CN105796053A - Method for measuring dynamic contrast ratio and estimating transverse flow with OCT - Google Patents

Abstract

The invention discloses a method for measuring the dynamic contrast ratio and estimating transverse flow with OCT. A method for determining a sample movement vector includes the steps that a sample image at the time T1 is obtained, a preset area in the sample image is selected as a tracking image, and image identifying information and the position at the time T1 of the tracking image are obtained; a sample image at the time T2 is obtained, and the position the same as the position of the tracking image at the time T1 is found to serve as a tracking box; one point of the tracking box serves as a reference point, and the maximum movement ranges of the reference range along the transverse axis and the vertical axis are determined, and are divided into a plurality of movement positions; image identifying information of the tracking image at the time T1 when the reference point moves according to vectors and conducts transversal on movement positions and the association coefficients of the image identifying information of images in the tracking box when the reference point moves to all the movement positions are calculated, and the reference-point movement vector corresponding to the maximum value of the association coefficients is the vector of a sample at the time T2 relative to the time T1.

Description

OCT is utilized to measure dynamic contrast and the method estimating horizontal flow

Technical field

The present invention relates to optical image technology field, utilize OCT to measure dynamic contrast and the method estimating horizontal flow particularly to a kind of.

Background technology

Optical coherence tomography (opticalcoherencetomography, OCT) is the optical image technology of a kind of non-intrusion type.In recent years, at medical domain, particularly in field of ophthalmology, optical coherence tomography system can to the shooting test object (retina such as optical fundus, arteria and vena centralis retinae etc.) tomography carry out three-dimensional visualization imaging, as the tomography of optical fundus specific part is carried out three-dimensional visualization imaging.When the tomography of optical fundus specific part is carried out three-dimensional visualization imaging, the position on optical fundus is determined firstly the need of optical fundus is carried out two-dimensional imaging, by finding the specific part on optical fundus at optical fundus two-dimensional imaging, the specific part on optical fundus is carried out OCT scan to carry out three-dimensional visualization imaging.But, because the ocular movement of watching attentively property can cause the movement on optical fundus, namely optical fundus has been moved.In prior art, carrying out the camera lens of OCT scan is do not move into Row sum-equal matrix with optical fundus, and so, the concrete position resulting in the optical fundus that OCT scan is scanned there occurs change, have impact on the accuracy of three-dimensional visualization imaging.

Summary of the invention

The invention provides a kind of method utilizing OCT measurement dynamic contrast and the horizontal flow of estimation, sample move to method for determination of amount and collecting method, solve because sample moves the technical problem that the specific part to sample caused carries out the poor accuracy of three-dimensional visualization imaging.

For reaching above-mentioned purpose, the present invention provides techniques below scheme:

A kind of sample move to method for determination of amount, it is characterised in that comprise the steps:

Obtain the sample image in T1 moment, select the predeterminable area of sample image in T1 moment as tracking image, obtain tracking image for the image recognition information identified and the position in tracking image T1 moment；

Obtain the sample image in T2 moment, find the position identical with the position in tracking image T1 moment as following the trail of frame in the sample image in T2 moment；

Plane right-angle coordinate is set up for initial point with the point following the trail of frame, with follow the trail of frame a point for reference point, determine that reference point is along the moveable maximum magnitude of transverse axis and vertical pivot, is divided into multiple shift position along the moveable minimum spacing of transverse axis and vertical pivot by moveable maximum magnitude；

Calculating reference point is according to vector (m, when n) moving and travel through shift position, the correlation coefficient of the image recognition information of image in frame is followed the trail of, the vector (m that the reference point that the maximum of described correlation coefficient is corresponding moves when the image recognition information of T1 moment tracking image and reference point move to each shift position_max,n_max) for the vector that the sample T2 moment moved relative to the T1 moment, the T2 moment is later than the T1 moment.

Sample provided by the invention move to method for determination of amount, tracking image from the sample image in T1 moment, obtaining two kinds of information, a kind of is distinguish tracking image and other parts of images so that the image recognition information being identified, and another kind is the tracking image position in the T1 moment；Then, obtain the sample image in T2 moment, now due to the movement of sample, the position of tracking image is moved, for the vector finding tracking image to move, it is necessary to utilize tracking image to find the frame of the position identical with its position as following the trail of frame in the sample image in T2 moment in the position in T1 moment；Afterwards, follow the trail of frame to move in its moveable maximum magnitude, calculating tracking frame follows the trail of the image recognition information of image in frame and the correlation coefficient of the image recognition information in tracking image T1 moment, the i.e. degree of correlation of the image recognition information of two images when moving to the every bit in moveable maximum magnitude；What the maximum of described correlation coefficient was corresponding follow the trail of, and position that frame moves to is the immediate position of image recognition information of two images, for the position moved to of tracking image.The movement of tracking image is to be caused by the movement of sample, and therefore, the vector of the movement of tracking image is exactly the vector moved relative to the T1 moment in the sample T2 moment.As such, it is possible to determine the vector of the movement of sample very easily, improve the specific part to sample and carry out the accuracy of three-dimensional visualization imaging.

Accompanying drawing explanation

Fig. 1-01 be the optical fundus of the present invention move to the schematic diagram of the eye fundus image in T1 moment in method for determination of amount；

Fig. 1-02 be the optical fundus of the present invention move to the schematic diagram of the eye fundus image in T2 moment in method for determination of amount；

What the optical fundus of Fig. 1-03 present invention was moved determines the tracking image schematic diagram in the position of the eye fundus image in T2 moment in method for determination of amount；

The principle derivation schematic diagram to method for determination of amount that the optical fundus of Fig. 1-04 present invention is moved；

Shown in Fig. 1 is the frame diagram of SDOCT system.Low-coherence light source S (k) is divided into reference arm and sample arm by fibre optic interferometer.Reflection light is assembled in spectrogrph and tests, thus the degree of depth of reflection profile is computed；

The phase place delta data of the simulation of the fluid mass motion of to be radian the be π (1/4th of imaging source wavelength) shown in Fig. 2；

Shown in Fig. 3 is scattering object intended phase variance test result figure within a period of time of motion；

What Fig. 4 A showed is the overview diagram that average OCT intensity image non-to sample in Fig. 4 B is relevant；In order to obtain this image, the agarose well of 2% is dissolved in the fat emulsion solution that density matching is 0.1%；In image, luminance contrast is confined to the edge of flowing and the boundary of air；

Fig. 5 A is the phase place variation diagram of period of time T=40us；

Fig. 5 B is the phase place variation diagram of period of time T=80us；

Fig. 6 A is the phase place variation diagram of period of time T=200us；

Fig. 6 B is the phase place variation diagram of period of time T=400us；

Fig. 7 A is the phase place variation diagram of period of time T=800us；

Fig. 7 B is the phase place variation diagram of period of time T=1.6ms；

Fig. 8 A is the phase place change comparison diagram of maximum phase transformation period cycle T 2=40T1；

Fig. 8 B is the phase place change comparison diagram of maximum phase transformation period cycle T 2=20T1；

Fig. 9 A is the phase place change comparison diagram of maximum phase transformation period interval T2=10T1；

Fig. 9 B is the phase place change comparison diagram of maximum phase place transformation period interval T2=5T1；

Shown in Figure 10 is the phase place delta data of single scattering object in water；Shown sphere diameter respectively 0.5um, 2um and 5um.Diameter is the impact that the spheroid of 2um demonstrates phase error, and it is weak to be primarily due in the phase place delta data of desired form OCT signal；

What Figure 11 A showed is the image of the lateral scanning pattern of MB-scanning；

The image of the lateral scanning pattern of the BM-scanning shown in Figure 11 B；

Shown in Figure 12 A is the OCT gray level image of the Brachydanio rerio tail using MB-scan mode to obtain；Shown in Figure 12 B is the phase place change contrast image using MB-scan mode to obtain Brachydanio rerio tail；The size of image is 900um*325um.T2=1ms, T1=40us；

Shown in Figure 13 A is the OCT gray level image of the Brachydanio rerio tail obtained with BM-scan mode；

Shown in Figure 13 B is the phase place change contrast image of the Brachydanio rerio tail using BM-scan mode to obtain；The size of Figure 13 A and Figure 13 B image is all 815um*325um；Phase error in Figure 13 B eliminates, and the period of time T 2 of contrast image is estimated as 40ms；Note that the areas imaging size of phase place change contrast image is 4 times of the contrast image of the MB-scanning shown in Figure 12 B；

Shown in Figure 14 A is the mean flow rate image of the time point that low speed transverse movement occurs, and this image sources is in the data of the BM-amphiblestroid transverse movement scanned；

Shown in Figure 14 B is the phase contrast image after noise elimination and medium filtering, and the BM-deriving from retina transverse movement scans data；Figure 14 B is the phase contrast image of the time point that there is a small amount of fluid transverse movement；

Shown in Figure 15 is the phase place change contrast image of uncorrected a large amount of fluid transverse movement；

Figure 16 A and Figure 16 B is the comparison diagram of one group of phase contrast image；Shown in Figure 16 A is the image of a large amount of transverse movement of uncorrected existence；

Shown in Figure 16 B be α=0 when, the image of corrected a large amount of transverse movements；

Figure 17 A and 17B is the contrast summary image through the 2.6S retina degree of depth collected；

Figure 17 A is without the image compensated before α=0；

Figure 17 B is the image after α=0 through overcompensation；

Shown in Figure 18 A is that BM-scans average OCT gray-scale map；

Figure 18 B, 18C, 18D are three phase place change contrast figure of different time points；Each image is all collect within 50ms.The region of arrow points is spine longitudinal direction angiosomes, segmental vessels (Se) two kinds different, spine aorta (DA), axial vein (AV)；

Shown in Figure 19 A and 19B is the OCT visual image of the Brachydanio rerio being quickly fertilized for 3 days；Figure 19 A is the MIcrosope image in the bright visual field, and Figure 19 B is the image of the cabrilla being quickly fertilized for 3 days, all illustrates the anatomical features of desired cabrilla；The line drawn in Figure 19 A and 19B represents the scanning area of OCT image；Further analyze average discharge and phase place changes the quality that can promote these images；

The OCT gray level image of Figure 20 A illustrates the internal structure of cabrilla in Figure 19；

Shown in Figure 20 B is the blood flow of endocardial, and this image has reduced phase noise and OCT image pixel in image；Phase contrast image after the elimination phase error impact that Figure 20 C describes；Figure 20 B clearly describes the outward appearance of heart, the flow direction of the yolk sac that the direction of arrow shown in Figure 20 C refers to, and the region expected in Figure 19 B and matches；

