Automatic microfluid flow measuring method based on spectral domain optical coherence tomography technology

Abstract

The invention discloses an automatic microfluid flow measurement method based on a spectral domain optical coherence tomography technology, which is characterized in that a Doppler frequency domain OCT technology is utilized, flow information is acquired by measuring Doppler frequency movement in microfluid, and how to measure a Doppler angle through a biplane method is described in detail. Through a method for eliminating random phase noise of a pipeline wall, the center position of the microfluid pipeline is determined by using an OCT image, and the reliability and accuracy of the method are verified through experiments. Experimental results show that the method can effectively measure the flow of the microfluid and has high precision and stability. The research provides a new solution for the automatic detection of the flow of the microfluid, and is beneficial to improving the working efficiency of the microfluid research.

Description

Technical Field

[0001] The invention relates to microfluid detection and its application field, and in particular to a microfluid flow automatic measurement method based on spectral domain optical coherence tomography technology. Background Art

[0002] Microfluidic detection technology is a science that precisely controls, manipulates and detects fluids at a microscopic scale. It is based on microelectronics, micromechanics, bioengineering and nanotechnology, and involves multiple disciplines. The core of microfluidic technology is to construct a microfluidic channel system to achieve complex fluid manipulation functions, such as pumps, valves, mixers, filters and separators. These microfluidic devices are widely used in scientific experiments such as biochips, sample DNA preparation, liquid chromatography, PCR reactions and electrophoresis detection. They quantitatively analyze the flow process in microchannels, emphasize the precise manipulation of fluids at a microscale, and can be of great help to microscopic research.

[0003] In the existing technology, the measurement of microfluid flow rate is relatively common, using Doppler OCT and the dual-plane method to measure the Doppler angle α and then obtain the real flow rate of the fluid. However, there are relatively few means for automatic detection of microfluid flow. In some microscopic experimental requirements, the lack of flow parameters may restrict the application of Doppler OCT technology in microfluid research. Summary of the invention

[0004] Purpose of the invention:

[0005] In order to solve the problems raised in the background technology, the present invention provides a method for automatically measuring microfluid flow based on spectral domain optical coherence tomography technology, which realizes fully automatic detection of microchannel liquid flow through comprehensive use of OCT images and Doppler phase images, combined with a center position denoising screening process and an integrated automated computer program.

[0006] Technical solution:

[0007] The present invention discloses a method for automatically measuring microfluid flow rate based on spectral domain optical coherence tomography technology, the method comprising the following steps:

[0008] S1 uses an OCT device to scan the microfluid twice at different positions to obtain OCT structural images and Doppler phase images, and uses a dual-plane method to measure the Doppler angle between the detection beam and the microfluid;

[0009] S2 performs binarization processing on the acquired OCT structural image to obtain a binarized image of the microfluidic cross section, and adjusts the amplitude threshold of the binarization processing according to the background noise intensity of the OCT system;

[0010] S3 averages the values ​​of the binary image of the microfluidic channel in the vertical direction to obtain the center position of the microfluidic channel in the horizontal direction.

[0011] S4 takes the horizontal center position as the center, averages and smoothes the values ​​of the binary image of the microfluidic channel in the horizontal direction, and obtains the vertical center position of the microfluidic channel;

[0012] S5 performs a phase recovery operation on the Doppler phase image, taking the center coordinate of the microfluidic channel as the center of the circle and selecting a circle with a diameter slightly smaller than the inner diameter of the capillary as the boundary for phase recovery;

[0013] S6 sets the microfluid flow ellipse boundary, adds the phase signals within the boundary to obtain a sum value, divides the calculated average of the sum value by the cosine of the Doppler angle, and obtains the flow result of the microfluid to be measured.

[0014] Furthermore, the process of obtaining the Doppler angle in S1 is as follows:

[0015] The process of obtaining the Doppler angle described in S1 is as follows:

[0016] The OCT scanning planes are defined as the x and z planes. In order to measure the angle θ between the OCT beam and the pipeline, the OCT sampling beam scans the pipeline twice alternately between the two scanning planes S1 and S2 with a small displacement Δy. The coordinates of the pipeline center in plane S1 are expressed as (x1, z1), and the coordinates in plane S2 are expressed as (x2, z2). The angle θ is obtained by calculating the center position of the fluid on the two scanning planes. According to the flow direction in the microchannel, the Doppler angle α=θ or α=180-θ is obtained.

