Abnormal blood flow distribution characterization method, device and equipment based on ultrasonic image

Abstract

The invention discloses an abnormal blood flow distribution characterization method based on an ultrasonic image. The method comprises the steps that 1, the ultrasonic image containing an abnormal blood flow signal is obtained and preprocessed; 2, extracting an abnormal blood flow signal area in the ultrasonic image; 3, characteristic parameters of the abnormal blood flow distribution area are calculated; 4, the characteristic parameters serve as independent variables, and a mathematical model between the abnormal blood flow signal characteristic parameter combination and the abnormal blood flow distribution condition is established. According to the method, blood flow abnormity can be analyzed and evaluated, and high efficiency, accuracy and expandability are achieved; compared with a traditional manual interpretation method, the method has the advantages that the efficiency and the accuracy of abnormal blood flow signal evaluation are remarkably improved, and the method can be applied to various medical imaging technologies and has a wide application prospect.

Description

Technical Field

[0001] The invention relates to a method for characterizing abnormal blood flow distribution based on ultrasonic images, and belongs to the field of medical ultrasonic imaging. Background Art

[0002] Abnormal blood flow is one of the important physiological characteristics of many diseases, especially closely related to the occurrence and progression of cardiovascular diseases, tumors, nervous system diseases and other diseases. Its morphology and distribution characteristics can reflect the occurrence, development and severity of lesions. The detection of abnormal blood flow generally adopts various medical imaging methods. Taking ultrasound imaging as an example, ultrasound images are widely used in the detection of abnormal blood flow signals due to their real-time, non-invasive and low-cost advantages. However, traditional blood flow abnormality analysis methods mainly rely on manual interpretation and experience judgment. This method has the following limitations: low efficiency, manual analysis of images requires a lot of time, and it is difficult to meet the needs of large-scale screening; strong subjectivity, different doctors may have different judgments on the same image, and the results lack consistency and objectivity; difficult to quantify, manual methods are difficult to accurately quantify abnormal blood flow signals, limiting its application in disease research and efficacy evaluation.

[0003] Based on the above problems, the present invention proposes a method for characterizing abnormal blood flow distribution based on ultrasound images. By extracting characteristic parameters of blood flow distribution through ultrasound blood flow images, combined with a quantitative analysis model, the degree of blood flow abnormality can be quantified efficiently and accurately to accurately characterize the abnormal blood flow distribution. Summary of the invention

[0004] The purpose of the present invention is to provide a method for characterizing abnormal blood flow distribution based on ultrasound images, obtain ultrasound blood flow images through an ultrasound blood flow imaging method, extract characteristic parameters of blood flow distribution therein, and establish a relationship model between the combination of characteristic parameters and the degree of blood flow abnormality, thereby realizing automated quantitative analysis and evaluation of blood flow abnormalities in images.

[0005] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0006] In a first aspect, the present invention provides a method for characterizing abnormal blood flow distribution based on ultrasound images, which comprises the following steps:

[0007] Step 1, obtaining an ultrasound image containing abnormal blood flow signals and preprocessing the image;

[0008] Step 2, extracting the abnormal blood flow signal area in the ultrasound image;

[0009] Step 3, calculating characteristic parameters of the abnormal blood flow distribution area;

[0010] Step 4: Using the above characteristic parameters as independent variables, a mathematical model between the abnormal blood flow signal characteristic parameter combination and the abnormal blood flow distribution is established;

[0011] Step 5: Analyze and quantitatively evaluate the abnormal blood flow signals in the image based on the mathematical model.

[0012] Furthermore, in step 1, the dynamic characteristics of blood flow are analyzed by using the interconvolution between the front and rear frames of the ultrasound image.