It it is the sketch in the direction of flow relative to imaging source direction shown in Figure 21；The axial flow component that the DopplerOCT conceptions of technology observe is used to use V_zDemarcate；

Figure 22 is shown that the phase noise of SNR-restriction and the graph of a relation of average OCT strength signal；

The contrast image of the cabrilla tail that the data using the collection of MB-scan mode shown in Figure 23 A, 23B, 23C and 23D produce；Figure 23 A is the gray level image of structural information；Arrow in Figure 23 A point out desired by the region that detects, this region is main two blood vessels of fish, spine tremulous pulse and axial blood vessel；Figure 23 B is phase place change contrast images, and the time cycle is T2=1ms and T1=40us, and radian changes from 0 to 2, and viewed region is sightless on DopplerOCT image, and DopplerOCT can only observe sufficiently stable static region.Arrow in Figure 23 B point out desired by the moving region observed.Figure 23 C is Doppler flow diagram picture, and the scope of application is+-0.12 radian=+-200um/s, and phase place change meansigma methods is 5；Shown in Figure 23 D is Doppler flow diagram picture, and the scope of application is+-0.12 radian=+-200um/s, and phase place change meansigma methods is 100.More and more stable static characteristic improves visual quality, and the position understanding each several part in advance is conducive to visual research；

Shown by Figure 24 A and 24B is the same area of Brachydanio rerio in Figure 23 A, 23B, 23C and 23D, but, the BM-scan mode using trace interval to be T=10ms；In order to select best parameter, the time used by the view data of 200 pixels across that gathers is 50ms；Gather the dynamic range of data owing to reducing DopplerOCT method, use the image of the method collection no longer to present；Compared with the phase variance contrast image of OCT mean flow rate figure and Figure 24 B of Figure 24 A, 5 B-scanning.Each durection component carries out medium filtering, radian excursion 0 to 3；The arrow of each image points out the region that spine master pulse is relevant with axial blood vessel.Figure 24 B is phase place change contrast image, it is possible to clearly observe the MB-same area scanned, but owing to the change of blood vessel refractive index has extra shade below blood vessel；

Shown in Figure 25 is the phase contrast summary image of the slices across of cabrilla cardiac position in 2.6s.Each time point collects in 50ms；This kind of method is equally used in Figure 17；Change in blood vessel and heart can be visible in detail；Shown in Figure 26 is the data image of time dependent heart contrast changing value.The change of contrast, relevant with the change of cabrilla normal cardiac rate in region and the changes in flow rate of viewing area；

It it is the contrast changing over certain lateral attitude of cabrilla shown in Figure 26；This contrast sums up 3 pixels across (7.2um), covers the entire depth of cabrilla；

Shown in Figure 27 A, 27B, 27C is the direct picture of cabrilla heart；Figure 27 A is OCT intensity sum graph picture in logarithmic range；Figure 27 B is phase place change sum graph picture, improves visualization compared to Figure 27 C；The confocal images of the cabrilla Green Fluorescent Protein of Figure 27 B and quick fertilization in 3 days has similar area；Shown in Figure 27 C is inject the Confocal Images that fluorescent material produces in the cabrilla body at similar age, and Figure 27 B motion contrast images shows relevant blood-vessel image；

Shown in Figure 28 A is average OCT gray level image；

Shown in Figure 28 B is that MB-scans the amphiblestroid doppler flow spirogram picture of mouse；The Doppler flow diagram picture of Figure 28 B, it does not have use any threshold value, excursion is+-2.5mm/s.The main fast flow of the retinal vessel observed is axial flow, but uses this image analysis technology to observe suprachoroid vascular flow；Figure 28 A shows the phase place change contrast image at a series of phase place transformation period interval；Extend phase place transformation period interval and can improve blood vessel (choroidal artery) visualization, but also increase the contrast shade below blood vessel simultaneously.The prolongation of interval equally also increases the contrast in the sensitivity of transverse movement and vertical direction；

Figure 29 A is the phase place change contrast figure of MB-scanning, and interval is 40us, and phase place is changed to 10；

Figure 29 B show MB-and scans phase place change contrast figure, and interval is 160us, and phase place is changed to 40；

Figure 29 C show MB-and scans phase place change contrast figure, and interval is 240us, and phase place is changed to 40；

Figure 29 D show MB-and scans phase place change contrast figure, and interval is 320s, and phase place is changed to 40；Shown in Figure 29 A, 29B, 29C and 29D is the retinal images after flattening, is mainly used to eliminate optical path and changes the impact of the change causing mouse eyeball curvature.Amphiblestroid flattening and identify that layer of retina separate confinement is the method for two kinds main extractions information in image, information retrieval in the depth areas of three dimensional grey scale image and contrast-data, is analyzed and is gathered data, the image in formation transverse direction or front；

Shown in Figure 30 A is the B-scanogram before retina flattening does not rearrange；

Shown in Figure 30 B is the B-scanogram after retina flattening rearranges；

Figure 31 A, 31B and 31C are the amphiblestroid BM-scanogram of mouse；

Figure 31 A is mean intensity scanogram；Figure 31 B is phase place modified-image, and this image does not eliminate any mathematical phase error, also without filtering；Figure 31 C is the phase place change contrast image after noise eliminates and after medium filtering, and excursion is 0 to 3 radians；

Shown in Figure 32 A is the direct picture of whole retina BM-scanning；

Shown in Figure 32 B is the front phase contrast image of whole retina BM-scanning；

The direct picture of the total intensity shown in Figure 33 A, Figure 33 B be retina the first half phase place change contrast image；The retinal vessel of the contrast image display surface of Figure 33 B, arrow indication is the region of blood capillary；

The direct picture of total intensity of display depth shown in Figure 34 A, Figure 34 B is the phase place change contrast image of retina the latter half；The contrast image of Figure 34 B shows the shade contrast of choroidal blood vessel and main retinal vessel；Arrow indication is the region of blood capillary；

Shown in Figure 35 A is that BM-scans front summation gray level image, and shown in Figure 35 B is the BM-phase place change contrast image scanning that front is total.Figure 35 A is whole amphiblestroid image, but the contrast image of Figure 35 B, it is only the image of BM-scanning retina the first half that pixels across is 200；The pixel of the single contrast image of Figure 35 B is 200*51, by the image of the whole independent scanning collection of amphiblestroid top area.What arrow was pointed out is some less visible blood vessels；

Total phase contrast image in the front shown in Figure 36 A, 36B, 36C and 36D, pixels across is the image of amphiblestroid the first half of the BM-scan mode collection of 100；System being used alone BM-and sweeps some images of continuous scanning collection, the pixel of these images is 100*50, concludes therefrom that on retina, the contrast of part is 100*100；Arrow marks some visible blood capillaries, but these blood vessels are because the intermittence of its contrast is not all can occur on other image；The transversal scanning region of image is the same with the region of Figure 34 A and 34B；

What two shown in Figure 37 A and 37B kind were different repeats average contrast's image of the acquisition method statistics of BM-scanning, for the image of retina the first half；What Figure 37 A and 37B gathered is the image being orthogonal to main horizontal scanning direction.Image can be inferred that pixel size is 100 × 100 as a comparison.

Detailed description of the invention

Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is only a part of embodiment of the present invention, rather than whole embodiments.Based on the embodiment in the present invention, the every other embodiment that those of ordinary skill in the art obtain under not making creative work premise, broadly fall into the scope of protection of the invention.

Embodiment one

The optical fundus of one embodiment of the present of invention move to method for determination of amount, comprise the steps:

As shown in Fig. 1-01, it is thus achieved that the eye fundus image 100 in T1 moment, select the predeterminable area of eye fundus image in T1 moment as tracking image 110, obtain tracking image for the image recognition information identified and the position in tracking image T1 moment；

As shown in Fig. 1-02, it is thus achieved that the eye fundus image 200 in T2 moment, find the frame of the position identical with the position in tracking image T1 moment as following the trail of frame 210 in the eye fundus image in T2 moment；

Plane right-angle coordinate is set up for initial point with the point following the trail of frame, with follow the trail of frame 210 a point for reference point, determine that reference point is along the moveable maximum magnitude of transverse axis and vertical pivot, along the moveable minimum spacing of transverse axis and vertical pivot, moveable maximum magnitude is divided into multiple shift position 220；

Calculating reference point is according to vector (m, n) during mobile traversal shift position, the correlation coefficient of the image recognition information of image in frame is followed the trail of, the vector (m that the reference point that the maximum of described correlation coefficient is corresponding moves when the image recognition information of T1 moment tracking image and reference point move to each shift position_max,n_max) for the vector that the optical fundus T2 moment moved relative to the T1 moment, the T2 moment is later than the T1 moment.

The optical fundus of the present embodiment moves to method for determination of amount, first, the tracking image from the eye fundus image in T1 moment, obtain two kinds of information, a kind of is distinguish tracking image and other parts of images so that the image recognition information being identified, and another kind is the tracking image position in the T1 moment；

Then, obtain the eye fundus image in T2 moment, now due to the movement on optical fundus, the position of tracking image is moved, for the vector finding tracking image to move, it is necessary to utilize tracking image to find the frame of the position identical with its position as following the trail of frame in the eye fundus image in T2 moment in the position in T1 moment；

Afterwards, follow the trail of frame to move in its moveable maximum magnitude, calculating tracking frame follows the trail of the image recognition information of image in frame and the correlation coefficient of the image recognition information in tracking image T1 moment, the i.e. degree of correlation of the image recognition information of two images when moving to the every bit in moveable maximum magnitude；What the maximum of described correlation coefficient was corresponding follow the trail of, and position that frame moves to is the immediate position of image recognition information of two images, for the position moved to of tracking image；Namely the position of tracking image in the eye fundus image in T2 moment is found according to the image recognition information of tracking image, as shown in Fig. 1-03.And the movement of tracking image is to be caused by the movement on optical fundus, therefore, the vector of the movement of tracking image is exactly the vector moved relative to the T1 moment in the optical fundus T2 moment.

As such, it is possible to determine the vector of the movement on optical fundus very easily, improve the specific part to optical fundus and carry out the accuracy of three-dimensional visualization imaging.In practical operation, realization tracking frame moves to the every bit in moveable maximum magnitude is unrealistic and unnecessary, follow the trail of one region during frame simultaneously, dynamic description of frameing shift is followed the trail of in order to simplify, introduce reference point, therefore, a point to follow the trail of frame is set up plane right-angle coordinate for initial point, with follow the trail of frame 210 a point for reference point, determine that reference point is along the moveable maximum magnitude of transverse axis and vertical pivot, is divided into multiple shift position along the moveable minimum spacing of transverse axis and vertical pivot by moveable maximum magnitude.So, it is achieved that follow the trail of the frame expression in the movement of its moveable maximum magnitude.

