K domain homogenization and dispersion compensation method based on SD-OCT technology

Abstract

The invention belongs to the technical field of optical coherence tomography, and relates to a K-domain homogenization and dispersion compensation method based on an SD-OCT (Secure Digital Optical Coherence Tomography) technology, which comprises the following steps: acquiring a first spectral signal; using the first spectral signal to generate a calibration index array K-MAP for K domain homogenization; acquiring a second spectral signal; based on the calibration index array K-MAP, resampling the second spectral signal by using a linear interpolation method to obtain a resampled spectral signal; acquiring a first phase signal of the resampling spectral signal, performing first-order polynomial fitting on the first phase signal and the equivalent wave number to obtain a second phase signal, and constructing a complex compensation item according to a difference value between the first phase signal and the second phase signal; and acquiring a third spectral signal of the to-be-detected sample, re-sampling the third spectral signal by adopting a calibration index array K-MAP and a linear interpolation method to obtain a K-domain homogenized signal, and analyzing the K-domain homogenized signal to obtain an analytic signal of the third spectral signal. And real-time processing of acquired signals is realized.

Description

Technical Field

[0001] The present application belongs to the field of optical coherence tomography technology, and specifically is a K-domain homogenization and dispersion compensation method based on SD-OCT technology. Background Art

[0002] OCT technology (Optical Coherence Tomography, referred to as OCT, optical coherence tomography) can achieve real-time, high-resolution, three-dimensional imaging of the internal microstructure of biological tissues, and has been widely used in clinical diagnosis and treatment and basic scientific research. Among them, SD-OCT technology (Spectral Domain Optical Coherence Tomography, Chinese abbreviation: spectral domain optical coherence tomography) has significant advantages in signal-to-noise ratio, sensitivity, and imaging speed, and is the most widely used OCT imaging technology. In SD-OCT imaging technology, the dispersion mismatch problem will affect the imaging quality from spectral data to the final image reconstruction process, so dispersion compensation is required, and the necessary pre-step of dispersion compensation is to perform K-domain homogenization on the signal. Therefore, it is particularly important to find an easy and accurate K-domain homogenization and dispersion compensation method.

[0003] Existing dispersion compensation methods for SD-OCT can be divided into hardware methods and software methods. Hardware-based dispersion compensation is difficult to achieve perfect, and compensation is often performed in post-processing, which increases the cost of hardware construction. It has been proven that the resolution of bandwidth limitations can be restored by numerically compensating for dispersion, thus achieving the need for hardware-based dispersion balance. Software algorithm methods such as polynomial fitting compensation and segmented compensation require precise wavelength pre-calibration of the spectrometer to complete the signal from wavelength space ( The process of rescaling the measured signal from the measured sample (in the K-space) to the wavenumber space (in the K-space) is complex, but the calibration process is complicated and the accuracy of the calibration result is difficult to guarantee. The traditional method directly processes the measured sample signal, which takes a long time and cannot guarantee real-time performance. The deep learning method can reduce the processing time by removing the dispersion mismatch phenomenon, but it requires excellent dispersion-free SD-OCT data as the true value for preliminary model training, which is also difficult to obtain. Summary of the invention

[0004] The embodiment of the present application provides a K-domain homogenization and dispersion compensation method based on SD-OCT technology, which at least solves the problem that the current method requires pre-calibration and takes a long time to process the sample signal.

[0005] The technical solution of the embodiment of the present application is as follows: The present application provides a K-domain homogenization and dispersion compensation method based on SD-OCT technology, including: Acquire a first spectral signal, where the first spectral signal is a spectral signal of a calibration sample collected by the SD-OCT system; Using the first spectral signal to generate a calibration index array K-MAP for K-domain homogenization; Acquire a second spectral signal, where the second spectral signal is a spectral signal of a reflector collected by the SD-OCT system; Based on the calibration index array K-MAP, a linear interpolation method is used to resample the second spectral signal to obtain a resampled spectral signal; Acquire a first phase signal of the resampled spectral signal, perform a first-order polynomial fitting on the first phase signal and the equivalent wave number to obtain a second phase signal, and construct a complex compensation term according to the difference between the first phase signal and the second phase signal; A third spectral signal of the sample to be tested is obtained, and the third spectral signal is resampled using the calibration index array K-MAP and linear interpolation method to obtain a K-domain homogenized signal, the K-domain homogenized signal is analyzed to obtain an analytical signal of the third spectral signal, and the analytical signal of the third spectral signal is multiplied by the complex compensation term to obtain an analytical signal after dispersion compensation.