[0017] Furthermore, the amplitude threshold in S2 is obtained as follows:

[0018] An amplitude threshold Ath is used. For the image pixel at position (i, j), where j is the pixel number in the depth direction, if its OCT amplitude A(i, j)>Ath, the signal intensity B(i, j) of the binary image is set to 512; otherwise, B(i, j)=0. The threshold Ath is determined according to the background noise level of the OCT system. The area above the surface of the OCT structural image contains background noise information. In order to determine the background noise, 900 vertical sampling lines are horizontally averaged. The concave area at the lower left of the curve corresponds to the area above the surface of the OCT structural image, and its signal amplitude represents the background noise level. By averaging the signals from the 10th pixel to the 30th pixel in the concave area of ​​the curve, the average noise level Anoise=232.8 is obtained. According to the signal fluctuation of each vertical sampling line, the threshold Ath is selected as 400.0, which is about 1.72 times of Anoise.

[0019] Furthermore, the detailed steps of S3 are as follows:

[0020] For each vertical line in the binarized image of the microfluidic channel after binarization, the binarized signal B(i,j) is averaged in the vertical direction within the range of N pixels from the image surface Pi downward. The average result BVA(i) can be expressed as

[0021]

[0022] Among them, the position Pi of the upper surface of the binary image of the microfluidic channel is determined according to the amplitude signal A(i,j) of the OCT. The vertical signal average curve is the average value BVA(i) when N=270, where the horizontal coordinate axis is the image vertical line number i. There are two negative peaks in the BVA(i) curve, which appear at i=iL and i=iR respectively. The negative peak at iL corresponds to the dark area of ​​the left wall of the capillary, and the negative peak at iR corresponds to the right wall of the capillary. After obtaining iL and iR by searching the minimum value in the BVA(i) curve, the horizontal center of the capillary ic=(iL+iR) / 2 can be calculated, so that the horizontal center position xc of the microfluid is xc=dx*ic, where dx is the unit length corresponding to each pixel in the horizontal direction.

[0023] Furthermore, with the horizontal center position i c As the center, the binary signals B(i,j) of (2D+1) vertical lines are horizontally averaged to obtain the average result BHA(ic,j):

[0024]

[0025] Where D is the average range selected for horizontal averaging, that is, the horizontal center position i c Take D vertical lines on each side as the center and average them.

[0026] Using (2U+1) pixels, BHA(ic,j) is processed by moving average in the vertical direction, and the smoothed result Bsm(ic,j) is:

[0027]

[0028] Among them, U is the range used for vertical moving average. For a certain point, take U points above and below the point, and calculate the binary value B within this range. HA Take the average and use the average as the smoothing result of this point.

[0029] P c The vertical line i represents the surface of the binary image of the microfluidic channel. c At the position, the smoothing result of the vertical center of the microchannel when U=6 is Bsm (i), the horizontal axis is the pixel number j in the vertical direction. The smoothed curve has two negative peaks, which appear at j=j. u and j = j b At j u The negative peak at corresponds to the dark area on the upper part of the glass capillary wall. b The negative peak at corresponds to the bottom of the capillary wall. sm (i c , j) The minimum value in the curve is obtained at position j u and j b After that, the vertical center j of the capillary can be calculated c , i.e. j c =(j u +j b ) / 2, and obtain the center coordinate z of the capillary in the vertical direction c For z c =dz*j c , where dz is the unit length corresponding to each pixel in the depth direction.

[0030] Furthermore, the detailed steps of S5 are as follows: after determining the center position of the microfluidic channel, a circle O with a center of (ic, jc) and a diameter of d is used as a boundary to perform phase unwrapping on the Doppler phase image. The diameter d of the circle O should be smaller than the inner diameter of the capillary and should be close to the inner wall of the channel. For the pixels on the circle O, since the flow velocity close to the channel wall is low, no wrapped phase will appear. The phase value of the pixel on the circle can be used as a boundary condition. The upper and lower semicircles of the circle O are used as the upper and lower boundaries of the phase recovery, respectively. The Doppler phase of the pixel in the circle O can be expanded in the vertical direction from the upper semicircle to the lower semicircle to obtain the restored Doppler phase image.

[0031] Furthermore, in the vertical direction of the restored Doppler phase image, for each vertical line, the Doppler signals of 10 consecutive pixel points below the image surface are averaged, and the average value is subtracted from the phase signal of the entire vertical line to obtain the phase signal ψ(i, j) caused by liquid flow.