Selection for trace regions, it is possible to have multiple choices mode, as the preferred mode of one, described tracking image be the T1 moment eye fundus image in be in the region of center.It is centrally located the tracking image of position, it is simple to follow the trail of.

Concrete, described tracking image is the picture element matrix being in center of the eye fundus image in T1 moment, representing with M × N, in Fig. 1-01, in picture element matrix, the pixel of the first row first row is initial point, the orientation of the first row pixel is transverse axis forward, the orientation of first row pixel is that vertical pivot forward sets up coordinate system, the transverse axis coordinate i=0 of picture element matrix, 1, ..., M-1；The vertical pivot coordinate j=0 of picture element matrix, 1 ..., N-1；So, the pixel in picture element matrix M × N coordinate (0,1), (0,2) ... (0, N-1) ... (M-1, N-1) represents.

Concrete, in Fig. 1-02, plane right-angle coordinate is set up for initial point with the pixel following the trail of the first row first row in frame, the reference point following the trail of frame is following the trail of the moveable maximum magnitude of frame according to vector (m, n) mobile and reference point travels through each pixel, and namely each pixel is shift position.

Concrete, the movement of tracking image is moved by optical fundus and is caused, and to move be because the ocular movement of watching attentively property causes on optical fundus.Reference point is determined along the moveable maximum magnitude of transverse axis and vertical pivot by the ocular movement of watching attentively property.Further, interval (-50 are belonged to according to the moveable maximum magnitude m that the optical fundus caused because of the ocular movement of watching attentively property is moved, 50), i.e. m ∈ (-50,50), n belongs to interval (-50,50), i.e. n ∈ (-20,20), as such, it is possible to the shift position in calculating the moveable maximum magnitude following the trail of frame is 100 × 40=4000.Now, 4000 cross-correlation can be obtained.Select the vector (m corresponding to cross-correlation maximum_max,n_max), precisely due to the vector of the movement on the optical fundus of watching attentively property ocular movement generation.

Concrete, the cross-correlation formula preset particularly as follows:

$r(m,n)=\frac{{\Σ}_{i=0}^{M-1}{\Σ}_{j=0}^{N-1}\[\left(x\right(i,j)-mx)\left(y\right(i-m,j-n)-my)\]}{\sqrt{{\Σ}_{i=0}^{M-1}{\Σ}_{j=0}^{N-1}{\left(x\right(i,j)-mx)}^2}\sqrt{{\Σ}_{i=0}^{M-1}{\Σ}_{i=0}^{N-1}{\left(y\right(i-m,j-n)-my)}^2}},$

Wherein, (i, j) is the image recognition information in tracking image T1 moment to x, y (i-m, j-n) is that reference point is according to vector (m, the image recognition information after n) moving, mx is that (my is y (i-m to x for i, average j), j-n) average, (m, n) is the image recognition information in tracking image T1 moment and the cross-correlation of image recognition information when reference point moves to each shift position to r, i=0,1 ..., M-1；J=0,1 ..., N-1.

Concrete, described image recognition information is gray value.

It should be noted that, the optical fundus of the present embodiment be a kind of concrete can mobile sample, in the defining method of the movement on optical fundus, pertain only to be likely to this characteristic mobile, be not related to optical fundus he manages it other, therefore, other can be suitable for above-mentioned defining method by mobile sample, namely can extending of the defining method of the movement on the optical fundus of the present embodiment is suitable for the sample likely moved, and the present invention extends to the defining method that sample moves.

Optical fundus is described below and moves to the principle derivation of method for determination of amount:

In nature or human society, if two variablees have certain contact on the size and Orientation of development and change, then claim between variable relevant.

Crosscorrelation is relevant one, while it represents between two variablees or non-concurrent be correlated with.For one-dimensional signal, cross correlation algorithm is the standard method of the dependency for evaluating two columns.Assume there is two columns x (i) and y (i), i=1,2...N-1, then two columns are about postponing as shown in correlation coefficient r (d) equation below of d:

$r\left(d\right)=\frac{{\mathrm{\Σ}}_{i=0}^{N-1}\[\left(x\right(i)-mx)\left(y\right(i-d)-my)\]}{\sqrt{{\mathrm{\Σ}}_{i=0}^{N-1}{\left(x\right(i)-mx)}^2}\sqrt{{\mathrm{\Σ}}_{i=0}^N{\left(y\right(i-d)-my)}^2}}$

Wherein, mx, the average of my respectively two columns, the span of r (d) is [-1,1], and r (d)=0 represents that two columns are uncorrelated；R (d)=-1, represents that two columns are maximum negative correlation；R (d)=1, represents that two columns are maximum positive correlation.

The cross correlation algorithm of one-dimensional signal is expanded in two dimensional image, it is possible to for the identification of specific region feature in image and tracking.As shown in Fig. 1-04, first take a borderline region picture element matrix mark1 of the 1st width image, be sized to M × N.First in t (t > 1) width image, find borderline region matrix mark2 according to the position of boundary region matrix mark1 inside the 1st width image, then borderline region matrix mark1 is according to vector (m, n) borderline region matrix mark2 is moved, as shown in Fig. 1-04.

Then mark1 and mark2 is made crosscorrelation, shown in equation below:

$r(m,n)=\frac{{\Σ}_{i=0}^{M-1}{\Σ}_{j=0}^{N-1}\[\left(x\right(i,j)-mx)\left(y\right(i-m,j-n)-my)\]}{\sqrt{{\Σ}_{i=0}^{M-1}{\Σ}_{j=0}^{N-1}{\left(x\right(i,j)-mx)}^2}\sqrt{{\Σ}_{i=0}^{M-1}{\Σ}_{i=0}^{N-1}{\left(y\right(i-m,j-n)-my)}^2}}$

(m n) is cross-correlation result to the r obtained.Find r maximum r_maxCorresponding vector (m_max,n_max), then this region vector of movement for the 1st width image in t width image is (m_max,n_max), the horizontal range of movement is | m_max|, the vertical dimension of movement is | n_max|。

Embodiment two

Carry out in the process of three-dimensional imaging utilizing means of optical coherence tomography, it is necessary to sample is carried out OCT scan and gathers data.

If sample is mobile, carry out optical coherence tomography galvanometer system not being corrected before and after sample moves of OCT scan, it will cause that the position of Sample Scan is changed by the camera lens of the optical coherence tomography galvanometer system of OCT scan.Therefore, to can mobile sample, need the vector finding sample to move before scanning, the corresponding optical coherence tomography galvanometer system adjusting OCT scan, the same position of sample is scanned gathering data by the camera lens making the optical coherence tomography galvanometer system of OCT scan, and implementing is realized by the collecting method of embodiment two.

The enforcement of the present invention provides a kind of collecting method, comprises the following steps:

Obtain sample at sample image formed by the T1 moment, select the predeterminable area of sample image formed by the T1 moment as trace regions, to set up plane right-angle coordinate with trace regions point for initial point；Per sample at sample image formed by the T1 moment, it is determined that the position of shooting test object, shooting test object is carried out first scan and gathers data；

Determining that trace regions is at the transverse axis of plane right-angle coordinate and the moveable maximum magnitude of vertical pivot, moveable maximum magnitude is divided into multiple shift position by the minimum spacing moved along transverse axis and vertical pivot；

According to vector, (m, during n) mobile and vector traversal shift position, with the vector of shift position trace regions one to one in its moveable maximum magnitude to calculate trace regions；

According to cross-correlation formula, calculate the position at trace regions T1 moment place and trace regions according to vector (m, the cross-correlation of the position after n) mobile, the vector (m of the trace regions that the maximum of described cross-correlation is corresponding_max,n_max) vector that moves relative to the T1 moment for the sample T2 moment；Wherein, the T2 moment is later than the T1 moment；

The optical coherence tomography galvanometer system T2 moment per sample is corrected relative to the motion-vector in T1 moment, and shooting test object is carried out scanning collection data again.

So, decrease sample and move the impact on optical coherence tomography galvanometer system scanning collection data, improve the accuracy of optical coherence tomography galvanometer system scanning imagery.

Embodiment three

Embodiments of the invention additionally provide a kind of method utilizing means of optical coherence tomography to measure dynamic contrast and estimate horizontal flow.

Utilize means of optical coherence tomography to measure dynamic contrast and the method estimating horizontal flow, said method comprising the steps of,

The collecting method adopting embodiment two carries out data acquisition, wherein, use optical coherence tomography system that sample carries out repeatedly B-scanning collection data, in the Multiple-Scan of transverse area, comprise each described B-scanning, carrying out data acquisition, data at least include phase information, strength information；

The determination of phase place delta data, wherein the determination of phase variance is based on the data of B-scanning collection, and, the motion contrast of sample depends on phase variance.

Concrete, wherein determine that phase place delta data comprises the following steps: to sample scanning once or even repeatedly, utilize time dependent motion phase variance differentiate and determine the movement of scattering object.

Concrete, use optical coherence tomography system determines the time fluctuation of the data of collection, and determines the motion contrast based on time fluctuation.

Concrete, estimate the change of refractive index at flow region, be simultaneously based on above estimation and determine motion contrast.

Concrete, estimate that in one or more phase contrast image, refractive index starts the time point of change and the composition of the refractive index estimated in flow region.

Concrete, B-scanning repeatedly should include MB-scanning, and BM-scans, or both has.

Concrete, wherein optical coherence tomography system should include Fourier optical coherence tomography analysis system.

Concrete, wherein Fourier optical coherence tomography is analyzed system and should be included domain optical coherence layer scanning technology, scanning source optical coherence tomography scanning technique, and optimal frequency domain be parsed into picture.

A kind of determine the method for the kinetic characteristic in different motion region in sample, including herein below, make to determine in aforementioned manners the motion contrast in multiple region in multiple regions, and differentiate the kinetic characteristic in one or more region of sample.

Concrete, one or more wherein said moving region is defined as three-dimensional.

Concrete, the motion 3D region of wherein said one or more definition includes the blood vessel of three-dimensional, target blood, or both have concurrently.