[0006] In one embodiment, using the first spectral signal to generate a calibration index array K-MAP for K-domain homogenization includes: Converting the first spectrum signal into a Fourier domain signal, setting a DC term and a conjugate image signal in the Fourier domain signal to zero and removing them, to obtain a Fourier domain signal of the first spectrum signal; The Fourier domain signal of the first spectral signal is sequentially subjected to eight-fold oversampling, zero filling, and inverse fast Fourier transformation to obtain an analytical signal of the first spectral signal, a phase signal of the analytical signal of the first spectral signal is extracted, and the phase signal of the first spectral signal is unwound and restored to obtain a phase signal of the first spectral signal; Perform linear interpolation on the phase signal of the first spectral signal to obtain a linearized phase signal, downsample the linearized phase signal eight times at equal intervals within the value range to obtain a new value range, use the value of the new value range as the query value, and use linear interpolation to calculate the original index in the phase signal of the first spectral signal corresponding to the value of the new value range to generate a calibration index array K-MAP for K-domain homogenization.

[0007] In one embodiment, based on the calibration index array K-MAP, a linear interpolation method is used to resample the second spectral signal to obtain a resampled spectral signal, including: Determine the query point according to the calibration index array K-MAP; The interpolation value of the second spectral signal at the query point is calculated using a linear interpolation method to form a resampled spectral signal.

[0008] In one embodiment, constructing a complex compensation term according to the difference between the first phase signal and the second phase signal includes: The difference is used as a phase compensation factor; Use the phase compensation factor as the phase of the complex compensation term.

[0009] In one embodiment, extracting the phase signal of the analytical signal of the first spectrum signal comprises: calculating the inverse tangent of the ratio of the imaginary part term to the real part term of the analytical signal of the first spectrum signal, and calculating the calculated result in The interval unwrapping restores the phase signal of the first spectral signal. The phase difference between the two waves is Radians or 180 degrees.

[0010] In one embodiment, obtaining a first phase signal from a resampled spectral signal includes: performing an inverse fast Fourier transform on the resampled spectral signal to obtain an analytical signal of the resampled spectral signal, extracting a phase signal of the analytical signal of the resampled spectral signal, and unwrapping and restoring the phase signal to obtain the first phase signal; extracting the phase signal of the analytical signal of the resampled spectral signal, and unwrapping and restoring the first phase signal includes: calculating an inverse tangent of the ratio of the imaginary part to the real part of the analytical signal of the resampled spectral signal, and calculating the calculated result in Interval unwinding.

[0011] In one embodiment, the DC term and the conjugate image signal in the Fourier domain signal are set to zero and removed, including: setting the zero range to , is the number of imaging points.

[0012] Compared with the prior art, the present application has the following beneficial effects: The method of the embodiment of the present application realizes K-domain homogenization and dispersion compensation method measurement, and the specially made calibration sample and reflector respectively complete the two processes, thereby avoiding the mutual influence of K-domain non-uniformity and dispersion mismatch in the traditional method.

[0013] The method of the present application selects to calculate the index array K-MAP required for K-domain homogenization and combines it with the linear interpolation method to realize the K-domain homogenization process, and calculates the complex compensation term to realize the dispersion compensation process. By directly acting on ordinary samples with the index array K-MAP of the calibration sample and the calculated complex compensation term, real-time processing of the collected signal can be realized, avoiding the problem of long processing time for each data of the measured sample in the traditional method. At the same time, the signal is always processed at the equivalent wavenumber, avoiding the process of accurate wavelength pre-calibration of the spectrometer to calculate the actual wavenumber required by the classical dispersion compensation method, thereby reducing the amount of calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1A method flow chart of a K-domain homogenization and dispersion compensation method based on SD-OCT technology provided in an embodiment of the present application; Figure 2 A schematic diagram of the structure of the SD-OCT system on which the K-domain homogenization and dispersion compensation method based on the SD-OCT technology provided in the embodiment of the present application is based; Figure 3 A schematic diagram of the calibration sample structure provided in an embodiment of the present application; Figure 4 The embodiment of the present application provides a plane mirror as a sample to be tested, and measures the original spectral signal of the SD-OCT system, the spectral signal after K-domain homogenization, and a comparison diagram of the point spread function of the spectral signal after dispersion compensation. DETAILED DESCRIPTION

[0015] In order to make the purpose, technical solution and advantages of the present application clearer and more intuitive, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0016] The executor of the method provided in the embodiment of the present application is an electronic device, which may include but is not limited to a mobile terminal, a computer terminal, a server, a processor of an SD-OCT system, or a similar computing device. The electronic device may be independent of the SD-OCT system or may serve as a part of the SD-OCT system.

