Multi-scale photoacoustic microscopic imaging method based on single-pixel imaging

Abstract

The invention discloses a multi-scale photoacoustic microscopic imaging method based on single pixel imaging. Comprising the following steps: 1) generating a series of stripe light illumination samples with different spatial frequencies on an XY two-dimensional plane through a spatial light modulator, applying phase shift to the stripe light with each spatial frequency by adopting a phase shift method, and collecting photoacoustic signals under each phase shift by utilizing a single ultrasonic probe; 2) calculating a Fourier spectrum coefficient according to the photoacoustic signal under each spatial frequency, introducing an impulse response function of the ultrasonic probe to correct the Fourier spectrum coefficient, and then reconstructing a photoacoustic image through inverse Fourier transform based on a simplified point spread function (PSF) according to the corrected Fourier spectrum coefficient; and (3) controlling the acquisition of the Fourier spectrum coefficient by controlling the spatial frequency of the stripe light so as to realize multi-scale imagines.The method solves the problem that the traditional photoacoustic microscopic imaging can only carry out imaging on one scale due to the fixed resolution, and multi-scale information cannot be fully utilized.

Description

Technical Field

[0001] The present invention relates to the field of life science imaging technology, and specifically to a multi-scale photoacoustic microscopy imaging method based on single-pixel imaging. Background Art

[0002] Photoacoustic imaging is a non-invasive imaging technology based on the photoacoustic effect that has emerged in the field of biomedical imaging in recent years. It combines the advantages of high contrast in optical imaging and high penetration in acoustic imaging, and has good clinical translation potential and application prospects.

[0003] As a new type of hybrid imaging technology, photoacoustic microscopy combines the high contrast of optics and the low scattering and high resolution of acoustics, and can achieve deeper imaging of biological tissues. Currently, it has a wide range of applications in biological imaging, such as tumor detection, blood vessels, molecular imaging, etc. However, traditional photoacoustic microscopy requires mechanical point scanning to obtain three-dimensional imaging information, resulting in slow imaging speed. At the same time, traditional photoacoustic microscopy can only image at one scale due to fixed resolution, and cannot fully utilize multi-scale information for integrated imaging, which limits the application of photoacoustic microscopy. Summary of the Invention

[0004] The purpose of the present invention is to address the deficiencies of the prior art and provide a multi-scale photoacoustic microscopy imaging method based on single-pixel imaging to solve the problems of long imaging time and slow volume speed caused by photoacoustic microscopy under ill-conditioned conditions (such as long mechanical point scanning time for volume imaging and fixed resolution).

[0005] To achieve the above purpose, the present invention adopts the following technical solutions.

[0006] A multi-scale photoacoustic microscopy imaging method based on single-pixel imaging includes the following steps: Step S1: Generate a series of fringe lights with different spatial frequencies in the XY two-dimensional plane through a spatial light modulator to illuminate the sample, and apply a phase shift to the fringe light of each spatial frequency using the phase-shift method. Use a single ultrasonic probe to collect the photoacoustic signals under each phase shift. Step S2: Calculate the Fourier spectrum coefficients based on the photoacoustic signals at each spatial frequency, introduce the pulse response function of the ultrasonic probe to correct the Fourier spectrum coefficients, and then reconstruct the photoacoustic image through inverse Fourier transform based on the simplified point spread function according to the corrected Fourier spectrum coefficients. Step S3: Control the acquisition of Fourier spectrum coefficients by controlling the spatial frequency of the fringe light, thereby achieving multi-scale imaging.

[0007] Specifically, in step S1, a series of fringe lights with different spatial frequencies are generated in the XY two-dimensional plane through a spatial light modulator to illuminate the sample, and for a given spatial frequency Sum phase fringe light, and the light intensity distribution of the fringe light is expressed as: ; In the above formula, x , y , z are respectively the xyz axis coordinates of the position where the sample is illuminated; The phase shift method is used to apply a phase shift to the fringe light of each spatial frequency, and a single ultrasonic probe is used to collect the photoacoustic signals under each phase shift. The collected photoacoustic signals are expressed as: ; In the above formula, is the light absorption distribution, is the area irradiated by the fringe light.

