Programmable annular led illumination-based high efficiency quantitative phase microscopy imaging method

Abstract

The invention discloses a programmable annular LED illumination-based high efficiency quantitative phase microscopy imaging method, the proposed method comprising the following steps: the derivation of system optical transfer function in a partially coherent illumination imaging system; the derivation of phase transfer function with the weak object approximation under the illumination of tilted axially symmetric coherent point illumination source; the extension of illumination from an axially symmetric coherence point source to a discrete annular point source, and the optical transfer function can be treated as an incoherent superposition of each pair of tilted axially symmetric coherent point sources. The acquisition of raw intensity dataset; the implementation of deconvolution for quantitative phase reconstruction. The invention derives the system phase transfer function under the tilted axially symmetric point light source in the case of partially coherent illumination, and promotes the optical phase transfer function of the discrete annular point light source. The programmability characteristic of LED array enables the annular illumination aperture to be flexibly adjustable, being applicable to different microscopic objects with different numerical apertures, and improving the compatibility and flexibility of the system.

Claims

1. A programmable annular LED illumination-based high efficiency quantitative phase microscopy imaging method, comprising: providing an imaging system comprising an LED array, a stage, a condenser, a sample to be tested, a microscope objective, an imaging tube lens and a camera, the system using an annular illumination pattern and adding the condenser to a light path, wherein the LED array is placed at a front focal plane of the condenser, a center of the LED array is on an optical axis of the microscope objective, a back focal plane of the microscope objective coincides with a front focal plane of the imaging tube lens, an imaging plane of the camera is placed at a back focal plane position of the imaging tube lens, the sample to be tested on the stage is adjusted to a front focal plane position of the microscope objective during imaging to form an infinity correction imaging system, the annular illumination pattern is displayed on the LED array, light emitted by each of the illuminated LED units is converged by the condenser into partially coherent light illuminating the sample to be tested which is placed on the stage, the light passes through the sample to be tested, is concentrated by the imaging tube lens and then reaches the imaging plane of the camera, and the stage is then moved along the optical axis and three intensity images are collected, the method further comprising the steps of: Step one of deriving an optical transfer function for the imaging system, wherein pupil of an illumination source displayed on the LED array and the microscopic objective are both symmetric about the optical axis, and the Step one further comprises a sub-step of deriving a corresponding expression of the optical transfer function under a weak object approximation in the imaging system; Step two of deriving a phase transfer function under weak object approximation for a tilted axially symmetric coherent point illumination source on the LED array, wherein the Step two comprises the sub-steps of: degenerating illumination displayed on the LEI) array from a circular partially coherent illumination pattern to a tilted axially symmetric coherent point source; matching two tilted axially symmetric discrete coherent point sources on the LED array with a pupil edge of the microscope objective, respectively; mapping the two tilted axially symmetric coherent point sources to the LED array on the focus plane of the condenser; calculating the corresponding optical transfer function of the imaging system with the LED array illumination pattern under a weak object approximation; generating a phase contrast intensity image containing phase information by introducing defocus, wherein a position of the intensity image does not locate at an in-focus plane; and transmitting the phase information of the sample to the phase contrast intensity image; Step three of extending illumination displayed on the LED array from an axially symmetric coherence point source to a discrete annular point source, wherein the Step three comprises sub-steps of decomposing an axially symmetric partially coherent illumination source into a plurality of axially symmetric discrete LED coherent point sources on a plane of the LED array; treating the optical transfer function as an incoherent superposition of each pair of tilted axially symmetric coherent point sources; and when the illumination source is in a shape of a discrete annular illumination pattern composed of discrete LED, obtaining the optical transfer function of the annular illumination pattern under the discrete condition; Step four of acquiring raw intensity dataset by the camera when the annular LED illumination pattern displayed on the LED array matches the pupil of the microscope objective, wherein the raw intensity dataset includes two defocus intensity images and an in-focusing intensity image with the movement of the stage along the optical axis; and Step five of implementing of deconvolution for quantitative phase reconstruction, the Step five comprising the sub-steps of subjecting the two defocus intensity images and an in-focusing intensity image acquired by the camera to central axial intensity difference; removing absorption components in the intensity images; performing a Fourier transform, wherein the Fourier transform is corresponding to a dividing of phase transfer function in a frequency domain; adding a regularization parameter to prevent an occurrence of dividing zero; and performing an inverse Fourier transform to obtain the quantitative phase microscopic image based on annular LED illumination.