[0017] like Figure 2 As shown, the structural schematic diagram of the SD-OCT system used in the embodiment of the present application, the method of the present application is to realize K-domain homogenization and dispersion compensation for the SD-OCT data of the SD-OCT system, and the commonly used SD-OCT system includes an SLD light source 1, a beam splitter / fiber coupler 2, a reference arm, a sample arm and a spectrometer. The beam splitter / fiber coupler 2 divides the light beam into two outputs at a ratio of 50:50, one output to the reference arm and the other output to the sample arm. The SLD in the SLD light source 1 refers to a superluminescent diode. The reference arm includes a first polarizer 3, a first collimator 4, an aperture 5 and a reference mirror 6; the spectrometer includes a second collimator 7, a scattering grating 8, a focusing lens 9 and a linear array CCD camera 10; the sample arm includes a second polarizer 11, a third collimator 12, a two-dimensional scanning galvanometer 13, a scanning lens 14, and a sample to be tested 15 placed behind the scanning lens 14. The first polarizer 3 of the reference arm and the second polarizer 11 of the sample arm respectively control the polarization states of the two optical signals of the reference arm and the sample arm.

[0018] The light signal emitted by the SLD light source 1 is split into two beams by the beam splitter / fiber coupler 2 and enters the reference arm and the sample arm respectively. After passing through various optical devices and the reference mirror or the reflection of the sample to be measured, the light signal passes through the beam splitter / fiber coupler 2 and enters the spectrometer, passes through the second collimator 7, the scattering grating 8, and the focusing lens 9, and finally is collected by the linear array CCD camera 10 to obtain the interference signal of the output light of the SD-OCT system. The linear array CCD camera 10 refers to a camera that uses a linear array charge-coupled device (CCD) as a photosensitive element.

[0019] The optical signal emitted by the SLD light source 1 is divided into two beams by the beam splitter / fiber coupler 2 and enters the reference arm and the sample arm respectively. After passing through various optical devices and the reference mirror or being reflected by the sample to be measured, the two beams pass through the beam splitter / fiber coupler 2 and enter the spectrometer, which specifically includes: the light beam of the reference arm passes through the first polarizer 3, the first collimator 4, the aperture 5 to the reference mirror 6 in sequence, and the reflected light of the reference mirror 6 returns to the beam splitter / fiber coupler 2 along the optical path of the reference arm and enters the spectrometer.

[0020] The light beam of the sample arm passes through the second polarizer 11, the third collimator 12, the two-dimensional scanning galvanometer 13, the scanning lens 14 to the sample 15 to be tested. The reflected light of the sample to be tested returns to the beam splitter / fiber coupler 2 along the optical path of the sample arm and enters the spectrometer.

[0021] The sample to be measured 15 is replaced with a calibration sample or a reflector according to the measurement requirements to obtain the signal of the calibration sample.

[0022] The calibration sample needs to have only two surfaces of the same material that reflect light of different optical paths and form an interference signal. The ideal surface is the upper and lower surfaces of a thin glass sheet. The material is the same but the optical paths are different. The optical path difference is the thickness of the glass sheet. A sufficiently thick gasket under the glass sheet ensures that no other interference signals will be measured. Glass sheets can include: soda-lime glass, borosilicate glass, quartz glass, tempered glass, low-iron glass, etc.

[0023] In one embodiment, Figure 3 The figure shows a schematic diagram of the calibration sample structure used, including a cover glass 21 made of glass, a first gasket 22, and a second gasket 23, wherein the first gasket 22 and the second gasket 23 have the same structure and are arranged in parallel, and the cover glass 21 is perpendicular to the ends of the first gasket 22 and the second gasket 23. The thickness of the cover glass 21 is 0.35 mm or less than 0.35 mm, and 0.35 mm is the thickness required by the embodiment of the present application. The first gasket 22 and the second gasket 23 are of the same thickness and are large enough. The large enough here refers to the range of the SD-OCT system as a reference, which needs to exceed the range of the SD-OCT system to ensure that no other interference signals are measured during measurement.