[0013] Furthermore, let the system receive two frames of echo signals s(t) and s(t+Δt), where t is time and Δt is the time delay, indicating the time interval between the previous and next frames; calculate the cross-convolution between the two frames:

[0014]

[0015] Where τ is the integration variable;

[0016] The amplitude information of the cross-convolution is used to extract the amplitude change of blood flow, reflecting the intensity of blood flow; the phase information of the cross-convolution result is analyzed to extract the phase change of blood flow, reflecting the direction and speed of blood flow.

[0017] Furthermore, in step 2, the abnormal blood flow signal is extracted by an image binarization method.

[0018] Furthermore, the abnormal blood flow area is selected in the ultrasound image, which should include an ultrasound image parallel to the nerve and an ultrasound image perpendicular to the nerve. Then, the image binarization method is used to classify the pixels as background or foreground according to the threshold set for each channel for the three color channels (red, green, and blue) of the image, that is:

[0019]

[0020] Among them, B(i,j) represents the binarized image (0 or 1), I(i,j) is the pixel value of the original image, and [T1,T2] is the given threshold range (for each color channel). The pixel value of the image is converted to 0 or 1 to extract abnormal blood flow signals.

[0021] Furthermore, in step 3, the characteristic parameters include: longitudinal length ratio, transverse width ratio, abnormal blood flow signal pixel ratio and abnormal blood flow signal pixel ratio under longitudinal scanning. The longitudinal length ratio is the ratio of the longitudinal length of the abnormal blood flow signal area to the nerve thickness; the transverse width ratio is the ratio of the transverse width of the abnormal blood flow signal area to the transverse width of the ROI area; the abnormal blood flow pixel ratio is the ratio of the number of pixels in the abnormal blood flow signal area to the total number of pixels in the ROI area. The extracted characteristic parameters are normalized so that the mean of each characteristic value is zero and the standard deviation is 1;

[0022] Furthermore, in step 4, the extracted features, the longitudinal length proportion x1, the transverse width proportion x2, the abnormal blood flow pixel proportion x3 under the longitudinal scan, and the abnormal blood flow pixel proportion x4 under the transverse scan are used as independent variables, and the blood flow distribution abnormality y is used as the dependent variable, and the model is established as follows:

[0023] y=a+b1x1+b2x2+b3x3+b4x4

[0024] Among them, a, b1, b2, b3, and b4 are model coefficients, which can be adjusted according to different application scenarios.

[0025] Furthermore, in step 5, the model established in step 4 is used to evaluate the ultrasound image containing abnormal blood flow signals, obtain quantitative results of its abnormal blood flow distribution, and realize automated quantitative analysis of blood flow abnormalities; the analysis step includes normalizing the extracted feature parameter set X=[x1, x2, x3, x4] so that the mean of each feature is zero and the standard deviation is 1; determining the model coefficients; and calculating the degree of blood flow distribution abnormality.

[0026] Furthermore, in the analysis of similar situations, for different detection objects, the ROI selection should be selected with the same position, shape, and size to ensure detection consistency.

[0027] In a second aspect, the present invention provides an abnormal blood flow analysis device for ultrasonic images, comprising:

[0028] An image acquisition module, used for acquiring an ultrasonic image containing abnormal blood flow signals and preprocessing the image;

[0029] An image extraction module, used for extracting abnormal blood flow signal areas in ultrasound images;

[0030] The model building module calculates the characteristic parameters of the abnormal blood flow distribution area and establishes a mathematical model between the abnormal blood flow signal characteristic parameter combination and the abnormal blood flow distribution situation;

[0031] The image analysis module analyzes and quantitatively evaluates abnormal blood flow signals in the image based on mathematical models.

[0032] In a third aspect, the present invention provides an electronic device comprising a processor, a memory, and a program or instruction stored in the memory and executable on the processor, wherein the program or instruction, when executed by the processor, implements the steps of the above-mentioned abnormal blood flow distribution characterization method.

[0033] In a fourth aspect, the present invention provides a readable storage medium storing a program or instruction, which implements the steps of the above-mentioned abnormal blood flow distribution characterization method when executed by a processor.

