CN104173038B - Based on the blood flow velocity measurement method of frequency domain laser speckle imaging - Google Patents

Abstract

The invention discloses a kind of blood flow velocity measurement method based on the imaging of frequency domain laser speckle, comprise the following steps: by laser beam irradiation on testee, recycling imaging system is to testee imaging, the original speckle image of testee is gathered by imageing sensor, then single pixel in the original speckle image gathered is in dynamic speckle intensity-conversion in time domain to frequency domain, and rated output spectrum density, and fitting of a polynomial is carried out to power spectral density obtain smoothed curve, then smoothed curve is transformed into time domain by Fourier transformation, calculate the auto-covariance function of pixel and be normalized, then the measurement model of blood flow rate is set up, obtain the relation between auto-covariance function and blood flow rate, final matching obtains Hemodynamic environment angle value, the present invention not only eliminates static noise, improves the accuracy of measurement of blood flow rate, and avoids the impact of Imagery environmental factor as the intensity of light source, irradiating angle etc., improves Measurement sensibility.

Description

Based on the blood flow velocity measurement method of frequency domain laser speckle imaging

Technical field

The present invention relates to biological tissue's blood flow imaging field, be specifically related to a kind of blood flow velocity measurement method based on the imaging of frequency domain laser speckle.

Background technology

Laser speckle contrast imaging is that a kind of optical imaging system that utilizes transmits laser speckle image to realize carrying out the whole audience blood flow technology of imaging.Laser speckle contrast imaging system is primarily of coherent source and image capture device composition, and coherent light is random superposition after biological tissue scatters, and gathers random speckle pattern by image capture device and carry out space and contrast analysis, finally estimates blood flow rate.Laser speckle contrast imaging method has the features such as simple to operate, practical, therefore obtains a wide range of applications in fields such as biomedical research, clinical diagnosis, surgical guidance, skin, dentistry, ophthalmology and neurosciences.

But when adopting laser speckle contrast imaging method to carry out blood flow velocity measurement, due to the impact by the scattering process (as VELOCITY DISTRIBUTION, static speckles and multiple speckle) in environmental condition (as the intensity of light source, irradiating angle, imaging thing and camera parameter) and biological tissue, there is comparatively big error in the blood flow rate that existing laser speckle contrast imaging method is recorded.

For above problem, the people such as Parthasarathy utilize many time of exposure speckle contrast formation method obtain auto-correlation function replace use single contrast value acquisition single-point auto-correlation function value, to reduce the impact that static scattering brings, the effective informations such as VELOCITY DISTRIBUTION type can be obtained simultaneously.Afterwards, Thompson and Andrews points out again on this basis, is the form of Doppler frequency spectrum by auto-covariance Curve transform, and the algorithm during laser-Doppler just can be used to measure estimates blood flow rate.Although improve by above method the effectiveness that speckle contrasts angle value to a great extent, but because speckle image still affects by multiple Imagery environmental factor, therefore utilize laser speckle contrast imaging method directly to measure the error problem existed in blood flow rate and still effectively do not solved.

Summary of the invention

The present invention, in order to overcome above deficiency, provides one can eliminate environment factor impact, improves the blood flow velocity measurement method based on the imaging of frequency domain laser speckle of accuracy of measurement.

In order to solve the problems of the technologies described above, technical scheme of the present invention is: a kind of blood flow velocity measurement method based on the imaging of frequency domain laser speckle, comprises the following steps:

S1: by laser beam irradiation on testee;

S2: utilize imaging system to testee imaging;

S3: utilize imageing sensor to gather the original speckle image of testee;

S4: calculate the single pixel in original speckle image, to obtain the auto-covariance function of single pixel, is normalized the auto-covariance function of single pixel, comprises the following steps:

S41: utilize public formula I to carry out Fourier transformation to the dynamic speckle intensity that single pixel (x, y) in the original speckle image gathered is in time domain, be transformed into frequency domain:

$\tilde{I}\left(\mathrm{\ω}\right)=\frac{1}{2\mathrm{\π}}\∫I\left(t\right)e^{-\mathrm{i\ωt}}\mathrm{dt}---\left(I\right)$

Wherein I (t) represents pixel (x, y) place light intensity sequence in time domain, represent pixel (x, y) place light intensity sequence in frequency domain, x and y represents abscissa and the vertical coordinate of pixel respectively;