[0024] The K-domain homogenization and dispersion compensation method based on SD-OCT technology provided in the embodiment of the present application is as follows Figure 2 The measurements acquired by the SD-OCT system shown are as follows Figure 3 The signals of the calibration sample and the reflector shown are the processing objects of this application.

[0025] The overall idea of ​​the embodiments of the present application is to obtain a K-domain uniformized calibration index array K-MAP through processing of a signal sampled by a calibration sample with the help of an SD-OCT system, and to obtain a complex compensation item of the phase signal through a reflector based on the signal sampled by the SD-OCT system and the calibration index array K-MAP obtained from the calibration sample, and to use the calibration index array K-MAP and the complex compensation item for dispersion compensation of the sample to be tested.

[0026] The above overall idea realizes real-time processing of the collected signal of the sample to be tested, avoiding the problem of long processing time of each data of the measured sample in the traditional method. At the same time, the signal is always processed at the equivalent wave number, avoiding the process of accurate wavelength pre-calibration of the spectrometer to calculate the actual wave number required by the classical dispersion compensation method.

[0027] Figure 1 A flowchart of a K-domain homogenization and dispersion compensation method based on SD-OCT technology provided in an embodiment of the present application, see Figure 1 As shown, the method includes: S1 acquires a first spectral signal, where the first spectral signal is a spectral signal of a calibration sample collected by the SD-OCT system; After the SD-OCT system is built, the number of imaging points of the linear array CCD camera is fixed to a value, for example, the number of imaging points is 2048, and the Figure 1The SD-OCT system replaces the sample 15 to be tested with the calibration sample. When measuring the calibration sample, the diaphragm 5 of the SD-OCT system should be closed so that the reference arm does not return a signal. In the subsequent step S3, the diaphragm 5 is first opened, and then the reflector is measured to obtain the second spectral signal. For the K-domain homogenization and dispersion compensation process of the SD-OCT system, the elimination of K-domain inhomogeneity and dispersion mismatch need to be performed separately. The signal collected during the K-domain homogenization process cannot contain the influence of the system dispersion mismatch, and the system dispersion mismatch is caused by the mismatch between the reference arm and the sample arm. Here, the signal of the self-interference of the sample arm is collected, which must not be affected by the mismatch between the two arms. The interference signal of the upper surface reflected light and the lower surface reflected light of the cover glass 1 can be collected by measuring the calibration sample, that is, the first spectral signal. The first spectral signal is not affected by the dispersion mismatch and can be used for the relevant calculation of the calibration index array K-MAP required for K-domain homogenization. The reason why the thickness of the first gasket 2 and the second gasket 3 exceeds the measurement range of the SD-OCT system is to ensure that the collected first spectral signal contains only the interference signal of the reflected light on the upper surface and the reflected light on the lower surface of the cover glass 1, and does not contain the interference signal between other planes.

[0028] The first spectral signal captured by the linear array CCD camera is calculated and the average reference spectrum is removed to obtain the result. Indicates that is the pixel, that is, the spectral signal represented by the pixel.

[0029] S2 uses the first spectral signal to generate a calibration index array K-MAP for K-domain homogenization; It can be understood that the first spectral signal is the signal after the average reference spectrum is removed in step S1, and K-MAP refers to the calibration index array used for K-domain homogenization. K-domain homogenization is to convert the spectral signal from the wavelength space to the wavenumber space and make the spectral signal uniformly distributed in the wavenumber space. This process can significantly improve the imaging quality, reduce artifacts, and optimize the axial resolution. The calibration index array K-MAP is an array, and each element in the calibration index array K-MAP represents the index of the corresponding original spectral data at a uniform wavenumber interval.

[0030] In one embodiment, the first spectrum signal is converted into a Fourier domain signal, and the DC term and the conjugate image signal in the Fourier domain signal are set to zero and removed to obtain a Fourier domain signal of the first spectrum signal; the zeroing range is set to , The number of imaging points is determined by the linear array CCD camera.