A kind of computer-readable recording medium being used for determining the computer executable command of sample motion contrast in optical coherence tomography analysis system, it is characterized in that, including herein below: using optical coherence tomography to analyze system and sample carries out B-scanning collection data repeatedly, wherein said scanning includes the data acquisition of the Multiple-Scan within the scope of transverse area；Determine that phase variance data, wherein said phase variance are based on B-scanning collection data and determine, and determine the motion contrast of sample based on phase variance；

A kind of optical coherence tomography analyzes system, and described analysis system includes:

The computer-readable media with computer executable command contrasts for the motion determining sample；

Using optical coherence tomography to analyze system and sample carries out B-scanning collection data repeatedly, wherein said scanning includes the data acquisition of the Multiple-Scan within the scope of transverse area.

Determine that phase variance data, wherein said phase variance are based on B-scanning collection data and determine, and determine the motion contrast of sample based on phase variance.

A kind of method determining sample motion contrast in optical coherence tomography analysis system, it is characterised in that described method includes: use optical coherence tomography to analyze system and sample is taken multiple scan collection data；Determine the phase variance of statistical data；It doesn't matter for wherein said phase variance and intensity data, and determine the motion contrast of sample based on phase variance.

Concrete, described one or many scanning includes B-scanning repeatedly.

First, the means of optical coherence tomography that utilizes introducing embodiments of the invention measures dynamic contrast and the background technology of the method estimating horizontal flow:

OCT is the optical image technology of a kind of non-intrusion type, can be produced the degree of depth reflected image of high-resolution sample by weak coherent light interferometer system.In the middle of various biosystems, OCT image can observe the three dimensional structure of sample interior, is that other imaging techniques can not be compared, and is not only used for observing the retina of eyes.

The visualization of blood vessel and the quantitative information of blood flow are very important for diagnosing and treating of numerous disease.In OCT image system, DopplerOCT (Doppler OCT) technology analyzing phse sensitivity is the important form of visualization of blood vessels and diagnosis.Phase place is a kind of method of high-resolution position that fathoms, and its direction is the direction reflected along the optical path of imaging system, is the cycle frequency of the light beam of optical source wavelength half.The change that wavelength is the depth location of optical source wavelength half can produce same phase measurement.Phase place change is proportional to axial flow.The flow component parallel with imaging direction is v (cos θ), and v is the speed of flow, and θ refers to flow direction and the angle in light beam of light source direction.In system, the phase noise based on local SNR determines minimum axial flow, it is prevented that when the very little visualization limiting flow of v or cos θ.Such as in the middle of retina, the direction of some flow directions and light source is almost vertical, then θ is equivalent to 90 degree, and cos θ is equivalent to 0.In these cases, flow must could could be visualized by height very by the speed of flow in this way.

The developing direction of OCT mainly develops towards fast imaging techniques, in the identical time, more large area is carried out imaging.In order to obtain quick image taking speed, DopplerOCT imaging technique only uses some continuous degree of depth method for reflection, is referred to as A-scans (typical numeral is about 5), average wherein phase place change.Between A-scans and phase measurement, limited statistical data and almost limit the minimum axial flow being able to observe that in the short time, the fastest flow can only be visualized.

For this situation, the phase variance that image has been computed does not increase extra motion contrast.Lacking extra contrast and be because phase transformation error, the signal to noise ratio of local has arranged the calculating of the phase variance in all regions, except using DopplerOCT technology in the traffic visualization region of identical speed.

Speckle analysis is conceived to the change of image intensity, has turned out in OCT field and limits normal operation.In OCT, major part is all towards this direction about the work of speckle, and the artifact of the multipath reflection generation reducing sample interior promotes picture quality.From the image of a single static, speckle analysis technology make use of the change of spatial density to go to identify flow region.These technology are merely able to analyze the region higher than image spatial resolution, and typically apply when not having OCT image to distinguish the degree of depth.In consideration of it, be necessary to propose a kind of accurate effective manner in OCT field to determine that the flow of biofluid is to carry out diagnosing and treating of disease.Especially, in OCT system, it is necessary to a kind of method of development is estimated horizontal flow speed and determines motion contrast.

In order to deepen the understanding of the present invention and understanding, below in conjunction with the drawings and specific embodiments, the invention will be further described and introduces.

The application further of this method is OCT image momentary fluctuation intensity.OCT image momentary fluctuation intensity can be used as the contrast of other forms to the flow observing within the scope of sample depth and absorption variations.

These methods and technology show motion contrast in OCT image.Motion contrast, particularly phase place change contrast can observe nano level Blang's diffusion motion and the motion of other nonfluids.Phase place change contrast can be used for distinguishing the moving region of different motion contrast, can also be determined the character of scattering campaign by phase information simultaneously.For flow region, described analysis method is capable of identify that zones of different, it is also possible to characterize kinetic characteristic.Quantitative flow is estimated to can determine flow, relative to the independent, direction of imaging direction.The shade of phase place change contrast and cymomotive force calculate and are used for determining the refractive index of flow region and the change of absorbance.

Described herein is a kind of poor efficiency, advanced IT application diagnostic method, and is high efficiency 3-D scanning acquisition method.Efficient method can make moving region three-dimensional visualization, for instance the blood vessel of sample.The motility of these inventive methods is in that to be capable of identify that intermittent flow region, for instance differentiate the motion of erythrocyte in capillary vessel.This ability differentiating intermittent blood flow region can assisted diagnosis and treatment patient in need.

Invention also contemplates that a kind of executable equipment of computer, be used for determining the motion contrast of sample.This computer can perform equipment, uses OCT equipment that sample carries out scanning once or even repeatedly, acquisition phase delta data, it is determined that based on the motion contrast of phase place change.It is gather cymomotive force or speckle information to determine the motion contrast of sample that this computer can perform the application further of equipment.Time dependent phase variance can differentiate and determine the mobile scattering object in OCT sample image.The executable equipment screen display motor region of computer and quantitative Diagnosis scattering object.

The equipment that a kind of OCT equipment includes computer executable command determines that the motion contrast of sample is contemplated.The OCT equipment of equipment can be performed including computer and sample is carried out scanning once or even repeatedly, acquisition phase delta data, it is determined that based on the motion contrast of phase place change.Comprising computer, can to perform that the OCT equipment of equipment applies further be gather cymomotive force or speckle information to determine the motion contrast of sample.Time dependent phase variance can differentiate and determine the mobile scattering object in OCT sample image.OCT equipment screen display motor region and the quantitative Diagnosis scattering object of equipment can be performed including computer.

The detailed description of invention:

OCT system in this proposition is a kind of spectral domain optical coherence tomography (SDOCT), and step is as it is shown in figure 1, light beam is divided into reference arm and sample arm by fibre optic interferometer.Collection described herein and analytical technology do not rely on already with OCT system, only with the degree of depth of each sample reflection speed, intensity and phase place relevant, be referred to as A-scan.

Phase change A Φ (z_i, T) and it is degree of depth z_iWith the function of interval T, in conjunction with influence factor once:

ΔΦ(z_i, T) and=ΔΦ_{motion,scanerer}(z_i,T)+ΔΦ_motion,bulk(T)+ΔΦ_error,SNR(z_i)+ΔΦ_error,other(z_i)

Phase change A Φ (z_i, T) not only only include degree of depth z_iThe independent scattering object motion ΔΦ at place_{motion,scatterer}(z_i, T) define, but include between sample and system the relative motion ΔΦ of whole fluids axially equally_motion,bulk(T)。ΔΦ_error,SNR(z_i) it is degree of depth z_iThe phase error of the SNR data that place calculates.The experimental result delivered has turned out out the noise ratio of test local signal and determines to measure the accuracy that phase place changes:

${\mathrm{\σ}}_{\mathrm{\Δ}\mathrm{\φ}SNR-error}\left(z\right)=1/\sqrt{SNR\left(z\right)}$

ΔΦ_error,other(z_i) including other phase transformation error, these errors are probably derived from the measurement of OCT phase place, but are not limited to lateral cross section scanning error, and the transverse movement of sample is, or be limited to the artifact that the sample axially-movable of sampling depth produces.In order to differentiate the motion ΔΦ of scattering object_{motion,scatterer}(z_i, T), it is necessary to eliminate or reduce the impact of phase noise of other forms.Fluid motion between degree of depth reflection and other parts of imaging system of sample, calculates each of which phase place.Independent measurement phase place changes, and is the moving region of mass motion and the indivedual independently moving that impossible distinguish sample and system axial.Not eliminating the impact of mass motion, the minimum motion that can measure of sample will be limited by the fluid mass motion of system and sample motion.

One of them may be used to determine the method for relative motion of sample fluid, it is simply that uses extra Motion Measuring Equipment, for instance interferometer, is used for determining reflex motor the strongest in sample.Fourier optical coherence tomography system, comprises domain optical coherence tomoscan and swept-source optical coherence tomography scanning (reference optical frequency image), and all of depth information of sample is measured in the same time.All degree of depth reflective informations are such as through phase change information acquisition, it is more likely that eliminate the relative motion of fluid.

In some cases, in order to eliminate relative motion, a branch of strong reflection light of sample is used as the reflection light of fixing reference.Many examples do not use the fixation reflex device of high reflectance, and therefore the degree of depth of whole sample must be used to calculate fluid motion.There is several method can move from all of depth analysis phase change information to calculate sample fluid.The big phase noise relevant with the degree of depth does not reflect or signal is near the grade of noise, and due to the signal conditioning that these are low, the meansigma methods of all of phase place change can distortion.The analysis of threshold of phase data can reduce the impact of these patterns.The computation schema of phase place change also is able to be used to determine that fluid moves, and can calculate fluid motion accurately by the parameter of computation schema.

Weighted average calculation allows the motion of the fluid to some sample examples to estimate.The calculating of the fluid motion in this method is summarized as ΔΦ_motion,bulk(T)=∑ [w (z_i)ΔΦ(z_i,T)]/∑[w(z_i)], weighter factor w (z_i) depend on imaging contexts.Weighter factor is by linear OCT intensity I²(z_i) determine, this intensity depends on the reflection that actionless sample interior is the strongest.Use OCT amplitude I (z_i) weighter factor, in testing can be sensitiveer to the phase noise of low frequency signal.Threshold value is brought into weighter factor, and in a lot of situations, the phase transformation effect of noise of low frequency signal will gradually reduce.Weighting pattern comprises spatial coherence to process special sample and kinetic property.Such as, the sample fixed area under high speed flow velocity is measured phase place change and be there will be unstable condition, and this is accomplished by sample different weightings occur, or higher than flow region or lower than flow region.Weighter factor can also comprise the weighting of the shape of local strength's change of degree of depth reflection.The secondary lobe of reflectance spectrum and the phase motion of inclination can produce artifact and cause less desirable difference, and this difference can result in the distortion of the estimation of the relative motion of fluid.