[0032] Furthermore, in order to calculate the volume flow rate of the microfluid, it is necessary to integrate the Doppler signal of the fluid in the flow channel. Due to the random phase noise in the capillary wall, the integration boundary is difficult to determine. In order to eliminate the influence of random phase noise on the microfluid flow calculation, the random phase noise in the capillary wall is reset to zero according to the OCT signal intensity A(i, j), and ψ(i, j) is converted to Γ(i, j). The specific method is as follows:

[0033] In the area outside the circle O in the Doppler phase image, if the OCT amplitude A(i,j)>Ath, then its phase value Γ(i,j)=ψ(i,j); otherwise, Γ(i,j)=0; for the area inside the circle O, the phase value remains unchanged, that is, Γ(i,j)=ψ(i,j), where the threshold Ath is the same as the threshold selected during image binarization. Through this method, the random phase noise on the pipe wall is replaced by zero value.

[0034] Furthermore, the detailed steps of S6 are as follows:

[0035] ①Set the microfluid flow ellipse boundary:

[0036] The ellipse centered at (ic, jc) represents the boundary of volume flow calculation. The lengths of the major and minor axes of the ellipse are |iR-iL| and |ju-jb|, respectively. The phase signals Γ(i, j) within the ellipse are added to obtain the sum σ.

[0037] ② Calculate the flow rate of the microfluid to be tested:

[0038] Multiple pairs of Doppler OCT images are collected each time the flow is measured. For the mth pair of images, the phase sums of the Doppler images collected in the S1 and S2 planes are σm1 and σm2, respectively. The flow Qm calculated using the mth pair of Doppler OCT images can be expressed as:

[0039]

[0040] Where θm is the angle calculated using the mth pair of images, βm is the angle between the OCT scanning plane and the plane perpendicular to the capillary axis, cosβm=Δy / R, and the flow result Q measured in one measurement is as follows:

[0041]

[0042] Where M is the total number of image pairs collected in one measurement.

[0043] Furthermore, the present invention also provides a computer storage medium having a computer program stored thereon, which, when executed by a processor, can implement the steps in the method for automatically measuring microfluid flow based on spectral domain optical coherence tomography technology in a non-contact manner.

[0044] Beneficial effects:

[0045] 1. The present invention starts with Doppler angle measurement and inspires microfluid flow measurement through existing flow velocity measurement methods, which effectively lowers the implementation threshold for technical personnel in this field and has a stronger adaptability prospect in various types of microfluid experiments.

[0046] 2. The present invention eliminates the influence of OCT signal intensity fluctuation and random phase noise of the pipeline wall through binarization processing and phase recovery, further improves the accuracy of microfluid flow measurement results, and ensures that the research results are controllable.

[0047] 3. The present invention adopts a non-contact automatic measurement method to reduce the interference of the method itself on the flow of microfluids and further improve the flow measurement accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is a schematic diagram of the measurement principle of microfluids by the spectral domain OCT device according to an embodiment of the present invention;

[0049] Figure 2 (a) is a structural image of a microfluid according to an embodiment of the present invention; (b) is a Doppler phase image of the microfluid;

[0050] Figure 3 The average value of the OCT amplitude in the horizontal direction of the embodiment of the present invention;

[0051] Figure 4 (a) Structural image of the sample fluid after binarization processing in the embodiment of the present invention; (b) Signal average result in the vertical direction B VA (i);

[0052] Figure 5 The smoothing result B for determining the vertical center of the microchannel in the embodiment of the present invention is sm (i);

[0053] Figure 6 (a) is a schematic diagram of the Doppler OCT phase recovery boundary according to an embodiment of the present invention; (b) is a Doppler phase diagram after phase recovery;

[0054] Figure 7 Determination of the integral boundary of microfluid flow calculation in an embodiment of the present invention;

[0055] Figure 8 These are images of microfluidic cross-sectional structures collected using a dual-plane method according to an embodiment of the present invention: (a) an image collected on the S1 plane; (b) an image collected on the S2 plane. DETAILED DESCRIPTION

[0056] The present application is further described in detail below in conjunction with the accompanying drawings. It is necessary to point out here that the following specific implementation methods are only used to further illustrate the present application and cannot be understood as limiting the scope of protection of the present application. Technical personnel in this field can make some non-essential improvements and adjustments to the present application based on the above application content.