S42: rated output spectrum density and it is right carry out fitting of a polynomial and obtain smoothed curve;

S43: utilize public formula II that the smoothed curve in step S42 is transformed into time domain by Fourier transformation, calculates the auto-covariance function of pixel (x, y):

$C_t\left(\mathrm{\τ}\right)={\∫\left|\tilde{I}\left(\mathrm{\ω}\right)\right|}^2{e}^{\mathrm{i\ω\τ}}\mathrm{d\ω}-{\⟨\tilde{I}\left(\mathrm{\ω}\right)\⟩}^2---\left(\mathrm{II}\right)$

Wherein τ represents interval;

And to C _t(τ) be normalized;

S5: the measurement model setting up blood flow rate, obtains auto-covariance function and the relation definitely between blood flow rate:

$C_t\left(\mathrm{\τ}\right)=e^{-\frac{M^2{v_0}^2{\mathrm{\τ}}^2}{{l_0}^2+M^2{\tilde{v}}^2{\mathrm{\τ}}^2}}[\frac{{l_0}^3}{{({l_0}^2+M^2{\tilde{v}}^2{\mathrm{\τ}}^2)}^{\frac{3}{2}}}+\frac{2M^4{v_0}^2{\tilde{v}}^2{l}_0{\mathrm{\τ}}^4}{{({l_0}^2+M^2{\tilde{v}}^2{\mathrm{\τ}}^2)}^{\frac{5}{2}}}]---\left(\mathrm{III}\right)$

Wherein M is the amplification of imaging system, and wherein τ represents interval; l ₀optical wavelength is irradiated in=0.41M λ/NA, λ representative, and NA is numerical aperture, v ₀the average speed of pixel (x, y), it is the root mean sequare velocity of pixel (x, y).

S6: by the auto-covariance function C through normalized in step S43 _t(τ) substitute into public formula III and carry out matching, obtain the blood flow rate v of pixel (x, y) ₀;

S7: repeat step S4-S6 to each pixel in original speckle image, to dynamic monitoring and the analysis of biological tissue specific region or focal area blood flow rate.

Further, described imaging system is battery of lens imaging system.

Blood flow velocity measurement method based on the imaging of frequency domain laser speckle provided by the invention, is first transformed into the original speckle image of imageing sensor collection in frequency domain and processes, obtain auto-covariance function, and be normalized; Then the measurement model of blood flow rate is set up, auto-covariance function after normalization and blood flow rate are connected, final blood flow rate is obtained by matching, static noise is not only eliminated by the method, improve the accuracy of measurement of blood flow rate, and avoid the impact of Imagery environmental factor as the intensity of light source, irradiating angle etc., improve Measurement sensibility.

Accompanying drawing explanation

Fig. 1 is the flow chart of the blood flow velocity measurement method based on the imaging of frequency domain laser speckle that the present invention proposes;

Fig. 2 is battery of lens imaging system structural representation of the present invention;

Fig. 3 a ~ 3c is the inventive method and the measurement result comparison diagram of traditional method in fluid simulation experiment;

Fig. 4 a ~ 4c is that the inventive method and traditional method are to the measurement result comparison diagram of rat auris dextra Ink vessel transfusing blood flow rate.

Shown in figure: 1, laser instrument; 2, testee; 3, battery of lens; 4, imageing sensor; 5, PC.

Detailed description of the invention

Below in conjunction with accompanying drawing, the present invention is described in detail:

As shown in Figure 1, the invention provides a kind of blood flow velocity measurement method based on the imaging of frequency domain laser speckle, comprise the following steps:

S1: by laser beam irradiation on testee;

S2: utilize imaging system to testee imaging;

S3: utilize imageing sensor to gather the original speckle image of testee;

S4: calculate the single pixel in original speckle image, to obtain the auto-covariance function of single pixel, is normalized the auto-covariance function of single pixel, comprises the following steps:

S41: utilize public formula I to carry out Fourier transformation to the dynamic speckle intensity that single pixel (x, y) in the original speckle image gathered is in time domain, be transformed into frequency domain:

$\tilde{I}\left(\mathrm{\ω}\right)=\frac{1}{2\mathrm{\π}}\∫I\left(t\right)e^{-\mathrm{i\ωt}}\mathrm{dt}---\left(I\right)$