[0034] The beneficial effects of the present invention are: the method of the present invention can effectively extract the characteristic parameters of ultrasonic blood flow images, automatically evaluate the abnormality of blood flow distribution in unknown images, significantly improve the efficiency and accuracy of abnormal blood flow signal analysis, and provide quantitative analysis results, and has wide applicability and versatility. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is the overall flow chart of the method of the present invention;

[0036] Figure 2 It is a schematic diagram of extracting abnormal blood flow signal areas under longitudinal scanning mode;

[0037] Figure 3 It is a schematic diagram of extracting abnormal blood flow signal areas under cross-sectional scanning mode. DETAILED DESCRIPTION

[0038] In order to elaborate the present invention in detail, the specific implementation process of the present invention is now described in detail in conjunction with the implementation examples in the attached drawings, thereby showing the process of extracting and analyzing abnormal blood flow characteristic signals of the present invention.

[0039] This embodiment proposes a method for characterizing abnormal blood flow distribution based on ultrasound images, and its implementation process is as follows: Figure 1 shown.

[0040] First, an image containing abnormal blood flow signals is obtained using the ultrasound blood flow imaging method, which should include an ultrasound image parallel to the nerve and an ultrasound image perpendicular to the nerve. Then, the abnormal blood flow signal area is extracted from the image using the binarization image method.

[0041] by Figure 2 , Figure 3 Take the ultrasound blood flow images under two scanning modes as an example. Figure 2 (a) is the original ultrasound blood flow image without blood flow abnormality under longitudinal scanning. Figure 2(c) is the original ultrasound blood flow image containing blood flow abnormalities under longitudinal scanning. Figure 2 (b) and Figure 2 (d) The corresponding binary images after blood flow signal extraction, where the white area represents the location where abnormal blood flow occurs. Figure 3 (a) is the original ultrasound blood flow image without blood flow abnormality under cross-sectional scanning. Figure 3 (c) is the original ultrasound blood flow image containing blood flow abnormalities under cross-sectional scanning. Figure 3 (b) and Figure 3 (d) The corresponding binary images after blood flow signal extraction, where the white area represents the location of abnormal blood flow. The four extracted features, the longitudinal length ratio x1, the transverse width ratio x2, the abnormal blood flow pixel ratio x3 and the abnormal blood flow pixel ratio x4 under longitudinal scanning, are used as independent variables, and the blood flow distribution abnormality y is used as the dependent variable, and brought into the model:

[0042] y=a+b1x1+b2x2+b3x3+b4x4

[0043] Among them, a=0.29, b1=3.35, b2=0.70, b3=3.71, b4=0.19 are the coefficients of the model.

[0044] Finally, we get the image y with normal blood flow. (-) =0.29, image y containing abnormal blood flow (+) =3.55. The model can determine whether there is abnormal blood flow and quantify the degree of abnormal blood flow through the abnormal blood flow distribution degree y. The coefficients selected when forming the parameter combination can be adjusted according to the specific situation and are not limited to the coefficients used in this embodiment.

[0045] This method can be used to analyze the abnormal blood flow signal area in ultrasound images. The extracted geometric features and model parameters can be adjusted according to the specific application scenario, and can be widely used in medical imaging, biological research and other fields.

[0046] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

[0047] The parts not involved in the present invention are the same as the prior art or can be implemented by using the prior art.

Claims

1. A method for characterizing abnormal blood flow distribution based on ultrasound images, characterized in that: The steps include: Step 1, obtaining an ultrasound image containing abnormal blood flow signals and preprocessing the image; Step 2, extracting the abnormal blood flow signal area in the ultrasound image; Step 3, calculating characteristic parameters of the abnormal blood flow distribution area; Step 4: Using the above characteristic parameters as independent variables, a mathematical model between the abnormal blood flow signal characteristic parameter combination and the abnormal blood flow distribution is established; Step 5: Analyze and quantitatively evaluate the abnormal blood flow signals in the image based on the mathematical model.