[0031] The Fourier domain signal of the first spectral signal is sequentially subjected to eight-fold oversampling, zero filling, and inverse fast Fourier transformation to obtain an analytical signal of the first spectral signal, a phase signal of the analytical signal of the first spectral signal is extracted, and the phase signal of the first spectral signal is unwound and restored to obtain a phase signal of the first spectral signal; Extracting the phase signal of the analytical signal of the first spectrum signal includes: calculating the inverse tangent of the ratio of the imaginary part term to the real part term of the analytical signal of the first spectrum signal, and calculating the calculated result in The interval unwrapping restores the phase signal of the first spectral signal. The phase difference between the two waves is Radians or 180 degrees, By unwrapping, the jumps in the phase signal can be eliminated and the real phase signal can be restored. The methods used include but are not limited to: first-order difference method, least square method, etc.

[0032] It can be understood that the first spectrum signal is converted into a Fourier domain signal through Fourier transform. The Fourier domain signal belongs to the frequency domain. The spectrum is expanded in the frequency domain through eight-fold oversampling and zero padding, so that the length of the spectrum is increased to 8 times the original, thereby improving the resolution of the signal.

[0033] The expanded signal is then converted to the wave number space through the inverse fast Fourier transform, and the analytical signal of the first spectral signal after the inverse fast Fourier transform is extracted in the wave number space, and the phase signal of the analytical signal of the first spectral signal is unwrapped and restored to obtain the phase signal of the first spectral signal.

[0034] In one embodiment, a linear interpolation method is performed on the phase signal of the first spectral signal to obtain a linearized phase signal, and the new value range is obtained by downsampling eight times at equal intervals within the value range of the linearized phase signal. The value of the new value range is used as a query value, and the original index in the phase signal of the first spectral signal corresponding to the value of the new value range is calculated using the linear interpolation method, thereby reducing the index value to 1 / 8 of the original value, and generating a calibration index array K-MAP for K-domain homogenization.

[0035] When calculating the calibration index array K-MAP required for K-domain homogenization, eight times oversampling zero padding is introduced to make the calculation of the calibration index array more precise and accurate, thereby improving the K-domain homogenization effect.

[0036] S3 acquires a second spectral signal, where the second spectral signal is a spectral signal of a reflector collected by the SD-OCT system; The reflector uses a single-layer reflective plane mirror, such as a plane reflector. Since there is no multi-layer reflective structure, the sample dispersion problem caused by the interference of reflection signals from different layers can be avoided; In the SD-OCT system, the sample arm and the reference arm need to be opened at the same time, and the aperture 5 in the reference arm needs to be opened to collect interference signals of the two signals in the sample arm and the reference arm.

[0037] It is understandable that the second spectrum signal still needs to be calculated and the average reference spectrum removed, and the result is recorded as , is pixels.

[0038] S4 resamples the second spectral signal by linear interpolation based on the calibration index array K-MAP to obtain a resampled spectral signal, where the resampled spectral signal refers to a spectral signal after the second spectral signal is homogenized; Since the collected second spectral signal is a non-uniform spectral signal, the second spectral signal is resampled to a uniform wavenumber space through the calibration index array K-MAP, and the query point is determined according to the calibration index array K-MAP; the linear interpolation method is used to calculate the interpolation value of the second spectral signal at the query point to form a resampled spectral signal.

[0039] S5 obtains a first phase signal of the resampled spectral signal, performs a first-order polynomial fitting on the first phase signal and the equivalent wave number to obtain a second phase signal, and constructs a complex compensation term according to the difference between the first phase signal and the second phase signal; The resampled spectral signal is now in a uniform wavenumber space as a homogenized spectral signal, which can be expressed as , is the equivalent wave number.

[0040] The resampled spectrum signal is subjected to fast Fourier transform to obtain the Fourier domain signal, and the DC term and conjugate image signal are set to zero and removed. The zero range is set to The obtained spectrum is expressed as ; is the equivalent optical path difference; The first phase signal is obtained from the resampled spectral signal, including: performing an inverse fast Fourier transform on the resampled spectral signal to obtain an analytical signal of the resampled spectral signal, extracting a phase signal of the analytical signal of the resampled spectral signal, and unwrapping and restoring the phase signal to obtain the first phase signal; extracting the phase signal of the analytical signal of the resampled spectral signal, and unwrapping and restoring the first phase signal includes: calculating the inverse tangent of the ratio of the imaginary part to the real part of the analytical signal of the resampled spectral signal, and calculating the calculated result in Interval unwinding.

[0041] The Fourier domain signal is obtained by performing fast Fourier transform on the resampled spectral signal, and the resampled spectral signal is converted from the wavenumber space to the frequency domain. The DC term and the conjugate image part are removed by zeroing in the frequency domain.

