CN109222952A

CN109222952A - A kind of laser speckle perfusion weighted imaging method

Info

Publication number: CN109222952A
Application number: CN201810786671.3A
Authority: CN
Inventors: 孔平; 黄钢; 张建青; 侯丽英; 李鹏; 黄辰
Original assignee: Shanghai University of Medicine and Health Sciences
Current assignee: Shanghai University of Medicine and Health Sciences
Priority date: 2018-07-17
Filing date: 2018-07-17
Publication date: 2019-01-18

Abstract

The present invention relates to a kind of laser speckle perfusion weighted imaging methods, comprising the following steps: 1) obtain each pixel of original laser image contrasts Value Data；2) it is pre-processed and is converted to blood perfusion amount to contrasting Value Data；3) with the different corresponding blood perfusion amounts of color mark different pixels point.Compared with prior art, the present invention fast, the prominent erythrocyte of calculating speeds such as has the advantages that largely to move faster region, practicability and generalization high for poly- sum aggregate.

Description

A kind of laser speckle perfusion weighted imaging method

Technical field

The present invention relates to laser speckle images to handle analysis field, more particularly, to a kind of laser speckle perfusion weighted imaging Method.

Background technique

Laser speckle blood current imaging technology is a kind of safety, and non-contact blood flow detection technology is widely used in body surface Microcirculation monitoring, sport injury prevention, the fields such as blood vessel state evaluation.The processing analysis of laser speckle image is main free at present Between contrast analysis (Laser Speckle Spatial Contrast Analysis, LSSCA) and time contrasts analysis (Laser Speckle Temporal Contrast Analysis, LSTCA), such as: " a kind of laser speckle blood current imaging and analyzing method " (Chinese invention patent CN101485565), " a kind of laser speckle blood stream imaging processing system and method " (Chinese invention patent CN102357033), " a kind of laser speckle blood flow imaging contrast analytical method " (Chinese invention patent CN102429650) etc., all It is to be contrasted by calculating come in the sequence time-varying speckle image of the still image or consecutive variations that reflect single frames, each region The relative value of blood flow velocity.But since the variation of the blood perfusion amount of skin surface is smaller, caused speckle contrasts value variation Amount is also smaller, causes to reflect that the effective information of blood flow variation is smaller in traditional laser speckle blood-stream image, mass data range It is occupied by static background.

Summary of the invention

It is an object of the present invention to overcome the above-mentioned drawbacks of the prior art and provide a kind of laser speckle blood flows Perfusion Imaging method.

The purpose of the present invention can be achieved through the following technical solutions:

A kind of laser speckle perfusion weighted imaging method, comprising the following steps:

1) obtain each pixel of original laser image contrasts Value Data；

2) it is pre-processed and is converted to blood perfusion amount to contrasting Value Data；

3) with the different corresponding blood perfusion amounts of color mark different pixels point.

The step 1) specifically includes the following steps:

11) using laser speckle blood current imaging device acquisition original laser figure f (x, y), picture size m*n；

12) analytic approach is contrasted using space and calculates that f (x, y) is corresponding contrasts value matrix C (x, y), contrast value matrix C (x, Y) size be (m-w+1) * (n-w+1), wherein w for used sliding window window size.

In the step 12), analytic approach is contrasted in space specifically:

Using the sliding window of a w*w, the ratio between the standard deviation and mean value of all pixels gray value in sliding window are obtained, is obtained The speckle contrast of central point in the forms, and by speckle contrast assignment in the corresponding pixel of the central point, when sliding window is sliding It when crossing entire image, amounts to and carries out (m-w+1) * (n-w+1) secondary calculating, and value composition is contrasted at obtained all centers and contrasts value Matrix C (x, y).

In the step 2), the change type of ratio C and blood perfusion amount K is served as a contrast are as follows:

Wherein, T is the time for exposure of CCD camera, τ_cFor the time constant of electric field auto-correlation function, β is optical system phase The dry factor (fixed value that optical system determines).

In the step 2), the dynamic range for serving as a contrast ratio C is [0.05-0.86], and measuring accuracy is 4 after decimal point, The dynamic range of blood perfusion amount K is [0-500], and measuring accuracy is to retain integer part.

The step 3) specifically includes the following steps:

31) it constructs the lining ratio C of three decimals of a reservation and retains the retrieval table of integer blood perfusion amount K, and use The method of linear interpolation, which to contrast each of value matrix, contrasts value, and corresponding blood perfusion is indexed in retrieval table Magnitude；

32) according to retrieval table will contrast it is all in value matrix C (x, y) contrast the corresponding blood perfusion value of value, be recorded in big In the small matrix P (x, y) for (m-w+1) * (n-w+1), and matrix P (x, y) is subjected to round processing, obtains size For the matrix P ' (x, y) of (m-w+1) * (n-w+1)；

33) according to blood perfusion amount K maximum in matrix P ' (x, y)_max, the range for defining shown color-bar CB is [0, CB_max], wherein CB_max=[K_max/ 3], [] is round operation, and the three primary colors of color-bar CB are RGB, Blood perfusion amount is mapped as to different color ranks, and divides display level according to brightness to each color rank；

34) matrix P ' (x, y) is transformed into the color matrix with brightness degree according to step 33), completed using pseudo- color The mode of chromatic graph states blood perfusion amount, and the more deep more partially red region of color illustrates that blood perfusion amount is bigger, corresponding erythrocyte Assemble more, movement velocity is faster.