Because the periodicity of phase measurement, phase place constant interval is-π to+π.Motion is interval bigger than this scope (being equal to 1/4th of light source bandwidth), and phase place interval occurs saltus step to cause mistake in computation (phase place change+π+δ is mistakened as work-π+δ).In phase test example, the fluid motion phase of sample is similar to +/-π, and phase error will cause that the phase place change profile calculated is similar with the data presented in the middle of Fig. 2.Not having the phase place change profile that additional correction is crossed, the phase place interval calculated is probably incorrect.Before the motion of fluid as stated above occurs, the distribution of phase place should be made zero again, to such an extent as to meansigma methods embodies kinestate more accurately.

From calculated phase change A Φ (z_i,T)-ΔΦ_motion,bulk(T) the fluid motion estimated is eliminated in, the summation of the variance that the variance of total quantity changes close to single factor:

σ_ΔΦ ²(z_i, T) and=σ_{ΔΦ,motion_scatterer} ²(z_i,T)+σ_{ΔΦ,SNR_error} ²(z_i)+σ_{error_other} ²(z_i)

The source spent as a comparison, phase change analysis target is scattering object motion phase version σ_{ΔΦ,motion_scatterer} ²(z_i,T).The phase transformation variances sigma of restriction SNR_{SNR_error} ²(z_i) it is determined by the noise ratio of local signal, this error is independently of the interval T's of phase test.The factor σ of last phase place change_{error_other} ²(z_i) and other phase error factor all without relation, mostly come from error and other various ΔΦs of fluid movement calculation method_error,other(z_i) impact.The phase error being limited to SNR generally can limit the visualization of scattering object motion.

By the phase measurement that scattering object moves, the part component of a lot of forms of motion can be observed, but is not limited to following component:

The change of axial flow component；

The impact of the horizontal mobility of incoherent scattering object；

The axial component of Production of Brown Type of Ammonia random motion；

The all static impact (multiple scattering body can be positioned in the high-resolution image of system and can recognise that the position of single scattering object) of incoherent scattering object.

The change of the motion of each form above-mentioned, the change at phase place interval over time and increase.In most cases, the phase error that the limiting factor of the most weak scattering object motion is each SNR restriction relevant to reflection is observed in impact.Because this phase variance is independent of the time, test waits the longer time, it is allowed to the phase place of scattering object motion varies more than the limits value of phase error.Extend interval further to may proceed to improve measurement phase variance ability.This process continues to until phase place change procedure reaches to be similar to the degree of the phase signal of completely random.Scattering object motion further strengthens the phase signal that the phase variance of test will not be made to exceed completely random.

Wherein impact on the one hand is that horizontal mobility has longer interval, causes the shade that there will be motion below flow region.This is to be caused by the refraction effect in horizontal mobility region.The remarkable change being based on given reflector locations of test of phase place in OCT, it is the change of optical path.Therefore, in whole period of time T, by also to measure the variance of all of refractive index while all of degree of depth of reflected light measurement sample.

$\mathrm{\Δ}\mathrm{\φ}(z,T)=\frac{4\mathrm{\π}}{{\mathrm{\λ}}_0}\left({\∫}_0^z\mathrm{\Δ}n\right(z^{\′},T){\mathrm{dz}}^{\′}+n(z\left)\mathrm{\Δ}z\right)$

For a fixing reflector, at average refractive index variableUnder, extend degree of depth z_n, the phase place change of calculating is:

$\mathrm{\Δ}\mathrm{\φ}(z,T)=\frac{4\mathrm{\π}}{{\mathrm{\λ}}_0}\mathrm{\Δ}\stackrel{\‾}{n}\left(T\right)z_n$

Such as, the method going to create the test of a kind of random phase completely, measure the blood flow of blood vessel of less than 15 microns by the method, for the wavelength of light source close to 800nm, it is necessary to minimum average B configuration refractive index variance be:

$\frac{\mathrm{\Δ}{\stackrel{\‾}{n}}_{nns}\left(T\right)}{\stackrel{\‾}{n}}\≈0.006=0.6\%$

By the knowledge of the time point being produced shade by variations in refractive index in understanding phase contrast image and flow region refractive index, horizontal flow and traffic intensity variance are able to determine.Correspondingly, the estimation of flow region refractive index variance contributes to determining motion contrast in OCT system.

Increase to prove the increase at motion variance interval over time, prove as an example with Brownian movement.The agarose well of 2% joins in Intralipid fat emulsions injection, is diluted the exercise intensity for mating agarose scattering object.Agarose is gelatin, is stable relative to the fat milk scattering object of motion.

The image-region that different intervals is different, phase place change calculations goes out phase variance.Fig. 5 A, Fig. 5 B, Fig. 6 A, Fig. 6 B, Fig. 7 A, Fig. 7 B are shown that from experiment imaging system shortest time point, and interval increases continuously.The shortest interval phase variance imaging is controlled by the phase noise of signal to noise ratio.Increasing over time, the phase variance value of calculation of the Intralipid fat emulsions injection including movement in region also increases.

In order to scattering object motion contrast is carried out imaging better, it is necessary to eliminate the phase noise of signal to noise ratio.One of which method is the phase variance utilizing different clock interval T1 and T2.If the phase error that we can assume that other is negligible, the phase variance at different time interval is calculated as follows:

σ² _Δφ(z_i,T₁)≌σ² _{Δφ,scatterer}(z_i,T₁)+σ² _Δφ,SNR(z_i)

σ² _Δφ(z_i,T₂)≌σ² _{Δφ,scatterer}(z_i,T₂)+σ² _Δφ,SNR(z_i)

T is set₂=β T₁, β is much larger than 1, it is assumed that the variances sigma of the scattering campaign in system² _{Δφ,scatterer}(z_i,T₂) much larger than σ² _{Δφ,scatterer}(z_i,T₁).Under these conditions, the basic phase contrast standard for phase variance contrast imaging is, σ² _{Δφ,scatterer}(z_i,T₂)-σ² _{Δφ,scatterer}(z_i,T₁) as follows:

σ² _{Δφ,scatterer}(z_i,T₂)-σ² _{Δφ,scatterer}(z_i,T₁)≌σ² _{Δφ,scatterer}(z_i,T₂)

Knowing the phase noise of the signal to noise ratio of form known, the mathematic expectaion inside Numerical value can be used for eliminating phase noise in image.The Numerical value of phase noise is based on the intensity of reflected signal in OCT and the noise characteristic of imaging system that describes before.Based on the degree of accuracy of the estimation to noise, this situation creates the image of similar contrast.

Change over time, phase variance can describe the kinetic characteristic of scattering campaign.Figure 10 shows it is observe the various sizes of microgranule of different time points data of the phase variance of Brownian movement in water, and in water, the diameter excursion of an independent scattering object is from 0.5 μm to 5 μm.Due to heat fluctuation, phase variance data can be used to analyze, and random motion is visualized.Expect that these scattering objects change variance over time equal to zero, it is necessary to calculate variance and motion is carried out imaging.The microgranule that diameter is 0.5 μm to 5 μm, to the very bright sense of OCT signal intensity, moves relative to scattering object, and these phase errors are negligible.For the microgranule of diameter 2 μm, when OCT signal is very weak, phase error cannot be ignored.The forms of motion estimated needs to contrast with other forms of motion, and Figure 10 shows the phase error in conjunction with forms of motion.Phase place change over time and change, gather the data of abundant phase variance, it is possible to therefrom extract exercise data.

For Brownian movement, the phase variance data (a lot of less relative to random phase variance data) of measurement, the relation such as following formula of measured forms of motion and the Brownian movement changing over little phase error:

σ_Δφ ²(z_i, T) and=A²+DT^γ

In order to make the change over time of the phase variance of measurement reach expected value, kinematic parameter (diffusion constant) D of scattering object may decide that its size.

Collecting method:

Going to identify or mark to levy moving region to gather more data, time interval must long enough.One simplest method is to wait in each lateral attitude, changes over time, obtains phase information, waits long enough for the transformation information gathering required statistical data and change over time.For the scanning technical term of OCT, A-scanning is the method utilizing the reflection measurement degree of depth for single position.Change over time is called M-scanning in same cross-section location repeatedly A-scanning.In whole cross-sectional extent, repeatedly A-scanning is called B-scanning.The process waited in each cross-section location produces M-scanning, and the M-scanning of the repetition in cross-sectional extent is called MB-scanning.

MB-scanning is that the kinetic characteristic in order to characterize scattering object the earliest is to gather required statistical data and the method for phase place change information.The flow that this feature description includes along imaging direction and the flow region of cross-sectional direction is quantitative.Other characterization information include some factors, including diffusion constant, scatter density and traffic flow information.The efficiency that the sole limitation of this method is collection data is low.Time 3 D stereo visualization in moving region needed by the method is oversize.It is the three dimensions needing scanning bigger for the visual quick acquisition method of some phase variance contrasts.It is in the same localities the passage of waiting time until three dimensions is sufficiently large produces motion variance contrast, before returning to original position, extra phase information might as well be obtained by scanning multiple position.As time goes on, carry out the method scanned of B-repeatedly at same cross section and be called BM-scanning.

In a long time interval, any image acquisition time is not sacrificed in BM-scanning, is a kind of high efficiency phase information acquisition method.This scan method is restricted in time, but the sightless of data analysis for continuous print A-scanning collection moves slowly, and this scan method is visible.Compared with MB-scan method, this scan method is limited for the quantitative analysis of flow.Figure 11 A, Figure 11 B are shown that some examples of these acquisition methods.

Figure 12 A, Figure 12 B and Figure 13 A, Figure 13 B are shown that being scanned by MB-and BM-scans the image of tail of Brachydanio rerio.But in both figures, the same region of phase variance contrast image identification, the contrast that BM-scanning produces is 4 times of MB-scanning.It is as expected, the sweep time used by BM-scanning is 40 times of the sweep time used by MB-scanning.From contrast figure it can be seen that the phase variance contrast image of BM-scanning has shade, this is because time interval is oversize causes that flow region exponentially changes.

The same area of the identical sample that Figure 12 A, Figure 12 B show with Figure 13 A, Figure 13 B.In whole imaging time, BM-scanning comprises the pixels across of about 2.5 times, is compared to MB-scanning and decreases 3 times.BM-can further reduce sweep time, by the adjustment of the quantity of lateral attitude, for calculating the adjustment of the statistical data of phase place change information variance, it is possible to reduce time of A-scan collecting system and improve the ability of system transversal scanning.