2. The programmable annular LED illumination-based high efficiency quantitative phase microscopy imaging method according to claim 1, wherein the Step one is performed according to the following formulas:
I(r)=|h(r)|.sup.2.Math.|t(r)|.sup.2.Math.I.sub.u(r), where r is two-dimensional variables in a spatial domain, h(r) is an amplitude point spread function of the imaging system, t(r) is a complex amplitude of an object, I.sub.u(r) represents the superposition of the intensity produced by all the point sources on source plane,
I(r)=|h(r).Math.t(r)|.sup.2,
I(r)=a.sub.0.sup.2TCC(0;0)+2a.sub.0 Re{∫TCC(u;0)[Δ)+ia.sub.0)]exp(i2πru)du}, where a.sub.0 is an average of amplitudes in a complex amplitude, TCC (0; 0) is a transmitted component of an incident ray to an object, {tilde over (ϕ)}(u) represents the Fourier transform of the phase of an object, and TCC (u; 0) is described as the optical transfer function under a weak object approximation (WOTF):
WOTF(u)≡TCC(u;0)=∫∫S(u′)P(u′+u)P(u′)du′ where u represents a two-dimensional variable of polar coordinate system in frequency domain, u is a temporary integral variable in frequency domain, S (u) is a distribution of illumination source on the front focal plane of a concentrator, and P (u) is the pupil function of a microscopic objective, and $.Math.P\left(u\right).Math.=\{\begin{array}{cc}1,& \mathrm{if}u\le {\rho }_P\\ 0,& \mathrm{if}u>{\rho }_P\end{array},$ where p.sub.p is a normalized cutoff frequency of the microscope objective pupil.

3. The programmable annular LED illumination-based high efficiency quantitative phase microscopy imaging method according to claim 2, wherein the Step two is performed according to the following formulas: $P\left(u\right)=.Math.P\left(u\right).Math.e^{ikz\sqrt{1-{\lambda }^2{.Math.u.Math.}^2}},.Math.u.Math.\lambda \le 1,$ $WOTF\left(u\right)=S\left(u^{\prime }\right).Math.P^{\star }\left(u^{\prime }\right).Math..Math.P\left(u^{\prime }+u\right).Math.\exp \left[ikz\left(-\sqrt{1-{\lambda }^2{.Math.u^{\prime }.Math.}^2}+\sqrt{1-{\lambda }^2{.Math.u+u^{\prime }.Math.}^2}\right)\right]d{u}^{\prime },$ where amplitude transfer function H.sub.A (u) and phase transfer function H.sub.P(u) correspond to a real part and an imaginary part of WOTF, respectively, and are expressed as:
H.sub.A(u)=2a.sub.0Re[WOTF(u)]
H.sub.P(u)=2a.sub.0Im[WOTF(u)] where Re and Im labels represent the real part and the imaginary part of the function, respectively,
S(u)=δ(u−ρ.sub.s)+δ(u+ρ.sub.s) where δ represents Dirac delta function, ρ.sub.s is a normalized frequency distance from the point source to the center of the source; S(u) is substituted into WOTF, and the point sources with different ρ.sub.s corresponds to the different distribution of illumination sources,
H.sub.p(u).sub.obl=|P(u−ρ.sub.s)|sin [kz(√{square root over (1−λ.sup.2|u−ρ.sub.s|.sup.2)}−√{square root over (1−λ.sup.2|ρ.sub.s|.sup.2)})]+|P(u+ρ.sub.s)|sin [kz(√{square root over (1−λ.sup.2|u−ρ.sub.s|.sup.2)}−√{square root over (1−λ.sup.2|ρ.sub.s|.sup.2)})], where |P(u−ρ.sub.s)| and |P(u+ρ.sub.s)| are a pair of objective pupil functions shifted by an tilted point source, and
H.sub.p(u)=|P(u)|sin(πλz|u|.sup.2)≈|P(u)|πλz|u|.sup.2.

