CN117576417B - Intelligent heart ultrasonic image feature extraction method - Google Patents
Intelligent heart ultrasonic image feature extraction method Download PDFInfo
- Publication number
- CN117576417B CN117576417B CN202410063150.0A CN202410063150A CN117576417B CN 117576417 B CN117576417 B CN 117576417B CN 202410063150 A CN202410063150 A CN 202410063150A CN 117576417 B CN117576417 B CN 117576417B
- Authority
- CN
- China
- Prior art keywords
- pixel
- neighborhood
- pixel point
- neighborhood range
- ultrasonic image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000605 extraction Methods 0.000 title claims abstract description 23
- 238000000034 method Methods 0.000 claims description 53
- 238000002604 ultrasonography Methods 0.000 claims description 41
- 230000000747 cardiac effect Effects 0.000 claims description 32
- 239000013598 vector Substances 0.000 claims description 21
- 239000006185 dispersion Substances 0.000 claims description 5
- 210000005003 heart tissue Anatomy 0.000 description 15
- 238000003745 diagnosis Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 210000001519 tissue Anatomy 0.000 description 2
- 238000010521 absorption reaction Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 239000000758 substrate Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30048—Heart; Cardiac
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Medical Informatics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Ultra Sonic Daignosis Equipment (AREA)
Abstract
The invention relates to the technical field of image processing, in particular to an intelligent heart ultrasonic image feature extraction method, which comprises the following steps: collecting a heart ultrasonic image; acquiring a neighborhood range of each pixel point in the heart ultrasonic image; acquiring the possibility that the neighborhood range of the pixel point is positioned at the edge of the heart ultrasonic image according to the heart ultrasonic image and the neighborhood range of the pixel point; and calculating the similarity between the neighborhood ranges of the two pixels according to the distance between the two pixels in the heart ultrasonic image and the gray value of the pixel in the neighborhood range of the two pixels. According to the invention, the pixel points are corrected in a local range according to the similarity between the pixel points and the possibility that the neighborhood range of the pixel points is positioned at the edge in the heart ultrasonic image, so that the detail characteristics in the heart ultrasonic image are maintained to a great extent while noise is removed.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to an intelligent heart ultrasonic image feature extraction method.
Background
Heart ultrasound is a safe and reliable image examination method, and doctors can observe the heart through heart ultrasound images to make accurate diagnosis. However, when the heart ultrasonic image is acquired, a great amount of noise exists in the heart ultrasonic image due to the factors of machine equipment, patient factors, environmental factors and the like, so that the heart ultrasonic image needs to be denoised. However, when the traditional method for denoising the heart ultrasonic image, the heart ultrasonic image may be blurred, and part of detail features may be lost.
Disclosure of Invention
The invention provides an intelligent extraction method for heart ultrasonic image features, which aims to solve the existing problems: when the traditional method for denoising the heart ultrasonic image, the heart ultrasonic image may be blurred, and part of detail features may be lost.
The intelligent heart ultrasonic image feature extraction method adopts the following technical scheme:
the method comprises the following steps:
collecting a heart ultrasonic image;
acquiring a neighborhood range of each pixel point in the heart ultrasonic image; acquiring gradient directions and gradient values of pixel points in the heart ultrasonic image according to the heart ultrasonic image; acquiring the gradient direction of the whole neighborhood range of the pixel point according to the gradient direction and the gradient value of the pixel point in the heart ultrasonic image; acquiring the possibility that the neighborhood range of the pixel point is positioned at the edge of the heart ultrasonic image according to the dispersion degree of the gray value of the pixel point in the neighborhood range of the pixel point and the difference between the gradient direction of each pixel point in the neighborhood range of the pixel point and the gradient direction of the whole neighborhood range of the pixel point;
calculating the similarity between the neighborhood ranges of the two pixels according to the distance between the two pixels in the heart ultrasonic image and the gray value of the pixel in the neighborhood range of the two pixels;
obtaining correction weights between two pixel points according to the similarity between the neighborhood ranges of the two pixel points and the possibility that the neighborhood ranges of the two pixel points are positioned at the edges in the heart ultrasonic image; obtaining a local range of each pixel point; and calculating the gray value of the corrected pixel point according to the correction weight between two pixel points in the local range of the pixel point.
Preferably, the method for acquiring the heart ultrasonic image comprises the following specific steps:
and acquiring a heart ultrasonic image through a heart ultrasonic instrument.
