination.

[0076] The intensity signal used for deconvolution reconstruction eliminates the nonlinear error of model mismatch, thereby enabling the deconvolution reconstruction algorithm to be used for accurate quantitative phase reconstruction of large-phase objects.

[0077] Step Six, Iterative Convergence Condition Judgment: Compare the estimated intensity generated by the reconstructed phase using the Abbe method with the directly acquired image to obtain the difference value and determine its magnitude. If the difference is large, repeat the iterative process from Step Two to Step Six; if the difference is so small as to be negligible, it indicates that the iteration has converged to a stable phase value, and the iteration can be terminated to proceed to Step Seven.

[0078] The Abbe method uses phase values ​​that are as close as possible to the true value of the object to estimate the error between the two intensity models during iteration. Ultimately, the intensity generated by the Abbe method will be consistent with the actual acquired intensity. At this point, it can be considered that the phase of the object has been recovered as much as possible.

[0079] Step 7: Output Results: After the above iterative processing, the nonlinear intensity error caused by model mismatch was successfully corrected, thus obtaining the accurate quantitative phase distribution of the sample.

[0080] After completing the large-phase deconvolution reconstruction of the sample phase, we used human breast cancer cells (MCF-4) as the observation sample to compare the cell morphology detection performance of the traditional one-step deconvolution differential phase-contrast imaging method and the large-phase deconvolution method. During sample preparation, the breast cancer cells were not subjected to any staining or labeling treatment, were immersed in paraformaldehyde liquid, and mounted with nail polish. Figure 2 The reconstructed phases under the two methods are shown and compared.

[0081] In the one-step deconvolution algorithm, to avoid ill-conditioned phase inversion, Tikhonov regularization is used to introduce a suitable regularization parameter. In the large-phase deconvolution algorithm, a stable reconstructed phase is obtained after 10 iterations, and the reconstructed phase across the entire field of view is as follows: Figure 2 As shown in (a), phase values ​​are extracted along the white dashed line in the figure, and a curve of the quantitative phase values ​​is plotted to compare the phase recovery accuracy of the two methods, yielding the following results. Figure 2 (b) and Figure 2 (c) Results. It is clearly observed that for mitotic cells with larger phase values, one-step deconvolution (green curve) provides weakened cell morphology features, while the large-phase deconvolution method (yellow curve) more accurately recovers the low-frequency components in the sample's quantitative information, demonstrating precise three-dimensional morphology data. Further, four regions of interest are selected for magnified display, such as... Figure 2 (d1)- Figure 2 (d2) Figure 2 (e1)- Figure 2 (e2) Figure 2 (f1)- Figure 2 (f2) Figure 2 (g1)- Figure 2 As shown in (g2), the cell outlines and internal subcellular structures of breast cancer cells are well represented under both methods. However, the regularization parameter, while removing noise, weakens low-frequency components, which, in addition to the mismatch of weak phase approximation, also leads to a reduction in the reconstructed phase. The large-phase deconvolution reconstruction algorithm, while maintaining high-resolution imaging, recovers the sample's accurate quantitative phase distribution as much as possible.

[0082] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

[0083] It should be understood that, in order to simplify the present invention and help those skilled in the art understand its various aspects, in the above description of exemplary embodiments of the present invention, various features of the present invention are sometimes described in a single embodiment or with reference to a single figure. However, the present invention should not be construed as including all features in the exemplary embodiments as essential technical features of the claims of this patent.

[0084] It should be understood that the modules, units, components, etc., included in the device of one embodiment of the present invention can be adaptively changed to be placed in a device different from that embodiment. Different modules, units, or components included in the device of the embodiment can be combined into a single module, unit, or component, or they can be divided into multiple sub-modules, sub-units, or sub-components.

Claims

1. A quantitative phase microscopic imaging method for large-phase objects, characterized in that, The imaging process steps are as follows: Step 1: Use a serial port to control a high-density programmable LED array, TFT-LCD, or spatial light modulator to generate a semi-circular illumination source and acquire intensity images of large-phase samples. Step 2: Perform differential calculation on the acquired intensity image and deconvolve it with the phase transfer function of the optical system to obtain the quantitative phase distribution of the sample as the initial phase value. Step 3: Construct a complex transmission function using initial phase value and uniform amplitude, and generate intensity difference images under asymmetric illumination using the Abbe method. At the same time, generate an intensity distribution that is linearly related to the phase using a weak phase transfer function. Subtracting the two intensity distributions yields the estimated nonlinear intensity error distribution; Step 4: Subtract the estimated nonlinear intensity error from the actual acquired intensity to obtain the linear intensity signal required for deconvolution; Step 5: Use the linear intensity estimated in Step 4 to perform deconvolution reconstruction to recover the updated phase signal of the current iteration; Step 6: Repeat steps 3 to 5 until the difference between the intensity signal of the acquired intensity differential image under asymmetric illumination generated by the Abbe method and the actual acquired intensity signal is lower than the set value, at which point the iteration terminates. Step 7: Perform an inverse Fourier transform on the phase spectrum obtained through iteration to obtain the quantitative phase result of the recovered large-phase object.