[0042] The Fourier domain signal after removing the DC term and conjugate image needs to be converted into the wave number space by fast Fourier inverse transform to obtain the analytical signal of the resampled spectral signal. The phase signal of the analytical signal of the resampled spectral signal is extracted in the wave number space, and the first phase signal is obtained by unwrapping and restoring, which is recorded as , the first phase signal and the equivalent wave number Perform a first-order polynomial fitting to obtain the second phase signal after fitting, recorded as , and Subtract the phase compensation factor to get the dispersion compensation , and then construct the complex compensation term: ;in, is an imaginary unit, is a complex compensation term, the spectrum It belongs to the frequency domain signal; After removing the DC term and the conjugate image part, the Fourier domain signal needs to be converted into the wave number space by fast Fourier inverse transform. There are discontinuities in the phase signal extracted in the wave number space. It is converted into a continuous phase signal, i.e., the first phase signal, by unwrapping, which is used to eliminate phase jumps and restore the real phase information. The unwrapping method may include but is not limited to the first-order difference method or the least squares method.

[0043] The first phase signal and the equivalent wave number The first-order polynomial fitting is to convert the equivalent wave number As a polynomial variable, a first-order polynomial is formed through two fitting parameters. It is expressed as: ,in, and is the fitting parameter, Refers to the phase after fitting, that is, the second phase signal.

[0044] The calibration index array K-MAP and the complex compensation terms are obtained through the above steps.

[0045] It can be understood that there is no particular order in which to obtain the calibration index array K-MAP and the complex compensation items. After the obtained calibration index array K-MAP and the complex compensation items are stored, they are used to perform dispersion compensation on the collected spectral signals of the samples to be tested.

[0046] The following content is to use the calibration index array K-MAP and the complex compensation term to perform dispersion compensation on the spectrum signal of the sample to be tested. The calibration index array K-MAP and the complex compensation term are the key elements for dispersion compensation.

[0047] S6, obtaining a third spectral signal of the sample to be tested, resampling the third spectral signal using a calibration index array K-MAP and a linear interpolation method to obtain a K-domain homogenized signal, analyzing the K-domain homogenized signal to obtain an analytical signal of the third spectral signal, and multiplying the analytical signal of the third spectral signal by a complex compensation term to obtain a dispersion-compensated analytical signal. The dispersion-compensated analytical signal can be directly used for image reconstruction.

[0048] For the spectral signals directly acquired by the SD-OCT system, the average reference spectrum needs to be removed. For the third spectral signal without the average reference spectrum, the calibration index array K-MAP and linear interpolation method are used for resampling in the same way as the processing of the second spectral signal in S3 to obtain the K-domain homogenized signal.

[0049] In one embodiment, the K-domain homogenized signal in S5 is analyzed to obtain the analyzed signal of the third spectral signal, including: performing a fast Fourier transform on the K-domain homogenized signal, setting the DC term and the conjugate image signal therein to zero and removing the zeroing range. , and then perform a fast inverse Fourier transform to obtain an analytical signal of the third spectral signal, and multiply the analytical signal of the third spectral signal by the complex compensation term to obtain an analytical signal after dispersion compensation. The analytical signal after dispersion compensation can be directly used for image reconstruction.

[0050] In order to verify the effect of the method provided in the embodiment of the present application, a plane mirror is used as a sample to be tested, and the point spread function (PSF) of the SD-OCT system is measured and the half-maximum full width is calculated. This value can reflect the axial resolution of the SD-OCT system to verify the imaging effect. The point spread function of the SD-OCT system before and after K-domain homogenization and before and after dispersion compensation is compared, as shown in FIG. Figure 4 As shown, Amp in the figure is the amplitude intensity of the point spread function of the SD-OCT system, and the horizontal coordinate Z (N) represents the optical path difference that changes with the number of coordinate points. The point spread functions of the original spectral signal, the spectral signal after K-domain homogenization, and the spectral signal after dispersion compensation are compared, where the point spread function of the spectral signal after dispersion compensation is the result obtained by the method of the embodiment of the present application. It can be calculated that the half-height full width of the point spread function of the original spectral signal, the spectral signal after K-domain homogenization, and the spectral signal after dispersion compensation are 16.7, 5.5, and 2.1, respectively. It can be seen that the method provided in the embodiment of the present application significantly eliminates the problems of K-domain inhomogeneity and dispersion mismatch, and improves the axial resolution of the SD-OCT system detection.