[0057] The non-contact measurement of microfluidic flow can be used to study fluid dynamics in microchannels, which is of great significance for the theoretical and applied research of microfluidics. In Doppler frequency domain OCT, the detection light scattered by the moving fluid will produce a Doppler frequency shift Δν, which is proportional to the flow velocity V of the fluid.

[0058] Δν=-2nVcosα / λ0,

[0059] Where n is the refractive index of the flowing medium, α is the angle between the OCT detection light and the liquid flow direction, and λ0 is the central wavelength of the detection light. In the spectral domain OCT system, this frequency shift will produce a phase shift in the detected spectral interference signal. After the fast Fourier transform, the transformation result is a complex function F(z), which contains the amplitude A(z) and the phase Φ(z). By calculating the phase difference between two adjacent vertical sampling lines, φ(z) = Φ2(z)-Φ1(z), the Doppler frequency shift Δν = φ(z) / (2πT) can be obtained, where T is the time interval between two adjacent vertical sampling lines. It can be seen from formula (1) that in Doppler OCT, the frequency shift is caused by the velocity component parallel to the detection light, and the value actually measured by Doppler OCT is Vcosα. In order to obtain the actual flow velocity V, multiple measurements are required to determine the Doppler angle α.

[0060] The principle of using the dual-plane method to measure the Doppler angle α is as follows: Figure 1 As shown, Figure 1 In the frequency domain OCT system, the outgoing light from the sample arm is first collimated by a lens, and then driven by a scanning galvanometer to scan the microfluid flowing through the glass capillary. The glass capillary is immersed in a paper cup filled with a diluted emulsion. Figure 1 The coordinate system shown in the sample arm in the figure defines the OCT scanning plane as the x and z planes. In order to measure the angle θ between the OCT beam and the pipeline, the OCT sampling beam scans the pipeline twice alternately between the two scanning planes S1 and S2 with a small displacement Δy. Assuming that the coordinates of the center of the pipeline on these two planes are expressed as (x1, z1) (S1 plane) and (x2, z2) (S2 plane), the angle θ can be obtained by calculating the center position of the fluid on the two scanning planes. Then, according to the flow direction of the emulsion in the microchannel, the Doppler angle α = θ or α = 180-θ can be obtained.

[0061] Figure 2 The OCT and Doppler phase images of the microfluid in the glass capillary acquired by the spectral domain OCT system are shown in Figure 1. Figure 2 As shown in (a), it consists of 900 vertical sampling lines. The dark ring in the image represents the wall of the glass capillary. Diluted milk flows through the capillary, and the coordinates of the center of the pipe are (x c , z c). The glass capillary is immersed in diluted milk. Therefore, the bright signal around the capillary in the OCT image represents the milk solution. In this patent application, we refer to the upper surface of the milk shown in the OCT image as the "image surface". On the i-th vertical line, the depth position of the image surface is denoted by P i Indicates that Figure 2 Indicated by the vertical dashed line in (a). Figure 2 (b) is a Doppler image of the microfluidic system, which shows that the Doppler frequency signal in the pipe presents a negative frequency shift, and the white signal in the central area is caused by phase wrapping. Since there is no backscattering of light on the capillary glass wall, there is random phase noise in the capillary wall area in the image, forming a noise ring.

[0062] In order to determine the center coordinates of the microfluidic channel, the OCT structural image was first binarized. When binarizing the image, we used an amplitude threshold A th For an image pixel at position (i, j), where j is the pixel number in the depth direction, if its OCT amplitude A(i, j)>A th , the signal strength B(i, j) of the binary image is set to 512; otherwise, B(i, j) = 0. Threshold A th It is determined based on the background noise level of the OCT system. Figure 2 In the structural image shown in (a), the area above the image surface contains background noise information. To determine the background noise, 900 vertical sampling lines are horizontally averaged, and the results are shown in Figure 3 As shown in the curve. Figure 3 The concave area at the lower left of the middle curve corresponds to the area above the surface of the OCT structural image, and its signal amplitude represents the background noise level. The average noise level A is obtained by averaging the signals from the 10th pixel (j=10) to the 30th pixel (j=30) in the concave area of ​​the curve. noise =232.8. Considering the signal fluctuation of each vertical sampling line, this study selects the threshold A th is 400.0, about A noise 1.72 times of. Figure 4 (a) shows the image after binarization processing.