[0008] Specifically, the calculation method of the Fourier spectrum coefficients in step S2 is as follows: The phase shift method is used to apply a phase shift to the fringe light of each spatial frequency, and a single ultrasonic probe is used to collect the photoacoustic signals of four phases, which are respectively: , , , ; Then the light absorption distribution corresponding to the Fourier spectrum coefficient of this spatial frequency is expressed as: ; In the above formula, j is the imaginary unit.

[0009] Furthermore, the impulse response function of the ultrasonic probe is introduced in step S2 to correct the Fourier spectrum coefficients. The impulse response function is expressed as: ; In the above formula, β and are non-linear parameters that control the attenuation rate of the photoacoustic signal, β and ; j is the imaginary unit; is the amplitude of the k th resonance peak of the photoacoustic signal; is the attenuation coefficient of the k th resonance peak of the photoacoustic signal; is the time delay of the k th resonance peak of the photoacoustic signal; is the phase of the k n-th resonant peak of the photoacoustic signal; The corrected Fourier spectrum coefficient is expressed as: ; In the above formula, is the Wiener filtering weight function, which is used to suppress the amplification of high-frequency noise; ; In the above formula, , is the conjugate impulse response function; is an empirical constant, , is the signal-to-noise ratio at the spatial frequency point.

[0010] Furthermore, in step S2, the photoacoustic image is reconstructed by inverse Fourier transform based on the corrected Fourier spectrum coefficient and the simplified point spread function, and the point spread function is expressed as: ; In the above formula, indicates that in the actual imaging system, the spatial frequency allowed to pass through the optical system is limited, is the maximum spatial frequency of the given fringe. For an objective lens with a given finite numerical aperture, it should satisfy , where is the cut-off frequency of the objective lens, is the numerical aperture of the objective lens, is the optical wavelength; exp[*] is the exponential function; i is the imaginary unit; c.c. is a constant factor; Ignoring the constant term and the constant factor, in polar coordinates, the point spread function is further simplified to: ; The photoacoustic image reconstructed by inverse Fourier transform based on the simplified diffusion function PSF is expressed as: .

[0011] Specifically, in step S3, the control of the spatial frequency of the fringe light is achieved by adjusting the wavelength of the light source, the fringe period, and the parameters of the projection system, and the Fourier spectrum coefficient is obtained based on the circular sampling strategy.

[0012] Furthermore, the circular sampling strategy defines a circular sampling region in the frequency domain, and the radius of this region corresponds to the required spatial frequency range; the Fourier spectrum coefficient is obtained by uniformly sampling within this circular region.

[0013] Compared with the prior art, the present invention has the following beneficial effects: 1. The method of the present invention uses stripe light with different spatial frequencies to illuminate the sample. The photoacoustic signal obtained by a single ultrasonic probe serves as the Fourier spectrum coefficient of the sample structure information at this spatial frequency, and the image can be reconstructed through inverse Fourier transform; by controlling the stripe spatial frequency, the acquisition of the Fourier spectrum coefficient can be controlled, thereby realizing multi-scale imaging, overcoming the defect that traditional photoacoustic microscopy can only image at one scale due to fixed resolution and cannot make full use of multi-scale information.

[0014] 2. Through imaging experiments, the method of the present invention verifies that in actual imaging, the method of the present invention can enhance the flexibility of photoacoustic imaging, expand the application range of photoacoustic imaging, and flexible and rapid multi-scale imaging can further enhance the integrated imaging ability of photoacoustic imaging. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a flowchart of a multi-scale photoacoustic microscopy imaging method based on single-pixel imaging of the present invention; Figure 2 is a schematic diagram of the principle of a multi-scale photoacoustic microscopy imaging method based on single-pixel imaging of the present invention; Figure 3 is a schematic diagram of a multi-scale photoacoustic imaging platform based on single-pixel imaging constructed using k-Wave; Figure 4 is the reconstruction result of multi-scale photoacoustic microscopy imaging of a single small ball absorber; Figure 5 is the reconstruction result of multi-scale photoacoustic microscopy imaging of multiple small ball absorbers; Figure 6 is the result diagram of reconstructing blood vessels using a reconstruction algorithm through the k-Wave: MATLAB simulation toolbox. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] To facilitate understanding and implementation of the present invention by those of ordinary skill in the art, the following details each step of the method proposed by the present invention. It should be understood that these embodiments are only for illustrating the present invention and not for limiting the scope of the present invention. In addition, it should be understood that after reading the content taught by the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms also fall within the scope defined by the appended claims of this application.