Wherein I (t) represents pixel (x, y) place light intensity sequence in time domain, represent pixel (x, y) place light intensity sequence in frequency domain, x and y represents abscissa and the vertical coordinate of pixel respectively;

S42: rated output spectrum density and it is right carry out fitting of a polynomial and obtain smoothed curve;

S43: utilize public formula II that the smoothed curve in step S42 is transformed into time domain by Fourier transformation, calculates the auto-covariance function of pixel (x, y):

$C_t\left(\mathrm{\τ}\right)={\∫\left|\tilde{I}\left(\mathrm{\ω}\right)\right|}^2{e}^{\mathrm{i\ω\τ}}\mathrm{d\ω}-{\⟨\tilde{I}\left(\mathrm{\ω}\right)\⟩}^2---\left(\mathrm{II}\right)$

Wherein τ represents interval;

And to C _t(τ) be normalized;

S5: the measurement model setting up blood flow rate, obtains auto-covariance function and the relation definitely between blood flow rate:

Suppose t ₀point (x when=0 on object plane ₀, y ₀) place only has a granule, movement velocity is v, under the irradiation of laser beam, and (x ₀, y ₀) the electric field amplitude distribution of point (x, y) place scattering in corresponding picture plane can be expressed as δ (x-x ₀-v τ, y-y ₀), and for the imaging system that numerical aperture is fixed, the electric field amplitude at point (x, y) place can be expressed as:

U(x,y)＝δ(x-Mx ₀-Mvτ,y-My ₀)*h(x,y)

(Ⅲ-1)

＝h(x-Mx ₀-Mvτ,y-My ₀)

Wherein, M is the amplification of imaging system, and h (x, y) is the optical transfer function of imaging system, and * is convolution algorithm, and τ represents interval.Suppose that pattern distortion can be ignored, then h (x, y) can be expressed as the some transfer function of system:

Wherein, J ₁single order Bessel function of the first kind, number, λ is the wavelength irradiating light, finds out that h (x, y) is for Airy disk pattern, radius r from above formula ₀for:

$r_0=0.61\frac{\mathrm{M\λ}}{\mathrm{NA}}---(\mathrm{III}-3)$

The granule assuming picture has single speed v, elapsed time interval τ, (x ₀, y ₀) point (x in corresponding picture plane ₁, y ₁) auto-covariance function C _tcan be expressed as after (v, τ) normalized:

$C_t(v,\mathrm{\τ})=\⟨\frac{\left|h{(x_1-M{x}_0,y_1-M{y}_0)}^{*}h(x_1-M{x}_0-\mathrm{Mv\τ},y_1-M{y}_0)\right|}{{\left|h(x_1-M{x}_0,y_1-M{y}_0)\right|}^2}\⟩---(\mathrm{III}-4)$

Assuming that t ₀when=0, any initial position all has abundant granule to object plane, movement velocity is all v, then for interval τ, corresponding for granules all on each initial position time average as the auto-covariance in plane can be converted to spatial averaging:

$C_t(v,\mathrm{\τ})=\frac{\∫\∫{\left|h{(x,y)}^{*}h(x-\mathrm{Mv\τ},y)\right|}^2\mathrm{dxdy}}{\∫{\left|h(x,y)\right|}^2\mathrm{dxdy}}---(\mathrm{III}-5)$

Here integration is carried out to whole x-y plane, x ₁-Mx ₀and y ₁-My ₀replace with x and y.

According to formula (III-5), C _t(v, τ) regards two as at a distance of M _{v τ}the overlapping region of Airy disk.Then C _t(v, τ) can approximate representation be following Gaussian function:

$C_t(v,\mathrm{\τ})=e^{\frac{-{\left(\mathrm{Mv\τ}\right)}^2}{{l_0}^2}}---(\mathrm{III}-6)$

Wherein l ₀=Mv τ ₀be decorrelation length, represent the e that two Airy disk overlapping regions are original overlapping regions ^-1, wherein original overlapping region refers to the region that two Airy disk patterns are completely overlapping.