4. The programmable annular LED illumination-based high efficiency quantitative phase microscopy imaging method according to claim 1, wherein in the Step three, annular sources finally displayed on LED array are represented as: $S\left(u\right)=\underset{i=0}{\overset{N}{.Math.}}\delta \left(u_i\right),.Math.u_i.Math.\approx .Math.{\rho }_p.Math.$ where N is number of LEDs on the source plane, a distance from each illuminated LED point source to a center of the LED array is ρ.sub.p, a numerical aperture of the microscopic objective lens is NA.sub.obj, magnification of imaging system is Mag, and a camera pixel size is Δx.sub.com, which satisfies $\frac{\mathrm{Mag}}{4\Delta x_{\mathrm{cam}}}\frac{\lambda }{{\mathrm{NA}}_{\mathrm{obj}}}>1.$

5. The programmable annular LED illumination-based high efficiency quantitative phase microscopy imaging method according to claim 1, wherein the Step five is performed according to the following formulas of: $-k\frac{\partial I\left(r\right)}{\partial z}=\nabla .Math.\left[I\left(r\right)\nabla \varphi \left(r\right)\right],$ $\frac{{\tilde{I}}_1\left(u\right)-{\tilde{I}}_2\left(u\right)}{4\tilde{I}\left(u\right)}=\left(\pi \lambda z{.Math.u.Math.}^2\right)\tilde{\varphi }\left(u\right),$ $\frac{{\tilde{I}}_{\Delta z}-{\tilde{I}}_{-\Delta z}\left(u\right)}{4{\tilde{I}}_0\left(u\right)}=\mathrm{Im}\left[\mathrm{WOTF}\left(u\right)\right]\tilde{\varphi }\left(u\right),$ and $\varphi \left(r\right)={-1}^{}\left\{\frac{{\tilde{I}}_{\Delta z}-{\tilde{I}}_{-\Delta z}\left(u\right)}{4{\tilde{I}}_0\left(u\right)}\frac{\mathrm{Im}\left[\mathrm{WOTF}\left(u\right)\right]}{{.Math.\mathrm{Im}\left[\mathrm{WOTF}\left(u\right)\right].Math.}^2+\alpha }\right\},$ where I(r) is the intensity distribution on a focal plane, ϕ(r) is a phase distribution of an object, Ĩ.sub.1(u), Ĩ.sub.2(u), Ĩ(U) are Fourier transform of three captured intensity images acquired in the case of weak defocus, πλz|u|.sup.2 is a Laplacian filter function, an inverse Laplacian function 1/(πλz|u|.sup.2) corresponds to the inverse form of the phase transfer function in the case of coherent illumination and weak defocus approximation, .sup.−1 represents the inverse Fourier transform, and parameter α represents regularization parameter avoiding the division by zero.

Description

DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 is a schematic diagram of a light source pattern for a coherent illumination, a partially coherent illumination, and a discrete annular illumination imaging system.

(2) FIG. 2 is a schematic diagram of a microscope optical path based on a programmable annular LED illumination based quantitative phase imaging.

(3) FIG. 3 is a schematic diagram of a comparison of system phase transfer functions corresponding to various illumination source modes at different defocus distances.

(4) FIG. 4 is a schematic flow chart of the microscopy imaging method of the present invention.

(5) FIG. 5 is a schematic diagram showing the results of quantitative phase imaging of human lung cancer cells by using the imaging method of the present invention.

(6) FIG. 6 is a schematic diagram showing the quantitative phase of high resolution human cervical cancer cells obtained by the imaging method of the present invention and a microscopic objective with a high numerical aperture.

THE EMBODIMENTS

(7) As shown in FIG. 2, the present invention is a programmable annular LED illumination-based high efficiency quantitative phase microscopy imaging system, the actual hardware platform of the imaging system comprises an LED array, a stage, a condenser, a sample to be tested, and a microscope objective, an imaging tube lens and a camera, the system using an annular illumination pattern and adding a condenser to the light path. The LED array is placed at the front focal plane of the condenser and the center of the LED array is on the optical axis of the microscope objective; the back focal plane of the microscope objective coincides with the front focal plane of the imaging tube lens, and the imaging plane of the camera is placed at the back focal plane position of the imaging tube lens; the sample to be tested on the stage is adjusted to the front focal plane position of the microscope objective during imaging to form an infinity correction imaging system. An annular illumination pattern is displayed on the LED array, and the light emitted by each of the illuminated LED units is converged by the condenser into partially coherent light illuminating the sample to be tested which is placed on the stage. The light passes through the sample to be tested, is concentrated by the imaging tube lens and then reaches the imaging plane of the camera. The stage is then moved along the optical axis and three intensity images are collected.