[0017] Embodiment As Figure 1 shown, a multi-scale photoacoustic microscopy imaging method based on single-pixel imaging includes the following steps: Step S1: Generate a series of fringe lights with different spatial frequencies in the XY two-dimensional plane through a spatial light modulator to illuminate the sample, apply a phase shift to the fringe light of each spatial frequency using the phase-shifting method, and collect the photoacoustic signals at each phase shift using a single ultrasonic probe; Step S2: Calculate the Fourier spectrum coefficients based on the photoacoustic signals at each spatial frequency, correct the Fourier spectrum coefficients by introducing the impulse response function of the ultrasonic probe, and then reconstruct the photoacoustic image through inverse Fourier transform based on the simplified point spread function according to the corrected Fourier spectrum coefficients; Step S3: Control the acquisition of the Fourier spectrum coefficients by controlling the spatial frequency of the fringe light, thereby achieving multi-scale imaging.

[0018] Specifically, in step S1, a series of fringe lights with different spatial frequencies are generated in the XY two-dimensional plane through a spatial light modulator to illuminate the sample. For the fringe light with a given spatial frequency and phase , the light intensity distribution of the fringe light is expressed as: ; In the above formula, x , y , z are the xyz axis coordinates of the position where the sample is illuminated respectively; Apply a phase shift to the fringe light of each spatial frequency using the phase-shifting method, collect the photoacoustic signals at each phase shift using a single ultrasonic probe, and the collected photoacoustic signals are expressed as: ; In the above formula, is the light absorption distribution, is the area irradiated by the fringe light.

[0019] Specifically, the calculation method of the Fourier spectrum coefficients in step S2 is as follows: Apply a phase shift to the fringe light of each spatial frequency using the phase-shifting method, and collect the photoacoustic signals of four phases using a single ultrasonic probe, which are respectively: , , , ; Then the Fourier spectrum coefficient corresponding to the light absorption distribution at this spatial frequency is expressed as: ; In the above formula, j is the imaginary unit.

[0020] Further, in step S2, the Fourier spectrum coefficients are corrected by introducing the pulse response function of the ultrasonic probe, and the pulse response function is expressed as: ; In the above formula, β and are non-linear parameters for controlling the attenuation rate of the photoacoustic signal, β and ; j is the imaginary unit; is the amplitude of the k th resonance peak of the photoacoustic signal; is the attenuation coefficient of the k th resonance peak of the photoacoustic signal; is the time delay of the k th resonance peak of the photoacoustic signal; is the phase of the k th resonance peak of the photoacoustic signal; The corrected Fourier spectrum coefficient is expressed as: ; In the above formula, is the Wiener filtering weight function for suppressing high-frequency noise amplification; ; In the above formula, , is the conjugate pulse response function; is an empirical constant, , is the signal-to-noise ratio at the spatial frequency point.

[0021] Furthermore, in step S2, the photoacoustic image is reconstructed by inverse Fourier transform based on the corrected Fourier spectrum coefficients and the simplified point spread function, and the point spread function is expressed as: ; In the above formula, indicates that in the actual imaging system, the spatial frequency allowed to pass through the optical system is limited, is the given maximum spatial frequency of the fringe, and for a given objective lens with a finite numerical aperture, it should satisfy , where is the cut-off frequency of the objective lens, is the numerical aperture of the objective lens, is the optical wavelength; exp[*] is the exponential function; i is the imaginary unit; c.c. is a constant factor; Ignoring the constant term and constant factor, in polar coordinates, the point spread function is further simplified to: ; The photoacoustic image reconstructed by inverse Fourier transform based on the simplified diffusion function PSF is expressed as: .