[0063] For each vertical line in the binary image, i In the range of N pixels below, the binary signal B(i,j) is averaged in the vertical direction, and the average result B VA (i) can be expressed as

[0064]

[0065] The position of the surface on the image is Pi It is determined based on the OCT amplitude signal A(i,j). Figure 4 The curve in (b) is the average value B when N = 270 VA (i), where the horizontal axis is the vertical line number i of the image. As can be seen from the figure, B VA (i) The curve has two negative peaks, appearing at i=i L and i=i R i L The negative peak at corresponds to the dark area on the left wall of the capillary, such as Figure 4 (a). Similarly, i R The negative peak at corresponds to the wall of the right side of the capillary. VA (i) The minimum value in the curve is i L and i R After that, the capillary horizontal center i can be calculated c =(i L +i R ) / 2, thus obtaining the horizontal center position x of the microfluid c For x c =dx*i c , where dx is the unit length corresponding to each pixel in the horizontal direction.

[0066] Next, with the horizontal center i c As the center, the binary signals B(i,j) of (2D+1) vertical lines are horizontally averaged to obtain the average result B HA (i c ,j):

[0067]

[0068] Among them, D is the average range selected when performing horizontal averaging, that is, the horizontal center position i c Take D vertical lines on each side as the center and average them.

[0069] Then, using (2U+1) pixels, we can HA (i c ,j) Perform moving average processing in the vertical direction, and the result after smoothing is B sm (i c ,j) is:

[0070]

[0071] Among them, U is the range used for vertical moving average. For a certain point, take U points above and below the point, and calculate the binary value B within this range. HA Take the average and use the average as the smoothing result of this point.

[0072] P c Indicates that the image surface is on the vertical line i c The location, such as Figure 4 (a) shown. Figure 5 The smoothing result B is shown when U=6 sm (i c ,j), the horizontal axis in the figure is the pixel number j in the vertical direction. Figure 5 It can be seen from the figure that there are two negative peaks in the smoothed curve, which appear at j=j. u and j = j b The negative peak at ju corresponds to Figure 4 (a) shows the dark area on the upper part of the glass capillary wall. b The negative peak at corresponds to the bottom of the capillary wall. Therefore, by searching for B sm (i c , j) The minimum value in the curve is obtained at position j u and j b After that, the vertical center j of the capillary can be calculated c , i.e. j c =(j u +j b ) / 2. Then, the vertical center coordinate z of the capillary can be obtained c For z c =dz*j c , where dz is the unit length corresponding to each pixel in the depth direction.

[0073] When Doppler OCT is used to measure fluid flow, the wrapped Doppler phase signal needs to be effectively recovered. Since there is random phase noise in the pipe wall, it is impossible to perform phase unwrapping from the image surface downward. Therefore, after determining the center of the microfluidic pipe, the center of the circle is (i c ,j c ), a circle O with a diameter of d is used as the boundary for phase unwrapping, such as Figure 6 (a) is shown. The diameter d of circle O should be smaller than the inner diameter of the capillary and close to the inner wall of the pipe. In this way, for the pixel points on circle O, since the flow velocity close to the pipe wall is low, no wrapped phase will appear, so the phase value of the pixel points on the circle can be used as a boundary condition. The upper and lower semicircles of circle O are used as the upper and lower boundaries of phase recovery, respectively, as shown in Figure 6 As shown in the direction of the arrow in (a), the Doppler phase of the pixel points within the circle O can be expanded in the vertical direction from the upper semicircle to the lower semicircle. Figure 6 (b) shows the restored Doppler phase image.

[0074] The Doppler information contains motion noise from the OCT system. If not corrected, the Doppler background noise generated by these movements will affect the measurement results. Therefore, in the vertical direction of the image, for each vertical line, the Doppler signals of 10 consecutive pixels below the image surface are averaged. Then, this average value is subtracted from the phase signal of the entire vertical line to obtain the phase signal ψ(i,j) caused by the liquid flow.