Laterally flow appraisal procedure:

The achievement in research that Park et al. delivers shows between continuous print phase measurement, when using the reflection light of the incoherent sample of transversal scanning as the part of light beam of light source, it is desirable to phase error there will be.Definition according to Park, the standard deviation of phase contrast is equal to the square root of phase contrast.

In order to create phase contrast within the shortest sweep time, the analysis result of Park is to determine the restrictive condition of phase contrast between continuous print A-scans, and transversal scanning creates a B-scanning simultaneously.Between sample irradiation and sample reflect, relative transverse movement is not mentioned to as the probability of the quantitative predication of transverse movement flow by this desired phase error.

If in period of time T, multiple A sweeps are separately performed in same lateral attitude, and relative lateral motion between reflecting due to sample irradiation and sample is it would appear that phase noise.In continuous measurement process, at same cross-section location, same light-source brightness, noise mostlys come from horizontal mobility, particularly from incoherent reflection, for instance find that class is noise-like in blood.Consider that a Gaussian beam is at focal point 1/e²Width of light beam=d, the lateral attitude identical in identical period of time T uses Method for Phase Difference Measurement.In identical period of time T, phase place change variance is to be determined by the transverse movement of scattering object Δ x, by lateral movement velocity V_xProduce.The bandwidth of definition light beam is Δ X/d, transverse movement the variance of the phase error caused is calculated as follows formula:

${{\mathrm{\σ}}^2}_{\mathrm{\Δ}\mathrm{\φ}}=\{1-\exp \[-2{\left(\frac{\mathrm{\Δ}x}{d}\right)}^2\]\}=\frac{4\mathrm{\π}}{3}\{1-\exp \[-2{\left(\frac{V_{x}T}{d}\right)}^2\]\}$

Due to other error in phase measurement process, the accuracy of this technology depends on the accuracy of the calibration standard of system and the elimination of other form phase errors.From the data that Park result of study shows, the dynamic range of the quantitative predication of horizontal flow is about 20%≤Δ x/d≤80%.The phase noise of signal to noise ratio restriction limits minimum lateral velocity and can measure by the method.The upper limit is only limitted between-π and π close to the random phase noise signal of saturation limit value.This equation is only allowed for the quantitative of flow, for instance V_xT/d >～0.8.The prolongation time can improve the dynamic range of horizontal flow measurement.

For retina image-forming, interval is 40 milliseconds, focuses on the diameter 20 μm of light beam, and the dynamic range of quantitatively horizontal flow is about 0.1mm/s to 0.4mm/s.

The dynamic range changing flow can be realized by the lateral resolution of change imaging system, or measures period of time T by changing.Such as, if retina image-forming time cycle be 10 milliseconds, beam diameter is 30 μm, if the dynamic range of horizontal flow is about 0.6mm/s to 2.4mm/s. in this case, the statistical data of phase measurement is enough, and the phase data that the sampling period is 10 milliseconds may also be used for the phase variance (measurement phase position per second) that the calculating sampling cycle is 20 milliseconds.The dynamic range of quantitative flow measurement will increase by 2 times, and dynamic range in this case is about 0.3mm/s to 2.4mm/s.

The second determines that the method for horizontal flow is that, due to the change of the refractive index in horizontal mobility region, the place that in image, flow is low can produce artifact.If it occur that time point it was determined that the depth bounds of flow region is it was determined that the refractive index mean change in region can also be determined.By the knowledge of the refractive index of flow region inscattering body it can be seen that horizontal mobility rate is confirmable.The method is likely to be useful when comprising polytype composition in flow region, such as the blood flow of blood vessel in blood plasma and hemocyte.

The method of the horizontal flow of another kind of quantitative measurement sample is combined with BM scanning and MB scanning.BM scanning phase variance contrast is the method for the mobility of the 3D region in a kind of efficient identification sample.Use BM-scanning, it is possible to identify three-dimensional blood vessel, can scan with this relative to the flow direction of imaging direction.MB scanning may identify which the blood vessel of target and optionally analyzes the flow of specific region.Depend on the flow that the average axial flow that average phase changes may decide that in region.Determining position by the method screened, the statistical data that MB scanning increases, small-sized axial flow component can calculate.The blood flow direction of known blood vessel, endovascular blood flow can calculate with geometry.Time dependent phase variance can pass through the calculating estimating to provide extra correlative flow of horizontal flow component.

Transverse movement noise eliminates:

One proposition relevant to acquisition method is accompanied by the additional noise of the growth of time.Moving when giving the scattering object more time, namely the sample acquisition more time moves.The axially-movable of sample is to be eliminated by the removing method of previously described fluid motion.The transverse movement of fluid is not supplementing of previous methods.Setting forth in document, the relative lateral motion between sample and imaging beam produces [1] that a phase error is determined by the size of the waist order of magnitude of imaging beam.That the derivation of phase noise is based in sample irrelevant scattering object it is assumed that reflection within the layer of sample, not all phase noise can so be derived, for instance retina.

Figure 14 A, Figure 14 B show the OCT image that Mouse Retina is cut into slices, the average OCT signal intensity image compared with phase variance contrast image.This contrast image has used Numerical value to eliminate signal to noise ratio phase noise and application medium filtering reduces artifact further.In the image of contrast, moving region is it can clearly be observed that include the extra shade not having the choroidal artery of any OCT signal generated below of the retinal vessel of top, artifact below blood vessel and bottom.In the image of this contrast, sample has a small amount of fluid generation transverse movement.Figure 15 is not the phase variance contrast image of that example, is same cross-section location place in retina, the image that big transverse movement after a while gathers when occurring.In this image, zone flow visualization remains possible, but a number of additional noise hinders visualization.

For processing fluid transverse movement, the method assumes that whole image contains the additional phase noise of the phase same level of the lateral attitude of all BM scanning.Owing to BM-scanning is all scan all of lateral attitude in a short period of time to gather data, all these points should experience identical motion, produces identical phase noise.After the phase noise of signal to noise ratio eliminates, all of contrast-data point is not zero, and the statistical data of picture contrast can be used to the impact attempting eliminating transverse movement.For the average value mu of non-zero contrast-data and the standard deviation sigma of image, picture contrast C (x, z) can include method of elimination and method for normalizing by a lot of method adjustment:

$\{C(x,z)-(\mathrm{\μ}+\mathrm{\α}\mathrm{\σ})\}\left\{\frac{3.3}{3.3-(\mathrm{\μ}+\mathrm{\α}\mathrm{\σ})}\right\}=C^{\′}(x,z)$

In this case, 3.3 radians are as phase variance contrast value, as the maximum expected value that a random phase is measured.Parameter alpha is adjustable in, and is used for improving the visualization in some region, and these regions likely disappear because of phase noise.When Figure 16 A, Figure 16 B show α=0 above-mentioned additive phase be removed after image.

Figure 17 A, Figure 17 B are shown that in retina contrast image the removal process of noise.Each two-dimensional contrast image summarises whole sample depth and is described as an one-dimensional straight line.The image overlay of multiple collections defines the two-dimentional contrastographic picture of this sample.The vertical line that left image shows, it was shown that scan contrast image at that time point BM and extra motion contrast noise occurs.Contrast image when right image is shown that α=0, after transverse movement elimination.In image medium vessels the change of position due in imaging process transverse movement cause.

Phase contrast image is based on, for the method for estimation of numeral elimination phase noise, the reflected signal S measuring phase noise².OCT strength signal I²It is reflected signal S²With noise signal N²Combination.The form of average OCT strength signal is as follows:

$<|\tilde{I}{|}^2>=S^2+<N^2>$

What Figure 21 described is measured phase noise and the relation of desired form.Work as S > > N, phase noise shows intended form as previously mentioned.

Other the method being used for eliminating transverse movement noise includes using extra statistical data to go to eliminate any transverse movement noise comprised in contrast image.Using an external motion tracker or OCT intensity and contrast image to analyze software, BM-scanning may identify which and eliminate obvious transverse movement.When less transverse movement occurs, repeat BM-scanning collection at the same area and can meet the requirement eliminating phase analysis process.

Intermittent flow identification:

The challenge of one motion contrast screening is the region identifying and not comprising scattering object in all time intervals.OCT needs to measure phase place according to the reflection in sample, but in some cases, is not that phase place any time all allows measured.It is good example for the requirement of sample reflectance, Brachydanio rerio and segmental vessels.The segmental vessels diameter of the embryo after being fertilized 3 days is 7-12 micron, and one of them big vessel branch is referred to as dorsal aorta.The confocal imaging of embryo shows that hemocyte can in specific position within a period of time.If the cross-section location that OCT image just scans does not have hemocyte, then Ink vessel transfusing would not have enough reflected signals to produce phase contrast signal.

Figure 18 shows several OCT image, is in the same cross-section location of different time points, the OCT scan image scanned from the tail of Brachydanio rerio to its yolk sac.In image, all of arrow specifies desired position in flow region: dorsal aorta (DA), axial vein (AV), segmental vessels (SE) and dorsal part stringer blood vessel (DLV).The position of these flow regions is inconspicuous in OCT intensity image, is lack enough absorptivities owing to lacking the intensity contrast in these regions and these thin vessels.Dorsal aorta and axial vein all occur in all images, but segmental vessels and dorsal part stringer blood vessel are not whenever all occur, result as expected.

Similar this situation, also comprises very little blood vessel, and is not all comprise motion contrast at all time points in retinal microvasculature.For any random phenomenon, visualization chance repeatedly and the increase of statistical data, whole event can be helped to visualize.BM scanning collecting method efficiency is high, expects the region having intermittent flow with the repeatable scanning of the method, such as blood capillary, for the visualization of motion contrast.

Intensity, speckle contrast:

Up to the present all of relative analysis method is all referring to the time dependent phase variance of scattering object.With identical acquisition method, also there is the available OCT intensity information data elapsed over time simultaneously.The character of many samples can cause the fluctuation of image: such as the change of optical coupling, power-supply fluctuation, and change over polarization variations relative in interferometer and can result in fluctuation.The example of the kinetic fluctuation of sample includes, but are not limited to:

The reflection interference of multiple obstacles in the resolution of system, as the interference that the little scattering object of Ink vessel transfusing independence reflects.

Variance based on reflecting part time dependent in flow region changes.

As time goes in the change of the absorbance of blood flow region.