[0022] In this embodiment, the illumination effect of fringe light with different spatial frequencies is equivalent to generating a virtual Bessel beam. The transverse resolution of the reconstructed photoacoustic image can be obtained from the full width at half maximum in the simplified point spread function expression as , by changing different transverse resolutions can be obtained, realizing large-range multi-scale imaging without movement. When the maximum spatial frequency of the fringe light is the cut-off frequency of the objective lens, that is, , the transverse resolution reaches the highest, which is ; assuming the numerical aperture of the objective lens , the wavelength , it is calculated that the transverse resolution at this time is about , and this transverse resolution can meet the high-resolution requirements for capillary imaging.

[0023] Specifically, the control of the spatial frequency of the fringe light in step S3 is achieved by adjusting the wavelength of the light source, the fringe period, and the parameters of the projection system, and the Fourier spectrum coefficients are obtained based on the circular sampling strategy.

[0024] Furthermore, the circular sampling strategy defines a circular sampling region in the frequency domain, and the radius of this region corresponds to the required spatial frequency range; by uniformly sampling within this circular region, the Fourier spectrum coefficients are obtained.

[0025] To verify the technical effect of the present invention, in this embodiment, the multi-scale imaging ability is tested by using the constructed single-pixel multi-scale imaging simulation model. The samples verified in this experiment are taken from a single spherical absorber, multiple spherical absorbers, and blood vessels respectively. When different fringe spatial frequencies ( , , ) are applied, most of the Fourier spectrum coefficients are obtained within the allowable frequency range of the system. The feasibility and effectiveness of the method of the present invention are confirmed by comparing the resolutions of the reconstructed images.

[0026] Figure 2 is a schematic diagram of the multi-scale photoacoustic microscopy imaging method based on single-pixel imaging of the present invention. Figure 2 In (a) is a schematic diagram that a sequence of fringe light illumination is equivalent to generating a virtual Bessel spot (system point spread function). Figure 2 In (b) -Figure 2 In (d), they are the equivalent Bessel light spots when the maximum spatial frequency of the striped light is , , respectively. Figure 2 In (e) is Figure 2 the light intensity distribution at the white dotted line in (b) - Figure 2 in (d) of [], reflecting the change in resolution; Figure 2 In [], NI represents the normalized light intensity, and PSF represents the point spread function.

[0027] Figure 3 is a schematic diagram of a multi-scale photoacoustic imaging platform based on single-pixel imaging constructed using k-Wave. The experimental platform is configured with a Windows 10 operating system with a 64-bit Intel® Core ™ i7-8750H CPU @ 2.20 GHz. A multi-scale photoacoustic imaging platform based on single-pixel imaging is constructed by using the k-wave simulation toolbox, and a series of coupled first-order equations are quickly solved by the k-wave simulation toolbox to simulate the propagation of photoacoustic signals. A three-dimensional acoustic medium space (including the perfectly matched layer PML for satisfying the boundary conditions of forward transmission) is established, and the size of each pixel is (at this time, the maximum striped spatial frequency supported by the model is ). The sample can be set as needed and placed in a three-dimensional grid. The surrounding medium is water, with the sound speed set to 1500 m / s and the density to 1000 kg / m 3 . The number of ultrasonic transducers is 1, the central frequency setting parameter is 50 MHz, and the bandwidth is 80%. The axial distance between the ultrasonic transducer and the plane where the sample is located is . Cosine striped light with different spatial frequencies is generated according to the method of the present invention, and the Fourier spectrum is obtained using a circular sampling strategy. All simulations assume a homogeneous acoustic medium with no absorption or scattering of sound.

[0028] As Figure 4 shows, it is the reconstruction result of multi-scale photoacoustic microscopy of a single spherical absorber. Figure 4 In [], NPS represents the normalized power spectrum, and NPA represents the normalized photoacoustic amplitude. A spherical absorber with a diameter of 2 microns is selected as the illumination sample. When the spatial frequency of the illuminated striped light is Figure 4 in (a) of [], that is, within the allowable frequency range of the system, most of the Fourier spectrum coefficients are obtained. At this time, since the high-frequency information reflecting the detailed information is obtained, the reconstructed image is relatively clear; as Figure 4 in (b) of [] shows, the contour of the reconstructed sphere is relatively clear. When the spatial frequency of the illuminated striped light isFigure 4 As shown in (c), at this time, due to the failure to acquire some high-frequency information, image details are lost; as Figure 4 shown in (d), the edges of the reconstructed image become less sharp. Further, when the spatial frequency of the illuminating fringe light is Figure 4 as shown in (e), at this time, the obtained Fourier coefficients are basically low-frequency coefficients, the details of the reconstructed image are insufficient, and the lateral resolution is low. As Figure 4 shown in (f), the Bessel sidelobes are obvious (indicated by the white arrow). When the spatial frequency of the illuminating fringe light is Figure 4 as shown in (g), at this time, only low-frequency Fourier coefficients are obtained, and the resolution of the reconstructed small ball is very poor. As Figure 4 shown in (h), at this time, the Bessel sidelobes are very obvious (indicated by the white arrow).