(8) In order to meet the minimum frequency domain sampling rate required by the imaging method, the numerical aperture of the microscope objective is NA.sub.obj, and the distance from each lit LED unit on the annular illumination pattern to the center of the LED array is l and satisfies

(9) ${\mathrm{NA}}_{\mathrm{obj}}=\frac{l}{\sqrt{l^2+f^2}},$
where f is the focal length of the condenser, generally between 10-20 mm. The magnification of the microscope objective is Mag, the pixel size of the camera is Δx.sub.cam, and the wavelength of illumination light is λ, and satisfies

(10) $\frac{\mathrm{Mag}}{4\Delta x_{\mathrm{cam}}}\frac{\lambda }{{\mathrm{NA}}_{\mathrm{obj}}}>1.$
In order to meet the needs of microscope objectives with different numerical apertures, the radius of the annular illumination pattern can be changed through reprogramming, that is, illumination sources always satisfies

(11) ${\mathrm{NA}}_{\mathrm{obj}}=\frac{l}{\sqrt{l^2+f^2}},$
at this point, the radius of an illumination ring always matches the numerical aperture of an objective lens, as shown in FIG. 1(d).

(12) An LED array includes a plurality of (at least 261) LED units that are equally spaced to form a two-dimensional array. Each of the LED units is colored red, green and blue, and its typical wavelength is red light 633 nm, green light 525 nm, and blue light 465 nm. The typical center distance d between each LED unit is 1-4 mm. An LED array does not need to be independently machined and can be directly purchased on the market. Table 1 shows product parameters of a commercially available LED array. In the LED array, the LED units have 32 rows and 32 columns, 1024 in total, and the brightness of each LED unit is above 2000 cd/m.sup.2.

(13) TABLE-US-00001 TABLE 1 Physical Parameters of the LED Array Items Parameters wavelength of LED unit red 633 nm, green 525 nm, blue 475 nm number of LED units 32 × 32 spacing of LED units 1.67 mm light emitting surface 150 μm size of LED unit brightness of LED unit 2000 cd/m.sup.2 array dimensions 55 mm × 55 mm × 17 mm weight 170 g lighting angle 150° voltage 5 v current the maximum 2 A

(14) Each LED unit in an LED array can be individually illuminated by programming. The specific method of lighting LED units is a conventional technology, and the implementation circuit can adopt (not limited to) existing technology such as a microcontroller unit, an ARM, or a programmable logic device. The specific implementation methods can refer to relevant literature (Guo Baozeng, Deng Yumiao: FPGA-based LED display control system design [J]. LCD and Display, 2010, 25(3): 424-428).

(15) The method for realizing high efficiency quantitative phase microscopy imaging using the imaging system based on the programmable annular LED illumination comprises the following steps:

(16) Step one, the derivation of optical transfer function for a partially coherent imaging system; considering that the optical pupil functions of the illumination source and microscopic objective are symmetric about the optical axis, the optical transfer function under a weak object approximation in a partially coherent illumination imaging system can be derived.

(17) The specific implementation process is: in an infinity-corrected imaging light-path composed of the programmable annular LED illumination-based high efficiency quantitative phase imaging system, the intensity image on the imaging plane for incoherent illumination is
I(r)=|h(r)|.sup.2.Math.|t(r)|.sup.2.Math.I.sub.u(r)
wherein r is two-dimensional variables in a spatial domain, and h(r) is an amplitude point spread function of the imaging system. t(r) is the complex amplitude of an object, I.sub.u(r) represents the superposition of the intensity produced by all the point sources on source plane; while the intensity images on the imaging plane for coherent illumination imaging systems can be expressed as:
I(r)=|h(r).Math.t(r)|.sup.2
for a partially coherent imaging system, the relationship between the intensity captured on the imaging plane and the imaging system is expressed as:
I(r)=a.sub.0.sup.2TCC(0;0)+2a.sub.0 Re{∫TCC(u;0)[Δ)+ia.sub.0)]exp(i2πru)du}
where a.sub.0 is the average of the amplitudes in the complex amplitude, TCC (0; 0) is the transmitted component of the incident ray to an object, represents the Fourier transform of the phase of an object, and TCC (u; 0) can be described as the optical transfer function under a weak object approximation (WOTF):
WOTF(u)≡TCC(u;0)=∫∫S(u′)P(u′+u)P(u′)du′
The above formula is the WOTF of the imaging system, where u represents the two-dimensional variable of polar coordinate system in frequency domain, u′ is the temporary integral variable in frequency domain, S(u) is the distribution of illumination source on the front focal plane of a concentrator, and P(u) is the pupil function of a microscopic objective, and the absolute value of the optical pupil function can be expressed as:

(18) $.Math.P\left(u\right).Math.=\{\begin{array}{cc}1,& \mathrm{if}u\le {\rho }_P\\ 0,& \mathrm{if}u>{\rho }_P\end{array}$
wherein ρ.sub.P is the normalized cutoff frequency of the microscope objective pupil.