As time goes on, the fluctuation of intensity and/or speckle analysis can analyze phase variance from identical collection data simultaneously.Assuming the transverse movement that there is fluid at a high speed in sample, intensity analysis is for identifying that flow locations is more useful.One at that time can the analytical technology of strength information be OCT intensity variance.In order to variance contrast is carried out rational imaging, its result is that strength information must standardization (such as, meansigma methods, median, maximum or minimum intensity level).The strength fluctuation announced and/or spot-analysis technology, as a comparison, use normalization space wave, the spatial resolution of restriction contrast image in single image.In one period of time T, Strength Changes variance is by the change (above-mentioned) of the contrast identified in moving region and absorbance.The strength fluctuation (such as, couple variations) of above-mentioned system also can cause the contrast of static region in image to be observed.One method attempting to reduce this unnecessary fluctuation is to use the Numerical value of the expection fluctuating margin based on OCT intensity and sample structure.The form of the fluctuating margin that one of which is estimated is:

$\{{{\mathrm{\&sigma;}}_{{\mathrm{\&Delta;I}}^2}}^2(z_i,T)-f(z_i,I^2(z_i\left)\right)\}/\left(I^2\right)$

The time-domain information analytical technology of another kind of fluctuating margin is the time domain distribution utilizing Fourier transformation to determine fluctuation.The shape of the frequency spectrum of Fourier transformation, bandwidth and amplitude information can be used for identifying average diameter and the diffusion constant that multiple mobility parameter includes but not limited to scattering object.

Shown in Fig. 1 is the frame diagram of SDOCT system.Low-coherence light source S (k) is divided into reference arm and sample arm by fibre optic interferometer.Reflection light is assembled in spectrogrph and tests, thus the degree of depth of reflection profile is computed；The phase place delta data of the simulation of the fluid mass motion of to be radian the be π (1/4th of imaging source wavelength) shown in Fig. 2；Shown in Fig. 3 is scattering object intended phase variance test result figure within a period of time of motion；What Fig. 4 A showed is the overview diagram that average OCT intensity image non-to sample in Fig. 4 B is relevant；In order to obtain this image, the agarose well of 2% is dissolved in the fat emulsion solution that density matching is 0.1%；In image, luminance contrast is confined to the edge of flowing and the boundary of air；Fig. 5 A is the phase place variation diagram of period of time T=40us；Fig. 5 B is the phase place variation diagram of period of time T=80us；Fig. 6 A is the phase place variation diagram of period of time T=200us；Fig. 6 B is the phase place variation diagram of period of time T=400us；Fig. 7 A is the phase place variation diagram of period of time T=800us；Fig. 7 B is the phase place variation diagram of period of time T=1.6ms；

Fig. 8 A is the phase place change comparison diagram of maximum phase transformation period cycle T 2=40T1；Fig. 8 B is the phase place change comparison diagram of maximum phase transformation period cycle T 2=20T1；Fig. 9 A is the phase place change comparison diagram of maximum phase transformation period interval T2=10T1；Fig. 9 B is the phase place change comparison diagram of maximum phase place transformation period interval T2=5T1；Shown in Figure 10 is the phase place delta data of single scattering object in water；Shown sphere diameter respectively 0.5um, 2um and 5um.Diameter is the impact that the spheroid of 2um demonstrates phase error, and it is weak to be primarily due in the phase place delta data of desired form OCT signal；What Figure 11 A showed is the image of the lateral scanning pattern of MB-scanning；The image of the lateral scanning pattern of the BM-scanning shown in Figure 11 B；Shown in Figure 12 A is the OCT gray level image of the Brachydanio rerio tail using MB-scan mode to obtain；Shown in Figure 12 B is the phase place change contrast image using MB-scan mode to obtain Brachydanio rerio tail；The size of image is 900um*325um.T2=1ms, T1=40us；Shown in Figure 13 A is the OCT gray level image of the Brachydanio rerio tail obtained with BM-scan mode；Shown in Figure 13 B is the phase place change contrast image of the Brachydanio rerio tail using BM-scan mode to obtain；The size of Figure 13 A and Figure 13 B image is all 815um*325um；Phase error in Figure 13 B eliminates, and the period of time T 2 of contrast image is estimated as 40ms；Note that the areas imaging size of phase place change contrast image is 4 times of the contrast image of the MB-scanning shown in Figure 12 B；Shown in Figure 14 A is the mean flow rate image of the time point that low speed transverse movement occurs, and this image sources is in the data of the BM-amphiblestroid transverse movement scanned；Shown in Figure 14 B is the phase contrast image after noise elimination and medium filtering, and the BM-deriving from retina transverse movement scans data；Figure 14 B is the phase contrast image of the time point that there is a small amount of fluid transverse movement；Shown in Figure 15 is the phase place change contrast image of uncorrected a large amount of fluid transverse movement；Figure 16 A and Figure 16 B is the comparison diagram of one group of phase contrast image；Shown in Figure 16 A is the image of a large amount of transverse movement of uncorrected existence；Shown in Figure 16 B be α=0 when, the image of corrected a large amount of transverse movements；Figure 17 A and 17B is the contrast summary image through the 2.6S retina degree of depth collected；Figure 17 A is without the image compensated before α=0；Figure 17 B is the image after α=0 through overcompensation；Shown in Figure 18 A is that BM-scans average OCT gray-scale map；Figure 18 B, 18C, 18D are three phase place change contrast figure of different time points；Each image is all collect within 50ms.The region of arrow points is spine longitudinal direction angiosomes, segmental vessels (Se) two kinds different, spine aorta (DA), axial vein (AV)；Shown in Figure 19 A and 19B is the OCT visual image of the Brachydanio rerio being quickly fertilized for 3 days；Figure 19 A is the MIcrosope image in the bright visual field, and Figure 19 B is the image of the cabrilla being quickly fertilized for 3 days, all illustrates the anatomical features of desired cabrilla；The line drawn in Figure 19 A and 19B represents the scanning area of OCT image；Further analyze average discharge and phase place changes the quality that can promote these images；The OCT gray level image of Figure 20 A illustrates the internal structure of cabrilla in Figure 19；Shown in Figure 20 B is the blood flow of endocardial, and this image has reduced phase noise and OCT image pixel in image；Phase contrast image after the elimination phase error impact that Figure 20 C describes；Figure 20 B clearly describes the outward appearance of heart, the flow direction of the yolk sac that the direction of arrow shown in Figure 20 C refers to, and the region expected in Figure 19 B and matches；It it is the sketch in the direction of flow relative to imaging source direction shown in Figure 21；The axial flow component that the DopplerOCT conceptions of technology observe is used to use V_zDemarcate；Figure 22 is shown that the phase noise of SNR-restriction and the graph of a relation of average OCT strength signal；The contrast image of the cabrilla tail that the data using the collection of MB-scan mode shown in Figure 23 A, 23B, 23C and 23D produce；Figure 23 A is the gray level image of structural information；Arrow in Figure 23 A point out desired by the region that detects, this region is main two blood vessels of fish, spine tremulous pulse and axial blood vessel；Figure 23 B is phase place change contrast images, and the time cycle is T2=1ms and T1=40us, and radian changes from 0 to 2, and viewed region is sightless on DopplerOCT image, and DopplerOCT can only observe sufficiently stable static region.Arrow in Figure 23 B point out desired by the moving region observed.Figure 23 C is Doppler flow diagram picture, and the scope of application is+-0.12 radian=+-200um/s, and phase place change meansigma methods is 5；Shown in Figure 23 D is Doppler flow diagram picture, and the scope of application is+-0.12 radian=+-200um/s, and phase place change meansigma methods is 100.More and more stable static characteristic improves visual quality, and the position understanding each several part in advance is conducive to visual research；Shown by Figure 24 A and 24B is the same area of Brachydanio rerio in Figure 23 A, 23B, 23C and 23D, but, the BM-scan mode using trace interval to be T=10ms；In order to select best parameter, the time used by the view data of 200 pixels across that gathers is 50ms；Gather the dynamic range of data owing to reducing DopplerOCT method, use the image of the method collection no longer to present；Compared with the phase variance contrast image of OCT mean flow rate figure and Figure 24 B of Figure 24 A, 5 B-scanning.Each durection component carries out medium filtering, radian excursion 0 to 3；The arrow of each image points out the region that spine master pulse is relevant with axial blood vessel.Figure 24 B is phase place change contrast image, it is possible to clearly observe the MB-same area scanned, but owing to the change of blood vessel refractive index has extra shade below blood vessel；Shown in Figure 25 is the phase contrast summary image of the slices across of cabrilla cardiac position in 2.6s.Each time point collects in 50ms；This kind of method is equally used in Figure 17；Change in blood vessel and heart can be visible in detail；Shown in Figure 26 is the data image of time dependent heart contrast changing value.The change of contrast, relevant with the change of cabrilla normal cardiac rate in region and the changes in flow rate of viewing area；It it is the contrast changing over certain lateral attitude of cabrilla shown in Figure 26；This contrast sums up 3 pixels across (7.2um), covers the entire depth of cabrilla；Shown in Figure 27 A, 27B, 27C is the direct picture of cabrilla heart；Figure 27 A is OCT intensity sum graph picture in logarithmic range；Figure 27 B is phase place change sum graph picture, improves visualization compared to Figure 27 C；The confocal images of the cabrilla Green Fluorescent Protein of Figure 27 B and quick fertilization in 3 days has similar area；Shown in Figure 27 C is inject the Confocal Images that fluorescent material produces in the cabrilla body at similar age, and Figure 27 B motion contrast images shows relevant blood-vessel image；Shown in Figure 28 A is average OCT gray level image；Shown in Figure 28 B is that MB-scans the amphiblestroid doppler flow spirogram picture of mouse；The Doppler flow diagram picture of Figure 28 B, it does not have use any threshold value, excursion is+-2.5mm/s.The main fast flow of the retinal vessel observed is axial flow, but uses this image analysis technology to observe suprachoroid vascular flow；Figure 28 A shows the phase place change contrast image at a series of phase place transformation period interval；Extend phase place transformation period interval and can improve blood vessel (choroidal artery) visualization, but also increase the contrast shade below blood vessel simultaneously.The prolongation of interval equally also increases the contrast in the sensitivity of transverse movement and vertical direction；Figure 29 A is the phase place change contrast figure of MB-scanning, and interval is 40us, and phase place is changed to 10；Figure 29 B show MB-and scans phase place change contrast figure, and interval is 160us, and phase place is changed to 40；Figure 29 C show MB-and scans phase place change contrast figure, and interval is 240us, and phase place is changed to 40；Figure 29 D show MB-and scans phase place change contrast figure, and interval is 320s, and phase place is changed to 40；Shown in Figure 29 A, 29B, 29C and 29D is the retinal images after flattening, is mainly used to eliminate optical path and changes the impact of the change causing mouse eyeball curvature.Amphiblestroid flattening and identify that layer of retina separate confinement is the method for two kinds main extractions information in image, information retrieval in the depth areas of three dimensional grey scale image and contrast-data, is analyzed and is gathered data, the image in formation transverse direction or front；Shown in Figure 30 A is the B-scanogram before retina flattening does not rearrange；Shown in Figure 30 B is the B-scanogram after retina flattening rearranges；Figure 31 A, 31B and 31C are the amphiblestroid BM-scanogram of mouse；Figure 31 A is mean intensity scanogram；Figure 31 B is phase place modified-image, and this image does not eliminate any mathematical phase error, also without filtering；Figure 31 C is the phase place change contrast image after noise eliminates and after medium filtering, and excursion is 0 to 3 radians；Shown in Figure 32 A is the direct picture of whole retina BM-scanning；Shown in Figure 32 B is the front phase contrast image of whole retina BM-scanning；The direct picture of the total intensity shown in Figure 33 A, Figure 33 B be retina the first half phase place change contrast image；The retinal vessel of the contrast image display surface of Figure 33 B, arrow indication is the region of blood capillary；The direct picture of total intensity of display depth shown in Figure 34 A, Figure 34 B is the phase place change contrast image of retina the latter half；The contrast image of Figure 34 B shows the shade contrast of choroidal blood vessel and main retinal vessel；Arrow indication is the region of blood capillary；Shown in Figure 35 A is that BM-scans front summation gray level image, and shown in Figure 35 B is the BM-phase place change contrast image scanning that front is total.Figure 35 A is whole amphiblestroid image, but the contrast image of Figure 35 B, it is only the image of BM-scanning retina the first half that pixels across is 200；The pixel of the single contrast image of Figure 35 B is 200*51, by the image of the whole independent scanning collection of amphiblestroid top area.What arrow was pointed out is some less visible blood vessels；Total phase contrast image in the front shown in Figure 36 A, 36B, 36C and 36D, pixels across is the image of amphiblestroid the first half of the BM-scan mode collection of 100；System being used alone BM-and sweeps some images of continuous scanning collection, the pixel of these images is 100*50, concludes therefrom that on retina, the contrast of part is 100*100；Arrow marks some visible blood capillaries, but these blood vessels are because the intermittence of its contrast is not all can occur on other image；The transversal scanning region of image is the same with the region of Figure 34 A and 34B；What two shown in Figure 37 A and 37B kind were different repeats average contrast's image of the acquisition method statistics of BM-scanning, for the image of retina the first half；What Figure 37 A and 37B gathered is the image being orthogonal to main horizontal scanning direction.Image can be inferred that pixel size is 100 × 100 as a comparison.