[0075] In order to calculate the volume flow of microfluids, it is necessary to integrate the Doppler signal of the fluid in the flow channel. Due to the presence of random phase noise in the capillary wall, the integration boundary is difficult to determine. In order to eliminate the influence of random phase noise on the calculation of microfluid flow, the present invention resets the random phase noise in the capillary wall to zero according to the OCT signal intensity A(i, j), and converts ψ(i, j) to Γ(i, j). The specific method is that in the area outside the circle O in the phase image, if the OCT amplitude A(i, j)>A th , then its phase value Γ(i,j)=ψ(i,j); otherwise, Γ(i,j)=0; for the area inside the circle O, the phase value remains unchanged, that is, Γ(i,j)=ψ(i,j). Figure 7 The image of Γ(i,j) after processing is shown, where the threshold A th The choice of is the same as the threshold selected during image binarization. By this method, the random phase noise on the pipe wall is replaced by zero value. Figure 7 China c ,j c ) represents the boundary of volume flow calculation. The lengths of the major and minor axes of the ellipse are |i R -i L | and |j u -j b |. The phase signals Γ(i,j) within the elliptical region are added to obtain the sum value σ.

[0076] Multiple pairs of Doppler OCT images are collected for each flow measurement. For the mth pair of images, it is assumed that the phase sum results of the Doppler images collected in the S1 and S2 planes are σ m1 and σ m2 , then the flow Q calculated using the mth pair of Doppler OCT images m It can be expressed as

[0077]

[0078] Among them, θ m is the angle calculated using the mth pair of images, β m is the angle between the OCT scanning plane and the plane perpendicular to the capillary axis, cosβ m =Δy / R. Then the flow rate result Q measured in one measurement is

[0079]

[0080] Where M is the total number of image pairs collected in one measurement.

[0081] The following is the specific implementation process and data results of this embodiment:

[0082] This embodiment uses a homemade spectral domain OCT system, which includes a superluminescent diode (SLD) with a central wavelength of 840nm. The light from the SLD passes through a fiber optic beam splitter and enters the Michelson interferometer. The detection light of the interferometer is guided by a two-dimensional scanning galvanometer to scan the fluid in the glass capillary. In this study, the inner diameter of the glass capillary used is about 380μm and the outer diameter is about 500μm. This study uses diluted milk as the test solution, which is pushed through the glass capillary by a syringe pump. The interference signal of the spectral domain OCT system is detected by a homemade spectrometer, which includes a high-speed linear array CCD camera. In this study, the time interval between two adjacent samplings of the CCD camera is set to T = 30μs. Taking into account the phase folding, the maximum measurable axial velocity component corresponding to this time interval is 5.3mm / s.

[0083] Figure 8 A pair of OCT structural images acquired using the OCT system are shown, wherein Figure 8 (a) and Figure 8 (b) The images were collected at positions S1 and S2, respectively, with the interval Δy between the two scanning surfaces being 1.25 mm. The center position of the fluid channel is marked with a crosshair in the figure, which is automatically calculated using the method in this study.

[0084] In this embodiment, five measurements are required for each flow setting. In each measurement, five pairs of Doppler OCT images need to be collected. Five images are sampled at the S1 position, and the other five images are collected from the S2 position. The collected OCT data are saved for subsequent processing. It takes about 3 minutes to collect and save 5 sets of data containing spectral interference signals.

[0085] When the interval between the two planes Δy=0.4 mm, the angle θ measured is shown in Table 1, where MT represents the number of measurements. In this measurement, the setting value of the microfluidic flow rate is 121 μl / min. The average angle of five measurements is 82.28 degrees, and the standard deviation is 0.14 degrees. Figure 2 (b) shows that the Doppler signal has a negative frequency shift, Δν<0. According to formula (1), we can deduce that cosα>0, so the Doppler angle α=θ=82.28 degrees.

[0086] Table 1

[0087]

[0088] Table 2 shows the measurement results of microfluid flow under different flow settings, where the first line is the set value of the microfluid flow, and the second line is the result measured using the method of the present invention. The relationship between the measured flow (Q) and the set flow (η) calculated using linear regression is Q=1.03η-3.38. At one flow setting value, for 5 measurements, the collection and storage of the interference spectrum signal takes about 3 minutes, so the measurement results in Table 2 are the average flow within 3 minutes. The experimental results show that the flow value automatically processed by the method proposed by the present invention is consistent with the set value, and the method proposed by the present invention is reliable.