In the step 33), blood perfusion amount is mapped as different color ranks by following equation:

Show blood perfusion amount

In the step 33), divide display level according to brightness to each color rank, specifically:

As blood perfusion amount K₀When positioned at the color rank of green, green point is CB altogether_max/ 4 grades, it is corresponding bright Spend grade n_gExpression formula are as follows:

As blood perfusion amount K₀When positioned at red color rank, red point is CB altogether_max/ 2 grades, it is corresponding bright Spend the expression formula of grade n are as follows:

As blood perfusion amount K₀When > CBmax, then most deep red is collectively expressed as.

Compared with prior art, the invention has the following advantages that

Laser speckle perfusion weighted imaging method proposed by the present invention contrasts value analysis method based on traditional Space Speckle, By the way of the quantization blood perfusion of color grading brightness classification, erythrocyte is big in the test serum that can give prominence to the key points Amount aggregation, moves faster region, and the problem of computational accuracy deficiency, is calculated using Fast Interpolation when compensating for tradition lining ratio Analysis Method skill improves calculating speed, and the blood perfusion amount of different zones is indicated with different color ranks, and human eye is utilized Visual characteristic be exaggerated the slight change between different quantized values, have practicability and generalization.

Detailed description of the invention

Fig. 1 is inventive algorithm schematic diagram.

Fig. 2 is actual imaging effect figure.

Specific embodiment

The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.

Embodiment

The example that the invention will now be described in detail with reference to the accompanying drawings.

As shown in Figure 1, the concrete mode for the Larger Dynamic range perfusion weighted imaging that the present invention describes are as follows:

Step 1: shooting original laser speckle image using matched laser speckle blood current imaging instrument, shooting position is Human body the back of the hand under stationary state.The resolution ratio of the image is.It is every with matrix f (x, y) the storage image of (1040*1392) size The gray value of one location of pixels.

Step 2: the ratio between the standard deviation and mean value of all pixels gray value in window are calculated using the sliding window of a 5*5, Obtain the speckle contrast C of central point in the forms, and by C assignment in the corresponding pixel of the central point, when sliding window slip over it is whole When width image f (x, y), (1036*1388) secondary similar calculating is carried out altogether, and obtained all centers contrast value and constitute (1036* 1388) value matrix C (x, y) is contrasted in the space of size.

Step 3: according to the conversion regime of formula 1, in advance with 1/C²As the initial value of Newton iteration, lining ratio C is calculated With the corresponding relationship of blood perfusion amount K, and concordance list is established.

In experiment, the dynamic range of C value is smaller, and range is at [0.05-0.86], it is therefore desirable to be metered into after decimal point 4 To improve the contrast between different pixels.And K is worth dynamic range larger, range only retains whole between [0-500] Number, the difference that can be also effectively reflected between different pixels.

Table 1 illustrates the partial content of concordance list

Table 1 contrasts value table (part) corresponding with blood perfusion amount

According to the searching algorithm of definition, if contrasting an Elements C in value matrix₀=0.1297, retrieval table is consulted, C is less than₀ Maximum value be 0.1291, be greater than C₀Minimum value be 0.1302, then corresponding blood perfusion amount be 58.5454, K after round numbers₀ It is 59.

In the above described manner, retrieval obtains contrasting the corresponding blood perfusion moment matrix P (x, y) of value matrix C (x, y), and size is same Sample is (1036*1388).

Step 4: by P (x, y) carry out round processing, obtain size be (1036*1388) matrix P ' (x, y)。

Step 5: taking out maximum blood perfusion amount K in P ' (x, y)_max=145, define the model of shown color-bar CB Enclose is [0, CB_max], wherein CB_max=[K_max/ 3]=48.Colored three primary colors are RGB respectively, by blood perfusion amount according to The mode of formula 2 is mapped as different color ranks.

Show blood perfusion amount

After distributing color, then a color component is divided into several grades according to brightness, on each color grade Color corresponding to K value is assigned, by taking green as an example, in present case, green can be divided into 12 grades, as 12≤K₀When < 24, K₀ Corresponding n grades of greens, wherein n is calculated by formula 3.

N=[K₀- 12] (formula 3)

If K₀=20, then it is exactly to indicate this pixel with the 8th grade of green.

P ' (x, y) is transformed into a color matrix in the method, that is, completes and states blood with the mode of pseudocolour picture The process of purling fluence, color is deeper, and more partially red region illustrates that blood perfusion amount is bigger, gets over to corresponding erythrocyte aggregation More, movement velocity is faster.