[0029] For quantitative analysis of the lateral resolution, take the full width at half maximum (FWHM) of the photoacoustic signal distribution at the position of the white dashed line in Figure 4 . As Figure 4 shown in (i), the changes in the lateral resolution are respectively shown when the spatial frequency of the illuminating fringe light is different. It can be seen that when , , , are Figure 4 as shown in (j), the corresponding measured lateral resolutions are respectively , , , ; the measured values are in agreement with the theoretical values, and the appropriate lateral resolution can be selected by quantitative analysis of these data for the lateral resolution.

[0030] As Figure 5 shown, the reconstruction results of multi-scale photoacoustic microscopy of multiple small ball absorbers are Figure 5 in which NPS represents the normalized power spectrum and NPA represents the normalized photoacoustic amplitude. Four randomly distributed small ball absorbers with a diameter of 2 microns are selected as the illuminating samples. As Figure 5 shown in (a), when the spatial frequency of the illuminating fringe light is Figure 5 , due to the high resolution, as Figure 5 shown in (b), the contours of the reconstructed small balls are relatively clear, and the 4 small balls can be clearly resolved. As shown in (c), when the spatial frequency of the illuminating fringe light Figure 5 isFigure 5 As shown in (e), when the spatial frequency of the illuminated fringe light is such that, at this time, the lateral resolution deteriorates further by a factor of 2, reaching the order of a dozen micrometers. As shown in Figure 5 (f), the 4 small balls can hardly be distinguished anymore.

[0031] The above details can be more clearly obtained from the locally magnified Figure 5 (g) - Figure 5 (i). Figure 5 (j) in Figure 5 (g) - Figure 5 (i) is the signal distribution at the position of the white dashed line. The change in resolution can also be clearly seen.

[0032] As Figure 6 shown is the result graph of reconstructing blood vessels using a reconstruction algorithm through the k-Wave: MATLAB simulation toolbox. Figure 6 In it, NPS represents the normalized power spectrum, and NPA represents the normalized photoacoustic amplitude. Virtual blood vessels are used for large-scale multi-scale imaging. The simulation parameters are consistent with the performance test experiment. Figure 6 (a) - Figure 6 (c) in are the Fourier spectra obtained when the spatial frequency of the illuminated fringe light , , is obtained. The corresponding reconstructed images are as shown in Figure 6 (d) - Figure 6 (f). It can be clearly seen from the figure that as more high-frequency information is obtained, the blood vessels are shown more clearly, and the blood vessel edges are more and more Rayleigh, reflecting the improvement of the resolution. Figure 6 (g) - Figure 6 (i) in Figure 6 (b) - Figure 6 (f) is the locally magnified view of the blood vessels in the white dashed box. Analyzing the width of the blood vessels at the white dashed line, as shown in Figure 6 (j), it can be seen that when the spatial frequency of the illuminated fringe light , , is such that, the measured widths are respectively , , ; the change in blood vessel width is consistent with the change in resolution. It shows that compared with the traditional mechanical moving scanning method and the microprobe scanning method, the method of the present invention has a reconstruction result closer to the labeled image for the image, that is, the imaging effect of the method of the present invention is better than that of the traditional imaging method.