[0089] Table 2

[0090]

[0091] The above embodiments are only for illustrating the technical concept and features of the present invention, and their purpose is to enable people familiar with the technology to understand the content of the present invention and implement it accordingly, and they cannot be used to limit the protection scope of the present invention. Any equivalent transformation or modification made according to the spirit of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for automatically measuring microfluid flow based on spectral domain optical coherence tomography technology, characterized in that: The method comprises the following steps: S1 uses an OCT device to scan the microfluid twice at different positions to obtain OCT structural images and Doppler phase images, and uses a dual-plane method to measure the Doppler angle between the detection beam and the microfluid; S2 performs binarization processing on the acquired OCT structural image to obtain a binarized image of the microfluidic cross section, and adjusts the amplitude threshold of the binarization processing according to the background noise intensity of the OCT system; S3 averages the values ​​of the binary image of the microfluidic channel in the vertical direction to obtain the center position of the microfluidic channel in the horizontal direction; S4 takes the horizontal center position as the center, averages and smoothes the values ​​of the binary image of the microfluidic channel in the horizontal direction, and obtains the vertical center position of the microfluidic channel; S5 performs a phase recovery operation on the Doppler phase image, taking the center coordinate of the microfluidic channel as the center of the circle and selecting a circle with a diameter slightly smaller than the inner diameter of the capillary as the boundary for phase recovery; S6 sets the microfluid flow ellipse boundary, adds the phase signals within the boundary to obtain a sum value, divides the calculated average of the sum value by the cosine of the Doppler angle, and obtains the flow result of the microfluid to be measured.

2. The method for automatically measuring microfluid flow based on spectral domain optical coherence tomography technology according to claim 1 is characterized in that: The process of obtaining the Doppler angle described in S1 is as follows: The OCT scanning planes are defined as the x and z planes. In order to measure the angle θ between the OCT beam and the pipeline, the OCT sampling beam scans the pipeline twice alternately between the two scanning planes S1 and S2 with a small displacement Δy. The coordinates of the pipeline center in plane S1 are expressed as (x1, z1), and the coordinates in plane S2 are expressed as (x2, z2). The angle θ is obtained by calculating the center position of the fluid on the two scanning planes. According to the flow direction in the microchannel, the Doppler angle α=θ or α=180-θ is obtained.

3. The method for automatically measuring microfluid flow based on spectral domain optical coherence tomography technology according to claim 1, characterized in that: The amplitude threshold in S2 is obtained as follows: An amplitude threshold A is used th， For an image pixel at position (i, j), where j is the pixel number in the depth direction, if its OCT amplitude A(i, j)>A th , the signal strength B(i, j) of the binary image is set to 512; otherwise, B(i, j) = 0, the threshold A th It is determined according to the background noise level of the OCT system. The area above the surface of the OCT structural image contains background noise information. In order to determine the background noise, 900 vertical sampling lines are horizontally averaged. The concave area at the lower left of the curve corresponds to the area above the surface of the OCT structural image. Its signal amplitude represents the background noise level. The average noise level A is obtained by averaging the signals from the 10th pixel to the 30th pixel in the concave area of ​​the curve. noise =232.8, select threshold A according to the signal fluctuation of each vertical sampling line th is 400.0, about A noise 1.72 times of.

4. The method for automatically measuring microfluid flow based on spectral domain optical coherence tomography technology according to claim 1, characterized in that: The detailed steps of S3 are as follows: For each vertical line in the binarized image of the microfluidic channel after binarization, i In the range of N pixels below, the binary signal B(i,j) is averaged in the vertical direction, and the average result B VA (i) can be expressed as Among them, the position P of the surface on the binary image of the microfluidic channel is i It is determined based on the amplitude signal A(i,j) of OCT. The vertical signal average curve is the average value B when N=270. VA (i), where the horizontal axis is the vertical line number i in the image, B VA (i) The curve has two negative peaks, appearing at i=i L and i=i R i L The negative peak at corresponds to the dark area on the left wall of the capillary. R The negative peak at corresponds to the wall of the right side of the capillary. VA (i) The minimum value in the curve is i L and i R After that, the capillary horizontal center i can be calculated c =(i L +i R ) / 2, thus obtaining the horizontal center position x of the microfluid c For x c =dx*i c , where dx is the unit length corresponding to each pixel in the horizontal direction.