Preferably, the method for obtaining the neighborhood range of each pixel point in the heart ultrasonic image includes the following specific steps:
for the first in the cardiac ultrasound imageA pixel point for adding the first pixel in the heart ultrasonic image>All pixels in eight neighborhoods of the pixel are marked as +.>Neighborhood range of individual pixels.
Preferably, the method for obtaining the gradient direction and the gradient value of the pixel point in the heart ultrasonic image according to the heart ultrasonic image comprises the following specific steps:
by means ofOperator acquisition->Gradient values in the horizontal direction and gradient values in the vertical direction of all pixel points in the neighborhood range of each pixel point are respectively marked as +.>、/>Obtain->Gradient values in the horizontal direction and gradient values in the vertical direction of all pixel points in the neighborhood range of each pixel point are determined by the +.>Gradient values in the horizontal direction and gradient values in the vertical direction of all pixel points in the neighborhood range of each pixel point to obtain the +.>The specific calculation formulas of the gradient directions of all the pixel points and the gradient values of all the pixel points in the neighborhood range of each pixel point are as follows:
in the method, in the process of the invention,indicate->The +.>Gradient direction of each pixel point; />Indicate->The +.>Gradient values of the individual pixels; />Indicate->The first pixel in the neighborhood of the pixel pointGradient values in the vertical direction of the individual pixel points; />Indicate->The +.>Gradient values in the horizontal direction of the individual pixel points; />Representing an arctangent function.
Preferably, the method for obtaining the gradient direction of the whole neighborhood range of the pixel point according to the gradient direction and the gradient value of the pixel point in the heart ultrasonic image comprises the following specific steps:
for the firstThe +.>A pixel dot of +.>The +.>Gradient values of the individual pixels are taken as the modulus of the vector, and the +.>The +.>The gradient of each pixel point is taken as the gradient direction of the vector and is marked as the +.>Pixel dotIs>Vectors of the individual pixels; get->Vectors of all pixels in the neighborhood of the pixel, will be +.>The sum of vectors of all pixels in the neighborhood of the pixel as +.>Vector of neighborhood range of each pixel point, will be +.>The direction of the vector of the neighborhood range of the individual pixel points as the +.>Gradient direction of the whole neighborhood range of each pixel point.
Preferably, the method for obtaining the possibility that the neighborhood range of the pixel point is positioned at the edge in the cardiac ultrasound image according to the dispersion degree of the gray value of the pixel point in the neighborhood range of the pixel point and the difference between the gradient direction of each pixel point in the neighborhood range of the pixel point and the gradient direction of the whole neighborhood range of the pixel point comprises the following specific steps:
first obtain the firstStandard deviation of gray values of all pixel points in a neighborhood range of each pixel point and average value of gray values of all pixel points; according to->Gradient direction of the neighborhood of the individual pixel point as a whole, first +.>Gradient direction of each pixel point in neighborhood range of each pixel pointFirst->Obtaining standard deviation of gray values of all pixels in a neighborhood range of each pixel and average value of gray values of all pixels, and obtaining the +.>The specific calculation formula of the possibility that the neighborhood range of each pixel point is positioned at the edge in the heart ultrasonic image is as follows:
in the method, in the process of the invention,indicate->The possibility that the neighborhood range of each pixel point is positioned at the edge of the heart ultrasonic image; />Indicate->Standard deviation of gray values of all pixel points in a neighborhood range of each pixel point; />Indicate->The average value of gray values of all pixel points in the neighborhood range of each pixel point; />Indicate->The +.>Of individual pixelsGradient direction; />Indicate->Gradient direction of the whole neighborhood range of each pixel point; />Indicate->The number of pixels in the neighborhood range of the individual pixels; />An exponential function based on a natural constant; />Representing an absolute value operation.