2. The method for characterizing abnormal blood flow distribution based on ultrasound images according to claim 1, characterized in that: In step 1, the dynamic characteristics of blood flow are analyzed by using the mutual convolution between the front and rear frames of the ultrasound image; the two frames of echo signals s(t) and s(t+Δt) received by the system are assumed, where t is time and Δt is time delay, indicating the time interval between the front and rear frames; the mutual convolution between the two frames is calculated: Where τ is the integration variable; The amplitude information of the cross-convolution is used to extract the amplitude change of blood flow, reflecting the intensity of blood flow; the phase information of the cross-convolution result is analyzed to extract the phase change of blood flow, reflecting the direction and speed of blood flow.

3. The method for characterizing abnormal blood flow distribution based on ultrasound images according to claim 1, characterized in that: In the step 2, the abnormal blood flow signal is extracted by using an image binarization method.

4. The method for characterizing abnormal blood flow distribution based on ultrasound images according to claim 3, characterized in that: The abnormal blood flow area is selected in the ultrasound image, including an ultrasound image parallel to the nerve and an ultrasound image perpendicular to the nerve; then, the image binarization method is used to divide the pixels into background or foreground according to the threshold set for each channel for the three color channels of the image, that is: Among them, B(i,j) represents the binarized image, I(i,j) is the pixel value of the original image, and [T1,T2] is the given threshold range. The pixel value of the image is converted to 0 or 1 to extract abnormal blood flow signals.

5. The method for characterizing abnormal blood flow distribution based on ultrasound images according to claim 1, characterized in that: In step 3, the characteristic parameters include: the longitudinal length ratio, the transverse width ratio, the abnormal blood flow signal pixel ratio under the longitudinal scan, and the abnormal blood flow signal pixel ratio under the transverse scan.

6. The method for characterizing abnormal blood flow distribution based on ultrasound images according to claim 1, characterized in that: In step 4, the extracted features, the longitudinal length proportion x1, the transverse width proportion x2, the abnormal blood flow pixel proportion x3 under longitudinal scanning, and the abnormal blood flow pixel proportion x4 under transverse scanning are used as independent variables, and the blood flow distribution abnormality y is used as the dependent variable, and the model is established as follows: y=a+b1x1+b2x2+b3x3+b4x4 Among them, a, b1, b2, b3, and b4 are model coefficients, which can be adjusted according to different application scenarios.

7. The method for characterizing abnormal blood flow distribution based on ultrasound images according to claim 1, characterized in that: In step 5, the analysis step includes normalizing the extracted feature parameter set X=[x1, x2, x3, x4] so that the mean of each feature is zero and the standard deviation is 1; determining the model coefficient; and calculating the abnormality of the blood flow distribution.

8. The method for characterizing abnormal blood flow distribution based on ultrasound images according to claim 1, characterized in that: In step 5, in the analysis of the same situation, for different detection objects, the ROI selection should be selected with the same position, the same shape, and the same size to ensure detection consistency.

9. An abnormal blood flow analysis device for ultrasonic images, characterized in that: include: An image acquisition module, used for acquiring an ultrasonic image containing abnormal blood flow signals and preprocessing the image; An image extraction module, used for extracting abnormal blood flow signal areas in ultrasound images; The model building module calculates the characteristic parameters of the abnormal blood flow distribution area and establishes a mathematical model between the abnormal blood flow signal characteristic parameter combination and the abnormal blood flow distribution situation; The image analysis module analyzes and quantitatively evaluates abnormal blood flow signals in the image based on mathematical models.

10. An electronic device, characterized in that: It comprises a processor, a memory and a program or instruction stored in the memory and executable on the processor, wherein the program or instruction, when executed by the processor, implements the steps of the abnormal blood flow distribution characterization method as described in any one of claims 1 to 8.

Images