(19) Step two, the derivation of the phase transfer function under weak object approximation for a tilted axially symmetric coherent point illumination source; when the illumination degenerates from a circular partially coherent illumination pattern to a tilted axially symmetric coherent point source, the two axially symmetric discrete coherent point sources are matched with the pupil edge of an objective lens, respectively; then these two tilted axially symmetric coherent point sources are mapped to the LED array on the source plane, and the optical transfer function of the imaging system under a weak object approximation can be calculated and finally, the phase contrast intensity image containing phase information is generated by introducing defocus, so that the phase information of a sample can be transmitted to the defocus intensity image.

(20) The specific implementation process is: if the source distribution S(u) and the microscope objective pupil P (u) are axially symmetrically distributed, WOTF is a real function and an even function; when the sample is in-focus, the phase information of object cannot be transmitted to the intensity image; Only when the phase contrast and the imaginary component are introduced into WOTF by axial defocusing mode, the pupil function at the back focal plane of a microscope objective can be expressed as

(21) $P\left(u\right)=.Math.P\left(u\right).Math.e^{\mathrm{ikz}\sqrt{1-{\lambda }^2{.Math.u.Math.}^2}},.Math.u.Math.\lambda \le 1$
then, substituting the pupil function to WOTF and obtaining the complex optical transfer function in the defocus condition:
WOTF(u)=S(u′)|P*(u′)∥P(u′+u)exp[ikz(−√{square root over (1−λ.sup.2)}+√{square root over (1−λ.sup.2|u+u′|.sup.2)})]du′
the amplitude transfer function H.sub.A(u) and phase transfer function H.sub.P(u) correspond to the real part and the imaginary part of WOTF, respectively, and these two transfer functions can be expressed as:
H.sub.A(u)=2a.sub.0Re[WOTF(u)]
H.sub.P(u)=2a.sub.0Im[WOTF(u)]
where Re and Im labels represent the real part and the imaginary part of the function, respectively; by introducing two tilted axially symmetric coherent point sources on the source planes where LED array is located, and the expression of source distribution S(u) is:
S(u)=δ(u−ρ.sub.s)+δ(u+ρ.sub.s)
where δ represents Dirac delta function, and ρ.sub.s is the normalized frequency distance from the point source to the center of the source. S(u) is substituted into WOTF, and the point sources with different ρ.sub.s corresponds to the different distribution of illumination sources. when ρ.sub.s≠0, the sources generate two-axially symmetric tilted illumination, and the phase transfer function at the moment is:
H.sub.p(u).sub.obl=|P(u−ρ.sub.s)|sin [kz(√{square root over (1−λ.sup.2|u−ρ.sub.s|.sup.2)}−√{square root over (1−λ.sup.2|ρ.sub.s|.sup.2)})]+|P(u+ρ.sub.s)|sin [kz(√{square root over (1−λ.sup.2|u−ρ.sub.s|.sup.2)}−√{square root over (1−λ.sup.2|ρ.sub.s|.sup.2)})]
where |P(u−ρ.sub.s)| and |P(u+ρ.sub.s)| are a pair of objective pupil functions shifted by an tilted point source; While ρ.sub.s=0, the two shifted pupil functions overlap each other at the center position, and the phase transfer function under the coherent condition is obtained for this situation; The transfer function of the transport of intensity equation is derived by introducing the paraxial approximation and weak defocus approximation. Thus, the transfer function of the transport of intensity equation in the frequency domain is:
H.sub.p(u)=|P(u)|sin(πλz|u|.sup.2)≈|P(u)|πλz|u|.sup.2.