For given VELOCITY DISTRIBUTION P (v), auto-covariance function can be expressed as:

$C_t\left(\mathrm{\τ}\right)={\∫}_0^{+\∞}P\left(v\right)C_t(v,\mathrm{\τ})\mathrm{dv}---(\mathrm{III}-7)$

Blood flow rate distribution is considered to that Brownian movement and Gauss distribution form jointly.Therefore, suppose that blood flow rate distribution P (v) is made up of two kinds of speed types, can be expressed as:

$P\left(v\right)=\frac{2}{\sqrt{\mathrm{\π}}{\tilde{v}}^3}{(v-v_0)}^2{e}^{-\frac{{(v-v_0)}^2}{{\tilde{v}}^2}}---(\mathrm{III}-8)$

C can be solved by (III-7) and (III-8) formula _t(τ) expression formula is:

$C_t\left(\mathrm{\τ}\right)=e^{-\frac{M^2{v_0}^2{\mathrm{\τ}}^2}{{l_0}^2+M^2{\tilde{v}}^2{\mathrm{\τ}}^2}}[\frac{{l_0}^3}{{({l_0}^2+M^2{\tilde{v}}^2{\mathrm{\τ}}^2)}^{\frac{3}{2}}}+\frac{2M^4{v_0}^2{\tilde{v}}^2{l}_0{\mathrm{\τ}}^4}{{({l_0}^2+M^2{\tilde{v}}^2{\mathrm{\τ}}^2)}^{\frac{5}{2}}}]---\left(\mathrm{III}\right)$

Wherein M is the amplification of imaging system, l ₀=0.41M λ/NA, λ represents optical wavelength, and NA is numerical aperture, v ₀the average speed of pixel (x, y), it is the root mean sequare velocity of pixel (x, y).

S6: by the auto-covariance function C through normalized in step S43 _t(τ) substitute into public formula III and carry out matching, obtain the blood flow rate v of pixel (x, y) ₀.

S7: repeat step S4-S8 to each pixel in original speckle image, to dynamic monitoring and the analysis of biological tissue specific region or focal area blood flow rate.

As shown in Figure 2, build battery of lens imaging system, laser instrument 1 Emission Lasers irradiates on testee 2, and scattering laser is collected through battery of lens 3, and carry out imaging by the imageing sensor 4 of high frame rate, picture signal is imported PC 5 pairs of blood flow rate into and is monitored and analyze the most at last.

Be below fluid simulation experiment:

Measurand 2 is the granules of polystyrene (0.1wt% at the uniform velocity flowed, average diameter is 3.2 μm), granules of polystyrene is controlled with 0.1mm/s by syringe pump, 0.5mm/s, the speed of 1mm/s at the uniform velocity flows, irradiate with the laser instrument of 780nm, then gather original laser speckle image by the high speed imaging sensor imaging be fixed on stereoscopic microscope (NA=0.2) with the speed that 1000 frames are per second, and measure.

As shown in Figure 3 a, wherein solid line selected areas is the region gathering original speckle to the laser speckle image that imageing sensor gathers; By calculating the VELOCITY DISTRIBUTION of institute's selection area under liquid different in flow rate as shown in Figure 3 b, wherein x-axis 0.0 place is the center in selected areas x direction, dotted line is the loose point value recorded, solid line carries out the parabolic curve after matching to dotted line, representing flow velocity from the bottom to top is respectively 0.1mm/s, the measurement result of 0.5mm/s, 1mm/s; Relation between the liquid velocity that measurement obtains and actual liquid speed as shown in Figure 3 c, wherein loose point value is experiment value, straight line representation theory value, as can be seen from the figure, for the liquid of different in flow rate, the result adopting the method for the present embodiment to record and theoretical value closely, therefore can find out that the method for the present embodiment has higher accuracy.

Be below the blood flow velocity measurement experiment of living animal:

Measurand is a SpragueDawley rat, uses chloral hydrate to anaesthetize it, and removes the hair of rat auris dextra with depilatory cream, under being fixed in stereoscopic microscope.The laser instrument of the 780nm of four same sizes is placed on the different position of four angles and (is labeled as angle 1 respectively, 2,3,4), from different angular illumination auris dextras, traditional method and context of methods is then adopted to measure the endovascular blood flow rate of auris dextra respectively.Because rat is in steady statue within the short time (5 minutes), therefore during this period, the blood flow rate of rat is in steady statue.In order to further illustrate performance of the present invention, traditional method and the inventive method two kinds of methods are adopted to measure.