5. The method for automatically measuring microfluid flow based on spectral domain optical coherence tomography technology according to claim 4 is characterized in that: The horizontal center position i c As the center, the binary signals B(i,j) of (2D+1) vertical lines are horizontally averaged to obtain the average result BHA(ic,j): Where D is the average range selected for horizontal averaging, that is, the horizontal center position i c Take D vertical lines on both sides as the center and average them; Using (2U+1) pixels, BHA(ic,j) is processed by moving average in the vertical direction, and the smoothed result Bsm(ic,j) is: Among them, U is the range used for vertical moving average. For a certain point, U points are taken above and below the point, and the binary value B is calculated within this range. HA Perform averaging and use the average value as the smoothing result of this point; P c The vertical line i represents the surface of the binary image of the microfluidic channel. c At the position, the smoothing result of the vertical center of the microchannel when U=6 is B sm (i), the horizontal axis is the pixel number j in the vertical direction. The smoothed curve has two negative peaks, which appear at j=j. u and j = j b At j u The negative peak at corresponds to the dark area on the upper part of the glass capillary wall. b The negative peak at corresponds to the bottom of the capillary wall. sm (i c , j) The minimum value in the curve is obtained at position j u and j b After that, the vertical center j of the capillary can be calculated c , i.e. j c =(j u +j b ) / 2, and obtain the center coordinate z of the capillary in the vertical direction c For z c =dz*j c , where dz is the unit length corresponding to each pixel in the depth direction.

6. The method for automatically measuring microfluid flow based on spectral domain optical coherence tomography technology according to claim 1, characterized in that: The detailed steps of S5 are as follows: after determining the center position of the microfluidic channel, use the circle center as (i c ,j c ), the circle O with a diameter of d is used as the boundary to perform phase unwrapping on the Doppler phase image. The diameter d of the circle O should be smaller than the inner diameter of the capillary and should be close to the inner wall of the pipe. For the pixels on the circle O, since the flow velocity close to the pipe wall is low, no wrapped phase will appear. The phase value of the pixel on the circle can be used as a boundary condition. The upper and lower semicircles of the circle O are used as the upper and lower boundaries of the phase recovery respectively. The Doppler phase of the pixel points in the circle O can be expanded in the vertical direction from the upper semicircle to the lower semicircle to obtain the restored Doppler phase image.

7. The method for automatically measuring microfluid flow based on spectral domain optical coherence tomography technology according to claim 6, characterized in that: In the vertical direction of the restored Doppler phase image, for each vertical line, the Doppler signals of 10 consecutive pixels downward from the image surface are averaged, and the average value is subtracted from the phase signal of the entire vertical line to obtain the phase signal ψ(i, j) caused by liquid flow.

8. The method for automatically measuring microfluid flow based on spectral domain optical coherence tomography technology according to claim 7, characterized in that: In order to calculate the volume flow rate of the microfluid, it is necessary to integrate the Doppler signal of the fluid in the flow channel. Due to the random phase noise in the capillary wall, the integration boundary is difficult to determine. In order to eliminate the influence of random phase noise on the microfluid flow calculation, the random phase noise in the capillary wall is reset to zero according to the OCT signal intensity A(i, j), and ψ(i, j) is converted to Γ(i, j). The specific method is as follows: In the area outside circle O in the Doppler phase image, if the OCT amplitude A(i,j)>A th , then its phase value Γ(i,j)=ψ(i,j); otherwise, Γ(i,j)=0; for the area inside the circle O, the phase value remains unchanged, that is, Γ(i,j)=ψ(i,j), where the threshold A th The choice of is the same as the threshold selected during image binarization, by which the random phase noise on the pipe wall is replaced by zero values.

9. The method for automatically measuring microfluid flow based on spectral domain optical coherence tomography technology according to claim 1, 4 or 5, characterized in that: The detailed steps of S6 are as follows: ①Set the microfluid flow ellipse boundary: Take (i c ,j c ) represents the boundary of volume flow calculation. The lengths of the major and minor axes of the ellipse are |i R -i L | and |j u -j b |, add the phase signals Γ(i,j) in the elliptical area to obtain the sum value σ; ② Calculate the flow rate of the microfluid to be tested: Multiple pairs of Doppler OCT images are collected for each flow measurement. For the mth pair of images, the phase sums of the Doppler images collected in the S1 and S2 planes are σ m1 and σ m2 , then the flow Q calculated using the mth pair of Doppler OCT images m It can be expressed as: Among them, θ m is the angle calculated using the mth pair of images, β m is the angle between the OCT scanning plane and the plane perpendicular to the capillary axis, cosβ m =Δy / R, then the flow rate result Q measured in one measurement is as follows: Where M is the total number of image pairs collected in one measurement.

10. A computer storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the automatic microfluid flow measurement method based on spectral domain optical coherence tomography technology as described in any one of claims 1 to 7 can be implemented in a non-contact manner.

Images