(22) Step three, the extension of illumination from an axially symmetric coherence point source to a discrete annular point source; any axially symmetric partially coherent illumination source can be decomposed into a plurality of axially symmetric discrete coherent point sources on the source plane, and the optical transfer function can be treated as an incoherent superposition of each pair of tilted axially symmetric coherent point sources, and when the illumination source is in the shape of a discrete annular illumination pattern composed of discrete LED, the optical transfer function of the annular illumination pattern under the discrete condition can be obtained as well.

(23) The specific implementation process is: any illumination pattern about the axially symmetric partially coherent illumination can be decomposed into a plurality of axially symmetric coherent point sources, And the transfer function of this illumination pattern can be composed of the transfer function incoherent superposition of each pair of tilted axially symmetric coherent point sources, so the annular sources finally displayed on LED array are represented as:

(24) $S\left(u\right)=\underset{i=0}{\overset{N}{.Math.}}\delta \left(u_i\right),.Math.u_i.Math.\approx .Math.{\rho }_p.Math.$
where N is the number of LEDs on the source plane, the distance from each illuminated LED point source to the center of the LED array is ρ.sub.p. Therefore, the annular LED illumination pattern is substantially matching a microscope objective pupil; In order to meet the minimum frequency domain sampling rate required by imaging, the numerical aperture of the microscopic objective lens is NA.sub.obj. The magnification of imaging system is Mag and the camera pixel size is Δx.sub.com, which satisfies

(25) $\frac{\mathrm{Mag}}{4\Delta x_{\mathrm{cam}}}\frac{\lambda }{{\mathrm{NA}}_{\mathrm{obj}}}>1;$
Thus, the phase transfer function of annular LED illumination under the discrete condition can be obtained through the coherence mode decomposition theory;
when the imaging system contains an objective lens with a different numerical aperture, the LED array can be re-programmed to change the size of annular illumination pattern, so the illumination pattern can re-match the pupil of the microscope objective, that is the relationship

(26) ${\mathrm{NA}}_{\mathrm{obj}}=\frac{l}{\sqrt{l^2+f^2}}$
can be satisfied. In the imaging system, the radius of annular the illumination is always matched with the numerical aperture of the objective lens, and the phase transfer function of annular illumination can be calculated based on the corresponding parameters of optical imaging system.

(27) As shown in FIG. 3, by comparing the amplitude magnitude of the phase transfer function in the optical system with the illumination modes of different light sources, the phase transfer function of the discrete annular point light source (FIG. 3(b) and FIG. 3(d)) is a basically relatively flat constant in the normalized spatial frequency with no zero crossing point (the zero crossing of transfer function will cause the spatial frequency near the zero crossing to be unrecoverable, so zero crossing of function should be avoided as much as possible); the annular illumination composed of the four LEDs cannot guarantee that the transfer function guarantee sufficient response in most directions, and therefore it can be seen from FIG. 3(d) that the two-dimensional phase transfer function can basically guarantee the spectral response coverage in most directions; Although as shown in FIG. 3(a), The amplitude of the phase transfer function response in the case of coherent illumination is larger, its spatial frequency is limited by the objective lens to coherent diffraction limit, resulting in a final phase reconstruction resolution reduction; although the phase transfer function in FIG. 3(c) can extend the cutoff frequency of the transfer function to twice the objective lens spatial frequency, the response amplitude is around 10′, that is, under traditional circular Kohler illumination with coherence parameter σ=0.99, the phase information contained in generated intensity images is insufficient to be detected by a camera; resulting in a very low reconstructed signal to noise ratio.

(28) when an imaging system is switched to a microscope objective with a different numerical aperture, the LED array is re-programmed to change the annular illumination pattern so that it can be re-cut into the pupil of the microscope objective, namely, satisfying

(29) ${\mathrm{NA}}_{\mathrm{obj}}=\frac{l}{\sqrt{l^2+f^2}},$
at the moment, the radius of the illumination circular ring is always matched with the numerical aperture of the objective lens, and the phase transfer function of the annular illumination corresponding to the parameters of the optical imaging system at the moment is calculated.

(30) Step four, the acquisition of raw intensity dataset; when the annular LED illumination pattern matches the objective lens pupil, the camera is used for the acquisition of two defocus intensity images and an in-focusing intensity image with the movement of stage along the optical axis direction.