As shown in fig. 4 a, in figure, solid line is the region that two kinds of methods gather original speckle to the laser speckle image that imageing sensor gathers.Traditional method and the inventive method is adopted under different measuring angle, to record angiocentric Hemodynamic environment angle value as shown in Figure 4 b, wherein left side represents the Hemodynamic environment angle value adopting the inventive method to record, the Hemodynamic environment angle value that right side representative adopts traditional method to record, when as can be seen from the figure adopting traditional method to measure, the blood flow rate deviation recorded under different taking measurement of an angle is comparatively large, illustrates that the impact of the change of illuminate condition on traditional method is larger.And for different illuminate conditions, the blood flow rate recorded by the inventive method only has very little deviation, therefore the inventive method has higher stability.Fig. 4 c shows the blood flow rate curve that two kinds of methods record same pickup area for 3 times in angle 2 and angle, can find out and adopt the value that records under two kinds of angles of method of the present invention closely, and the value adopting traditional method to record under two kinds of angles exists larger gap, illustrate that the inventive method has better stability.

In sum, the blood flow velocity measurement method based on the imaging of frequency domain laser speckle provided by the invention, is first transformed into the original speckle image of imageing sensor collection in frequency domain and processes, obtain auto-covariance function, and be normalized; Then the measurement model of blood flow rate is set up, auto-covariance function after normalization and blood flow rate are connected, final blood flow rate is obtained by matching, not only eliminate static noise, improve the accuracy of measurement of blood flow rate, and avoid the impact of Imagery environmental factor as the intensity of light source, irradiating angle etc., improve Measurement sensibility.

Although be illustrated embodiments of the present invention in description, these embodiments just as prompting, should not limit protection scope of the present invention.Carry out various omission, displacement and change without departing from the spirit and scope of the present invention all should be included in protection scope of the present invention.

Claims (5)

1., based on a blood flow velocity measurement method for frequency domain laser speckle imaging, it is characterized in that, comprise the following steps:

S1: by laser beam irradiation on testee;

S2: utilize imaging system to testee imaging;

S3: utilize imageing sensor to gather the original speckle image of testee;

S4: calculate the single pixel in original speckle image, to obtain the auto-covariance function of single pixel, is normalized the auto-covariance function of single pixel;

S5: the measurement model setting up blood flow rate, obtains the relation between auto-covariance function and blood flow rate;

S6: matching is carried out to the measurement model of the auto-covariance function after normalized and blood flow rate, obtains the blood flow rate of single pixel;

S7: repeat step S4 ~ S6, obtain the blood flow rate of each pixel in original speckle image, and then the dynamic monitoring carried out biological tissue's focal area blood flow rate and analysis.

2. the blood flow velocity measurement method based on the imaging of frequency domain laser speckle according to claim 1, is characterized in that, described imaging system is battery of lens imaging system.

3. the blood flow velocity measurement method based on the imaging of frequency domain laser speckle according to claim 1, it is characterized in that, described step S4 comprises:

S41: utilize public formula I to carry out Fourier transformation to the dynamic speckle intensity that single pixel (x, y) in the original speckle image gathered is in time domain, be transformed into frequency domain:

Wherein I (t) represents pixel (x, y) place light intensity sequence in time domain, represent pixel (x, y) place light intensity sequence in frequency domain, x and y represents abscissa and the vertical coordinate of pixel respectively;

S42: rated output spectrum density and it is right carry out fitting of a polynomial and obtain smoothed curve;

S43: utilize public formula II that the smoothed curve in step S42 is transformed into time domain by Fourier transformation, calculates the auto-covariance function of pixel (x, y):

Wherein τ represents interval;

To C _t(τ) be normalized.

4. the blood flow velocity measurement method based on the imaging of frequency domain laser speckle according to claim 3, is characterized in that, in described step S5, the pass between auto-covariance function and blood flow rate is:

Wherein M is the amplification of imaging system, and τ represents interval, l ₀=0.41M λ/NA, λ are for irradiating optical wavelength, and NA is the fixed numbers aperture of imaging system, v ₀the blood flow rate of pixel (x, y), it is the root mean sequare velocity of pixel (x, y).

5. the blood flow velocity measurement method based on the imaging of frequency domain laser speckle according to claim 4, is characterized in that, by the auto-covariance function C through normalized in step S43 _t(τ) substitute into public formula III and carry out matching, obtain the blood flow rate v of pixel (x, y) ₀.