Obviously, the embodiment of the present invention can be carried out various change and modification without deviating from the spirit and scope of the present invention by those skilled in the art.So, if these amendments of the present invention and modification belong within the scope of the claims in the present invention and equivalent technologies thereof, then the present invention is also intended to comprise these change and modification.

Claims (22)

1. a sample move to method for determination of amount, it is characterised in that comprise the steps:

Obtain the sample image in T1 moment, select the predeterminable area of sample image in T1 moment as tracking image, obtain tracking image for the image recognition information identified and the position in tracking image T1 moment；

Obtain the sample image in T2 moment, find the position identical with the position in tracking image T1 moment as following the trail of frame in the sample image in T2 moment；

Plane right-angle coordinate is set up for initial point with the point following the trail of frame, with follow the trail of frame a point for reference point, determine that reference point is along the moveable maximum magnitude of transverse axis and vertical pivot, is divided into multiple shift position along the moveable minimum spacing of transverse axis and vertical pivot by moveable maximum magnitude；

Calculating reference point is according to vector (m, when n) moving and travel through shift position, the correlation coefficient of the image recognition information of image in frame is followed the trail of, the vector (m that the reference point that the maximum of described correlation coefficient is corresponding moves when the image recognition information of T1 moment tracking image and reference point move to each shift position_max,n_max) for the vector that the sample T2 moment moved relative to the T1 moment, the T2 moment is later than the T1 moment.

2. sample according to claim 1 move to method for determination of amount, it is characterised in that described tracking image is the region being in center of the sample image in T1 moment.

3. sample according to claim 1 move to method for determination of amount, it is characterised in that described tracking image is the picture element matrix being in center of the sample image in T1 moment, represents with M × N.

4. sample according to claim 3 move to method for determination of amount, it is characterized in that, plane right-angle coordinate is set up for initial point with the pixel following the trail of the first row first row in frame, following the trail of the moveable maximum magnitude of frame, according to vector, (m, n) mobile and reference point travels through each pixel to the reference point of tracking frame.

5. sample according to claim 4 move to method for determination of amount, it is characterised in that the formula of cross correlation algorithm particularly as follows:

Wherein, (i, j) is the image recognition information in tracking image T1 moment to x, y (i-m, j-n) is that reference point is according to vector (m, the image recognition information after n) moving, mx is that (my is y (i-m to x for i, average j), j-n) average, (m, n) is the image recognition information in tracking image T1 moment and the cross-correlation of image recognition information when reference point moves to each shift position to r, i=0,1 ..., M-1；J=0,1 ..., N-1.

6. sample according to claim 5 move to method for determination of amount, it is characterised in that described image recognition information is gray value.

7. a collecting method, it is characterised in that comprise the following steps:

Obtain the sample image in T1 moment, select the predeterminable area of sample image in T1 moment as tracking image, obtain tracking image for the image recognition information identified and the position in tracking image T1 moment；Sample image according to the T1 moment, it is determined that the position of shooting test object, carries out first scan to shooting test object and gathers data；

Obtain the sample image in T2 moment, find the position identical with the position in tracking image T1 moment as following the trail of frame in the sample image in T2 moment；

Plane right-angle coordinate is set up for initial point with the point following the trail of frame, with follow the trail of frame a point for reference point, determine that reference point is along the moveable maximum magnitude of transverse axis and vertical pivot, is divided into multiple shift position along the moveable minimum spacing of transverse axis and vertical pivot by moveable maximum magnitude；

Calculating reference point is according to vector (m, when n) moving and travel through shift position, the correlation coefficient of the image recognition information of image in frame is followed the trail of, the vector (m that the reference point that the maximum of described correlation coefficient is corresponding moves when the image recognition information of T1 moment tracking image and reference point move to each shift position_max,n_max) for the vector that the sample T2 moment moved relative to the T1 moment, the T2 moment is later than the T1 moment；

The optical coherence tomography galvanometer system T2 moment per sample is corrected relative to the vector of the movement in T1 moment, and shooting test object is carried out scanning collection data again.

8. utilize OCT to measure dynamic contrast and the method estimating horizontal flow, it is characterised in that said method comprising the steps of,

The collecting method described in claim 7 is adopted to carry out data acquisition, wherein, use optical coherence tomography system that sample carries out repeatedly B-scanning collection data, in the Multiple-Scan of transverse area, comprise each described B-scanning, carrying out data acquisition, data at least include phase information, strength information；

The determination of phase place delta data, wherein the determination of phase variance is based on the data of B-scanning collection, and, the motion contrast of sample depends on phase variance.

9. the method utilizing OCT measurement dynamic contrast and the horizontal flow of estimation according to claim 8, it is characterized in that, wherein determine that phase place delta data comprises the following steps: to sample scanning once or even repeatedly, utilize time dependent motion phase variance differentiate and determine the movement of scattering object.

10. the method utilizing OCT measurement dynamic contrast and the horizontal flow of estimation according to claim 8, it is characterised in that use optical coherence tomography system determines the time fluctuation of the data of collection, and determines the motion contrast based on time fluctuation.

OCT is utilized to measure dynamic contrast and the method estimating horizontal flow 11. according to claim 8, it is characterised in that to estimate the change of refractive index at flow region, be simultaneously based on above estimation and determine motion contrast.

OCT is utilized to measure dynamic contrast and the method estimating horizontal flow 12. according to claim 11, it is characterized in that, estimate that in one or more phase contrast image, refractive index starts the time point of change and the composition of the refractive index estimated in flow region.

OCT is utilized to measure dynamic contrast and the method estimating horizontal flow 13. according to claim 11, it is characterised in that B-scanning repeatedly should include MB-scanning, and BM-scans, or both has.

OCT is utilized to measure dynamic contrast and the method estimating horizontal flow 14. according to claim 8, it is characterised in that wherein optical coherence tomography system should include Fourier optical coherence tomography analysis system.

OCT is utilized to measure dynamic contrast and the method estimating horizontal flow 15. according to claim 14, it is characterized in that, wherein Fourier optical coherence tomography analysis system should include domain optical coherence layer scanning technology, scan source optical coherence tomography scanning technique, and optimal frequency domain is parsed into picture.

16. one kind is determined the method for the kinetic characteristic in different motion region in sample, it is characterized in that, including herein below, use the method in claim 8 to determine the motion contrast in multiple region in multiple regions, and differentiate the kinetic characteristic in one or more region of sample.

The method of the kinetic characteristic in different motion region in sample is determined, it is characterised in that one or more wherein said moving region is defined as three-dimensional 17. according to claim 16.

The method of the kinetic characteristic in different motion region in sample is determined, it is characterised in that the motion 3D region of wherein said one or more definition includes the blood vessel of three-dimensional, target blood, or both have concurrently 18. according to claim 17.

19. the computer-readable recording medium being used for determining the computer executable command of sample motion contrast in optical coherence tomography analysis system, it is characterized in that, including herein below: using optical coherence tomography to analyze system and sample carries out B-scanning collection data repeatedly, wherein said scanning includes the data acquisition of the Multiple-Scan within the scope of transverse area；Determine that phase variance data, wherein said phase variance are based on B-scanning collection data and determine, and determine the motion contrast of sample based on phase variance.

20. an optical coherence tomography analyzes system, it is characterised in that described analysis system includes:

The computer-readable media with computer executable command contrasts for the motion determining sample；

Using optical coherence tomography to analyze system and sample carries out B-scanning collection data repeatedly, wherein said scanning includes the data acquisition of the Multiple-Scan within the scope of transverse area.

Determine that phase variance data, wherein said phase variance are based on B-scanning collection data and determine, and determine the motion contrast of sample based on phase variance.

21. a kind of method determining sample motion contrast in optical coherence tomography analysis system, it is characterised in that described method includes: use optical coherence tomography to analyze system and sample is taken multiple scan collection data；Determine the phase variance of statistical data；It doesn't matter for wherein said phase variance and intensity data, and determine the motion contrast of sample based on phase variance.

22. the method determining sample motion contrast according to claim 21, it is characterised in that described one or many scanning includes B-scanning repeatedly.