Preferably, the calculating the similarity between the two pixel neighborhood ranges according to the distance between the two pixel points in the heart ultrasonic image and the gray value of the pixel point in the two pixel neighborhood ranges includes the following specific steps:
for calculation of the firstNeighborhood range and +.>The degree of similarity between the neighborhood ranges of the individual pixels; first obtain +.>Pixel dot and->The Euclidean distance between the individual pixels is then obtained>Gray of all pixels in neighborhood of each pixelDegree value and->Gray values of all pixel points in a neighborhood range of each pixel point; according to->Pixel dot and->Euclidean distance between individual pixels, th->Gray values of all pixels in the neighborhood of the pixel>Gray values of all pixels in a neighborhood range of each pixel are calculated as +.>Neighborhood range and first pixel pointThe similarity degree between the neighborhood ranges of the pixel points is calculated according to the following specific formula:
in the method, in the process of the invention,indicate->Neighborhood range and +.>The degree of similarity between the neighborhood ranges of the individual pixels; />Indicate->The average value of gray values of all pixel points in the neighborhood range of each pixel point; />Indicate->The average value of gray values of all pixel points in the neighborhood range of each pixel point; />Indicate->The +.>Gray values of the individual pixels; />Indicate->The +.>Gray values of the individual pixels; />Representing the number of pixel points in the neighborhood range; />Indicate->Pixel dot and->The Euclidean distance between the individual pixel points; />Expressed in natural constantIs an exponential function of the base.
Preferably, the method for obtaining the correction weight between two pixels according to the similarity between the neighborhood ranges of the two pixels and the possibility that the neighborhood ranges of the two pixels are located at the edge in the cardiac ultrasound image includes the following specific steps:
for calculation of the firstPixel dot and->Correction weights between the individual pixels are first obtained +.>The possibility that the neighborhood range of all the pixel points in the neighborhood range of each pixel point is positioned at the edge of the heart ultrasonic image; then, obtain->The possibility that the neighborhood range of all the pixel points in the neighborhood range of each pixel point is positioned at the edge of the heart ultrasonic image; next, get the->Neighborhood range and +.>Degree of similarity between neighborhood regions of individual pixels +.>Finally according to the firstLikelihood that the neighborhood range of all pixel points in the neighborhood range of each pixel point is positioned at the edge in the heart ultrasonic image, and +.>Possible that neighborhood range of all pixel points in neighborhood range of each pixel point is positioned at edge in heart ultrasonic imageSex +.>Get->Pixel dot and->The specific calculation formula of the correction weight among the pixel points is as follows:
in the method, in the process of the invention,indicate->Pixel dot and->Correction weights among the pixel points; />Indicate->The +.>The possibility that the neighborhood range of each pixel point is positioned at the edge of the heart ultrasonic image; />Indicate->The +.>Neighborhood range of each pixel point is positioned at edge of heart ultrasonic imageAn energy property; />Indicate->Neighborhood range and +.>The degree of similarity between the neighborhood ranges of the individual pixels; />Representing the number of pixel points in the neighborhood range; />An exponential function based on a natural constant is represented.
Preferably, the obtaining the local range of each pixel point includes the following specific methods:
presetting a local range side lengthFor->A pixel dot of +.>A pixel point is taken as the center, and a single pixel is builtLocal range of sizes, denoted by>Local extent of individual pixels.
Preferably, the calculating the gray value of the corrected pixel point according to the correction weight between two pixel points in the local range of the pixel point includes the specific method that:
for the firstA pixel dot for adding->Marking all pixel points in the local range of each pixel point as target pixel points, and acquiring gray values and the +.>Correction weights between the individual pixel points and the target pixel point; according to gray values of all target pixels +.>Correction weight between each pixel point and the target pixel point is calculated and corrected +.>The gray value of each pixel point is calculated according to the following specific formula:
in the method, in the process of the invention,indicating post-correction->Gray values of the individual pixels; />Representing the number of pixels in the local range; />Indicate->Pixel dot and->Correction weights among the target pixel points; />Indicate->Pixel dot and->Correction weights among the target pixel points; />Indicate->Gray values of the target pixel points; />Representing rounding-to-rounding functions.
The technical scheme of the invention has the beneficial effects that: according to the embodiment, through analyzing the heart ultrasonic image, the possibility that the neighborhood range of each pixel point is positioned at the edge in the heart ultrasonic image and the similarity degree between the neighborhood ranges of the two pixel points are obtained; then obtaining correction weights among the pixels according to the possibility that the neighborhood range of each pixel is positioned at the edge in the heart ultrasonic image and the similarity degree between the neighborhood ranges of the two pixels; because the larger the correction weight between the pixel points is, the two pixel points are likely to be in the same heart tissue, the detail characteristics can be reserved to a great extent by correcting the pixel points through the correction weight, and the condition that the detail characteristics are lost when the heart ultrasonic image is denoised is avoided.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flow chart of steps of an intelligent extraction method for heart ultrasonic image features.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following is a detailed description of specific implementation, structure, characteristics and effects of the intelligent extraction method for heart ultrasonic image features according to the invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of the intelligent extraction method for heart ultrasonic image features provided by the invention with reference to the accompanying drawings.