[0033] The above are only the preferred embodiments of the present invention, and are not intended to limit the present invention in other forms. Any person skilled in the art may use the technical content disclosed above to make changes or modifications into equivalent embodiments with equivalent changes. However, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the technical solution content of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A multi-scale photoacoustic microscopy imaging method based on single pixel imaging, characterized in that: The following steps are involved: Step S1, generating a series of stripe lights with different spatial frequencies in the XY two-dimensional plane to illuminate the sample through a spatial light modulator, applying a phase shift to the stripe lights of each spatial frequency using a phase shift method, and collecting photoacoustic signals under each phase shift using a single ultrasonic probe; Step S2, calculating the Fourier spectrum coefficients according to the photoacoustic signal at each spatial frequency, introducing the impulse response function of the ultrasound probe to correct the Fourier spectrum coefficients, and then reconstructing the photoacoustic image through inverse Fourier transform based on the corrected Fourier spectrum coefficients based on a simplified point spread function; Step S3, controlling the spatial frequency of the stripe light to control the acquisition of Fourier spectrum coefficients, thereby achieving multi-scale imaging.

2. The multi-scale photoacoustic microscopy imaging method based on single pixel imaging according to claim 1, characterized in that: In step S1, a series of stripe lights with different spatial frequencies are generated in the XY two-dimensional plane by a spatial light modulator to illuminate the sample. and Phase Stripe light, intensity distribution of stripe light It is expressed as: ； In the above formula, x , y , z are the positions of the illumination samples. xyz Axis coordinates; The phase shift method is used to apply a phase shift to the stripe light of each spatial frequency, and a single ultrasonic probe is used to collect the photoacoustic signal under each phase shift. It is expressed as: ； In the above formula, is the light absorption distribution, It is the area illuminated by the stripe light.

3. The multi-scale photoacoustic microscopy imaging method based on single pixel imaging according to claim 2, characterized in that: The calculation method of the Fourier spectrum coefficient in step S2 is as follows: The phase shift method is used to apply a phase shift to the stripe light of each spatial frequency, and a single ultrasonic probe is used to collect and obtain photoacoustic signals of four phases, which are: 、 、 、 ； The light absorption distribution Fourier spectrum coefficient corresponding to the spatial frequency It is expressed as: ； In the above formula, j Is an imaginary unit.

4. The multi-scale photoacoustic microscopy imaging method based on single pixel imaging according to claim 3, characterized in that: In step S2, the impulse response function of the ultrasonic probe is introduced to correct the Fourier spectrum coefficients. It is expressed as: ； In the above formula, β and is a nonlinear parameter that controls the attenuation speed of the photoacoustic signal. β and ; j is an imaginary unit; Photoacoustic signal k The amplitude of the resonance peak; Photoacoustic signal k The attenuation coefficient of the resonance peak; Photoacoustic signal k The time delay of a resonance peak; Photoacoustic signal k The phase of the resonance peak; Corrected Fourier spectrum coefficients It is expressed as: ； In the above formula, is the Wiener filter weight function, which is used to suppress high-frequency noise amplification; ； In the above formula, , is the conjugate impulse response function; is an empirical constant, , is the signal-to-noise ratio at the spatial frequency point.

5. The multi-scale photoacoustic microscopy imaging method based on single pixel imaging according to claim 4, characterized in that: In step S2, the photoacoustic image is reconstructed by inverse Fourier transform based on the modified Fourier spectrum coefficients and the simplified point spread function. It is expressed as: ； In the above formula, This means that in actual imaging systems, the spatial frequency allowed by the optical system is limited. For a given maximum spatial frequency of fringes, for a given objective with a finite numerical aperture, ,in is the cut-off frequency of the objective lens, is the numerical aperture of the objective lens, is the wavelength of light; exp[*] is the exponential function; i is the imaginary unit; cc is the constant factor; Ignoring constant terms and constant factors, the point spread function is further simplified in polar coordinates to: ； Photoacoustic image reconstructed by inverse Fourier transform based on simplified spread function PSF It is expressed as: 。 6. The multi-scale photoacoustic microscopy imaging method based on single pixel imaging according to claim 1, characterized in that: The control of the spatial frequency of the stripe light in step S3 is achieved by adjusting the wavelength of the light source, the stripe period and the parameters of the projection system, and the Fourier spectrum coefficients are obtained based on a circular sampling strategy.

7. The multi-scale photoacoustic microscopy imaging method based on single pixel imaging according to claim 6, characterized in that: The circular sampling strategy defines a circular sampling area in the frequency domain, the radius of which corresponds to the required spatial frequency range; and obtains Fourier spectrum coefficients by uniformly sampling in the circular area.

Images