2. The quantitative phase microscopic imaging method for large-phase objects according to claim 1, characterized in that, The acquired intensity image is represented as follows: I(u)=Bδ(u)+A(u)ATF(u)+iΦ(u)WPTF(u) In the formula, u represents the frequency component in the frequency domain, and B represents the background term, expressed as B=∫∫S(u j )|P(u j )| 2 d 2 u j u j S(u) represents the phase shift of a single angular illumination source in the frequency domain. j P(u) represents the tilted illumination light at a single angle, and P(u) represents the pupil function of the objective lens in the Fourier plane. j The phase shift u represents the phase shift of a single angular illumination source in the frequency domain. j Let A(u) be the pupil function of the objective lens on the Fourier plane, A(u) be the amplitude spectrum distribution of the sample, Φ(u) be the phase spectrum distribution of the sample, and ATF(u) be the amplitude transfer function, expressed as ATF(u)=∫∫S(u) j )P(u j )[P(u+u j )+P * (uu j )]d 2 u j , P(u+u j ) and P(uu j This indicates a phase shift of u in the frequency domain. j and -u j The objective lens has a pupil function in the Fourier plane, WPTF(u) represents the phase transfer function, and δ(u) represents the impulse function in the frequency domain.

3. The quantitative phase microscopic imaging method for large-phase objects according to claim 1, characterized in that, The specific steps for performing differential calculations on the acquired intensity images and deconvolving them with the phase transfer function of the optical system to obtain the quantitative phase distribution of the sample as the initial phase value are as follows: The acquired intensity images are differentially analyzed to obtain the intensity signal after differential calculation. The intensity signal after differential calculation The phase transfer function of the optical system is deconvolved with the phase transfer function of the optical system to obtain the solution. In the formula, u represents the frequency component in the frequency domain, u j S represents the phase shift of a single angular illumination source in the frequency domain. lr (u j P(u) represents the tilted illumination light under differential illumination. j The phase shift u represents the phase shift of a single angular illumination source in the frequency domain. j The pupil function of the objective lens in the Fourier plane, P(u+u) j ) and P(uu j This indicates a phase shift of u in the frequency domain. j and -u j The pupil function of the objective lens in the Fourier plane; The quantitative phase distribution of the sample obtained by deconvolution is as follows: In the formula, WPTF i DPC (u) represents the phase transfer function, I i DPC (u) represents the light intensity distribution, α represents the regularization parameter, and i represents the asymmetry axis of the illumination.

4. The quantitative phase microscopy imaging method for large-phase objects according to claim 1, characterized in that, A complex transmission function is constructed using an initial phase value and a uniform amplitude. First, the Abbe method is used to generate an intensity difference image under asymmetric illumination. Then, a weak phase transfer function is used to generate an intensity distribution linearly related to the phase. Finally, the two intensity distributions are subtracted to obtain the specific formula for the nonlinear intensity error distribution: In the formula, S(u j ) represents the lighting source, where u j e represents the spatial frequency corresponding to a single illumination angle. jφ (x) is the complex transmittance function of the object, P(u) is the pupil function, and WPTF(u) is the complex transmittance function of the object. j The phase transfer function of the weak object is given by φ(u), which is the phase distribution obtained in step two, and φ(x) represents the phase distribution of the sample.

5. The quantitative phase microscopy imaging method for large-phase objects according to claim 1, characterized in that, The linear intensity signal required for deconvolution is specifically: In the formula, This represents the estimated linear intensity signal. This represents the intensity image after intensity difference analysis. As a nonlinear intensity error, this calculation can make the intensity signal approximate the ideal input of the deconvolution reconstruction algorithm.

6. The quantitative phase microscopic imaging method for large-phase objects according to claim 1, characterized in that, The specific formula for deconvolution reconstruction using the linear intensity estimated in step four is as follows: In the formula, WPTF i (u) represents the phase transfer function. Let represent the linear intensity signal, α represent the regularization factor, n represent the number of iterations, and i represent the asymmetry axis of the illumination.

Images