(31) Step five, the implementing of deconvolution for quantitative phase reconstruction; three intensity images acquired by the camera are subjected to central axial intensity difference, and the absorption components in the intensity images are removed; and then the Fourier transform is performed, and this transform is corresponding to the dividing of phase transfer function in the frequency domain; moreover, the regularization parameter is added to prevent the occurrence of dividing zero; finally, the inverse Fourier transform is performed to obtain the quantitative phase microscopy image based on annular LED illumination.

(32) The specific implementation process is: the conventional transport of intensity equation is expressed as:

(33) 0$-k\frac{\partial I\left(r\right)}{\partial z}=\nabla .Math.\left[I\left(r\right)\nabla \varphi \left(r\right)\right]$
where I(r) is the intensity distribution on a focal plane, and ϕ(r) is the phase distribution of object. The transport of intensity equation is solved by using the Poisson equation-based fast Fourier transform, and the detailed formula can be written as the Laplacian filter form in the frequency domain:

(34) $\frac{{\tilde{I}}_1\left(u\right)-{\tilde{I}}_2\left(u\right)}{4\tilde{I}\left(u\right)}=\left(\pi \lambda z{.Math.u.Math.}^2\right)\tilde{\varphi }\left(u\right)$
where Ĩ.sub.1(u), Ĩ.sub.2(u), Ĩ(U) are the Fourier transform of three captured intensity images acquired in the case of weak defocus, πλz|u|.sup.2 is the Laplacian filter function, and the inverse Laplacian function 1/(πλz|u|.sup.2) corresponds to the inverse form of the phase transfer function in the case of coherent illumination and weak defocus approximation. Generally, the forward form of the optical transfer function under weak object approximation for more general partially coherent illumination can be expressed as:

(35) $\frac{I_{\Delta z}^{\%}-I_{-\Delta z}^{\%}\left(u\right)}{4I_0^{\%}\left(u\right)}={\mathrm{Im}\left[\mathrm{WOTF}\left(u\right)\right]}_{\varphi }^{\%}\left(u\right)$
where the phase transfer function of an imaging system corresponds to the imaginary part of optical transfer function under the weak object approximation. Thus, the Fourier transform of the quantitative phase of an object can be obtained through the inverse form of the phase transfer function, and the quantitative phase distribution of an annular LED illumination can be obtained through inverse Fourier transform:

(36) $\varphi \left(r\right)={-1}^{}\left\{\frac{I_{\Delta z}^{\%}-I_{-\Delta z}^{\%}\left(u\right)}{4I_0^{\%}\left(u\right)}\frac{\mathrm{Im}\left[\mathrm{WOTF}\left(u\right)\right]}{{.Math.\mathrm{Im}\left[\mathrm{WOTF}\left(u\right)\right].Math.}^2+\alpha }\right\}$
where .sup.−1 represents the inverse Fourier transform and parameter α represents regularization parameter avoiding the division by zero.

(37) As shown in the reconstruction flowchart FIG. 4, the imaging method of the present invention is applied to quantitative phase imaging of unstained human cancer cells through the above steps and procedure, FIG. 5 is the quantitative phase imaging result of the imaging method of the present invention on human lung cancer cells, wherein FIG. 5 (a1) and FIG. 5 (a2) are quantitative phase results on the entire camera imaging region under coherent illumination and annular illumination respectively; three sub-region quantitative phases are selected from FIG. 5 (a) and their phase gradient images are calculated to do comparison, as shown in FIG. 5 (b), FIG. 5 (c) and FIG. 5 (d), the phase imaging resolution under annular illumination is higher than that of a traditional coherent illumination. Two points are respectively selected from three phase sub-regions under the two illumination modes, and the phase change curve between the two points is drawn. From the curve changes, it can be seen that the resolution of annular illumination program double that of conventional coherent illumination (in the first phase one-dimensional drawing post line diagram, the variable aperture frequency of triangular dotted lines is twice the solid line of the original point), so the annular LED illumination proposed by the present invention can achieve twice the imaging resolution of objective lens. FIG. 6 is an obtained high-resolution quantitative phase diagram and a high-resolution phase gradient diagram of human cervical cancer cells by using the imaging method of the present invention and a high numerical aperture microscope objective. The numerical aperture of the microscope objective is 0.75, the annular illumination pattern is matched to the pupil of objective lens by re-programming, and the final imaging resolution is twice resolving power of lens. By selecting three sub-regions to be zoomed in, as shown in FIG. 6(c1)-(c3), the granular organelles in cytoplasm and the nucleus of cervical cancer cells can be clearly observed.