As shown in Fig. 2, Fig. 2 is using the obtained perfusion weighted imaging figure of the present invention, shooting is hand under normal circumstances Situation is carried on the back, at finger-joint, finger tip color is deeper, meets the faster physiology reality of the two position blood flow velocities, illustrates this Method has practicability.

The present invention uses the blood perfusion amount K value of Larger Dynamic range as quantitative criteria, to do to microcirculation in human body situation Objective assessment out will not bring subjective judgement into.By it was verified that blood perfusion amount is in computational accuracy, numberical range etc. is square Face, which is better than, traditional contrasts value quantitative criteria.

The invention avoids operations each time will carry out a large amount of duplicate Newton iteration processes, can be realized pseudocolour picture Real-time display, meet the requirement of practical application, and describe and Larger Dynamic range quantized value is presented using Pseudo-color Technique Technology is counted within the scope of showing different data using three kinds of RGB different color features using the vision difference of human body Value can pull open the difference between different data range, obtain more intuitive display.

Claims

1. a laser speckle blood perfusion imaging method, is characterized in that, comprises the following steps:

1) Obtain the contrast value data of each pixel of the original laser image;

2) Preprocess the contrast value data and convert it into blood perfusion volume;

3) Mark the blood perfusion volume corresponding to different pixel points with different colors.

2. A laser speckle blood perfusion imaging method according to claim 1, wherein the step 1) specifically comprises the following steps:

11) Using a laser speckle blood flow imaging device to collect the original laser image f(x,y), the image size is m*n;

12) Using the spatial contrast analysis method to calculate the contrast value matrix C(x,y) corresponding to f(x,y), the size of the contrast value matrix C(x,y) is (m-w+1)*(n- w+1), where w is the window size of the sliding window used.

3. A laser speckle blood perfusion imaging method according to claim 2, characterized in that, in the step 12), the spatial contrast analysis method is specifically:

A w*w sliding window is used to obtain the ratio of the standard deviation to the mean value of all pixel gray values in the sliding window, obtain the speckle contrast of the center point in the window, and assign the speckle contrast to the corresponding center point. Pixels, when the sliding window slides across the entire image, a total of (m-w+1)*(n-w+1) calculations are performed, and all the central contrast values obtained form a contrast value matrix C(x,y) .

4. A laser speckle blood perfusion imaging method according to claim 1, wherein in the step 2), the conversion formula of the contrast ratio value C and the blood perfusion volume K is:

Among them, T is the exposure time of the CCD camera, τ _c is the time constant of the electric field autocorrelation function, and β is the coherence factor of the optical system.

5 . The laser speckle blood perfusion imaging method according to claim 4 , wherein in the step 2), the dynamic range of the contrast value C is [0.05-0.86], and the measurement accuracy is after the decimal point. 6 . 4 bits, the dynamic range of blood perfusion K is [0-500], and the measurement precision is reserved for the integer part.

6. A laser speckle blood perfusion imaging method according to claim 4, wherein the step 3) specifically comprises the following steps:

31) Construct a retrieval table that retains three decimal places for the contrast value C and retains the integer blood perfusion volume K, and uses the method of linear interpolation to make each contrast value in the contrast value matrix index to the corresponding value in the retrieval table. blood perfusion value;

32) Record the blood perfusion values corresponding to all the contrast values in the contrast value matrix C(x, y) according to the retrieval table in the matrix P(x) whose size is (m-w+1)*(n-w+1). , y), and round the matrix P(x, y) to obtain the matrix P′(x, y) whose size is (m-w+1)*(n-w+1);

33) According to the maximum blood perfusion K _max in the matrix P'(x, y), define the range of the displayed color bar CB as [0, CB _max ], where CB _max =[K _max /3], [ ] is the rounding operation, the three primary colors of the color bar CB are red, green and blue, the blood perfusion is mapped to different color levels, and each color level is displayed according to the brightness level;

34) According to step 33), the matrix P'(x, y) is transformed into a color matrix with a brightness level, and the blood perfusion volume is expressed in the form of a pseudo-color map. The darker the color, the redder the area indicates the blood perfusion volume The larger it is, the more red blood cells gather and the faster the movement speed.

7. A kind of laser speckle blood perfusion imaging method according to claim 6, is characterized in that, in described step 33), blood perfusion amount is mapped into different color levels according to the following formula:

8. A laser speckle blood perfusion imaging method according to claim 7, wherein in the step 33), each color level is divided into display levels according to brightness, specifically:

When the blood perfusion volume K ₀ is at the green color level, the green color is divided into CB _max /4 levels, and the expression of the corresponding brightness level n _g is:

When the blood perfusion K ₀ is at the red color level, the red color is divided into CB _max /2 levels, and the expression of the corresponding brightness level n _r is:

When blood perfusion K ₀ >CBmax, it is uniformly expressed as the deepest red.