[0051] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A K-domain homogenization and dispersion compensation method based on SD-OCT technology, characterized in that: include: Acquire a first spectral signal, where the first spectral signal is a spectral signal of a calibration sample collected by the SD-OCT system; Using the first spectral signal to generate a calibration index array K-MAP for K-domain homogenization; Acquire a second spectral signal, where the second spectral signal is a spectral signal of a reflector collected by the SD-OCT system; Based on the calibration index array K-MAP, a linear interpolation method is used to resample the second spectral signal to obtain a resampled spectral signal; Acquire a first phase signal of the resampled spectral signal, perform a first-order polynomial fitting on the first phase signal and the equivalent wave number to obtain a second phase signal, and construct a complex compensation term according to the difference between the first phase signal and the second phase signal; A third spectral signal of the sample to be tested is obtained, and the third spectral signal is resampled using the calibration index array K-MAP and linear interpolation method to obtain a K-domain homogenized signal, the K-domain homogenized signal is analyzed to obtain an analytical signal of the third spectral signal, and the analytical signal of the third spectral signal is multiplied by the complex compensation term to obtain an analytical signal after dispersion compensation.

2. The K-domain homogenization and dispersion compensation method based on SD-OCT technology according to claim 1, characterized in that: The first spectral signal is used to generate a calibration index array K-MAP for K-domain homogenization, comprising: Converting the first spectrum signal into a Fourier domain signal, setting a DC term and a conjugate image signal in the Fourier domain signal to zero and removing them, to obtain a Fourier domain signal of the first spectrum signal; The Fourier domain signal of the first spectral signal is sequentially subjected to eight-fold oversampling, zero filling, and inverse fast Fourier transformation to obtain an analytical signal of the first spectral signal, a phase signal of the analytical signal of the first spectral signal is extracted, and the phase signal of the first spectral signal is unwound and restored to obtain a phase signal of the first spectral signal; Perform linear interpolation on the phase signal of the first spectral signal to obtain a linearized phase signal, downsample the linearized phase signal eight times at equal intervals within the value range to obtain a new value range, use the value of the new value range as the query value, and use linear interpolation to calculate the original index in the phase signal of the first spectral signal corresponding to the value of the new value range to generate a calibration index array K-MAP for K-domain homogenization.

3. The K-domain homogenization and dispersion compensation method based on SD-OCT technology according to claim 1, characterized in that: Based on the calibration index array K-MAP, the second spectral signal is resampled by a linear interpolation method to obtain a resampled spectral signal, including: Determine the query point according to the calibration index array K-MAP; The interpolation value of the second spectral signal at the query point is calculated using a linear interpolation method to form a resampled spectral signal.

4. The K-domain homogenization and dispersion compensation method based on SD-OCT technology according to claim 1, characterized in that: A complex compensation term is constructed according to the difference between the first phase signal and the second phase signal, including: The difference is used as a phase compensation factor; Use the phase compensation factor as the phase of the complex compensation term.

5. The K-domain homogenization and dispersion compensation method based on SD-OCT technology according to claim 2, characterized in that: Extracting the phase signal of the analytical signal of the first spectrum signal includes: calculating the inverse tangent of the ratio of the imaginary part term to the real part term of the analytical signal of the first spectrum signal, and calculating the calculated result in The interval unwrapping restores the phase signal of the first spectral signal. The phase difference between the two waves is Radians or 180 degrees.

6. The K-domain homogenization and dispersion compensation method based on SD-OCT technology according to claim 2, characterized in that: The first phase signal is obtained from the resampled spectral signal, including: performing an inverse fast Fourier transform on the resampled spectral signal to obtain an analytical signal of the resampled spectral signal, extracting a phase signal of the analytical signal of the resampled spectral signal, and unwrapping and restoring the phase signal to obtain the first phase signal; extracting the phase signal of the analytical signal of the resampled spectral signal, and unwrapping and restoring the first phase signal includes: calculating the inverse tangent of the ratio of the imaginary part to the real part of the analytical signal of the resampled spectral signal, and calculating the calculated result in Interval unwinding.

7. The K-domain homogenization and dispersion compensation method based on SD-OCT technology according to claim 2, characterized in that: Zero out the DC term and conjugate image signal in the Fourier domain signal, including: Set the zeroing range to , is the number of imaging points.

Images