Referring to fig. 1, a flowchart of steps of a method for intelligent extraction of cardiac ultrasound image features according to an embodiment of the present invention is shown, the method includes the following steps:
step S001: and acquiring a heart ultrasonic image.
It should be noted that, the cardiac ultrasound is a safe and reliable image inspection method, and a doctor can observe the heart through the cardiac ultrasound image to make an accurate diagnosis. However, when the heart ultrasonic image is acquired, a great amount of noise exists in the heart ultrasonic image due to the factors of machine equipment, patient and environment. In order to better extract the features of the heart ultrasonic image, the embodiment needs to perform denoising processing on the heart ultrasonic image, and in order to perform denoising processing on the heart ultrasonic image, the embodiment needs to collect a heart ultrasonic image first.
Specifically, a cardiac ultrasound image is acquired by a cardiac ultrasound apparatus.
Thus, a heart ultrasonic image is obtained.
Step S002: acquiring a neighborhood range of each pixel point in the heart ultrasonic image; acquiring gradient directions and gradient values of pixel points in the heart ultrasonic image according to the heart ultrasonic image; acquiring the gradient direction of the whole neighborhood range of the pixel point according to the gradient direction and the gradient value of the pixel point in the heart ultrasonic image; and acquiring the possibility that the neighborhood range of the pixel point is positioned at the edge of the heart ultrasonic image according to the dispersion degree of the gray value of the pixel point in the neighborhood range of the pixel point and the difference between the gradient direction of each pixel point in the neighborhood range of the pixel point and the gradient direction of the whole neighborhood range of the pixel point.
It should be noted that, as a method for intelligently extracting features of a cardiac ultrasound image, the present embodiment mainly performs denoising processing on the cardiac ultrasound image, so as to achieve the purpose of better extracting features of the cardiac ultrasound image. However, the traditional method for denoising the heart ultrasonic image is non-local mean filtering, and when the heart ultrasonic image is denoised by using the non-local mean filtering, the heart ultrasonic image may be blurred, and part of detail characteristics are lost, so that diagnosis of a doctor on a patient is not facilitated; therefore, the embodiment provides an intelligent heart ultrasonic image feature extraction method.
It should be further noted that, the detail part in the cardiac ultrasound image is mainly represented by the edge in the cardiac ultrasound image, and the detail is lost due to the edge blurring in the cardiac ultrasound image caused by the traditional denoising method, so that the embodiment firstly calculates the possibility that each pixel point in the cardiac ultrasound image is located at the edge, and then denoising the cardiac ultrasound image according to the possibility that each pixel point in the cardiac ultrasound image is located at the edge pixel point, thereby avoiding the edge blurring and the detail loss caused when denoising the cardiac ultrasound image.
Specifically, for the first in the cardiac ultrasound imageFirst, the first pixel point in the heart ultrasonic image is added with the first pixel point>All pixels in eight neighborhoods of the pixel are marked as +.>A neighborhood range of the individual pixels;
then, utilizeOperator acquisition->Gradient values in the horizontal direction and gradient values in the vertical direction of all pixel points in the neighborhood range of each pixel point are respectively marked as +.>、/>Wherein->Operators are prior art and are not summarized in detail in this embodiment; get->Gradient values in the horizontal direction and gradient values in the vertical direction of all pixel points in the neighborhood range of each pixel point are determined by the +.>Gradient values in the horizontal direction and gradient values in the vertical direction of all pixel points in the neighborhood range of each pixel point to obtain the +.>The specific calculation formulas of the gradient directions of all the pixel points and the gradient values of all the pixel points in the neighborhood range of each pixel point are as follows:
in the method, in the process of the invention,indicate->The +.>Gradient direction of each pixel point; />Indicate->The +.>Gradient values of the individual pixels; />Indicate->The first pixel in the neighborhood of the pixel pointGradient values in the vertical direction of the individual pixel points; />Indicate->The +.>Gradient values in the horizontal direction of the individual pixel points; />Representing an arctangent function.
Next, for the firstThe +.>A pixel dot of +.>The +.>Gradient values of the individual pixels are taken as the modulus of the vector, and the +.>The +.>The gradient of each pixel point is taken as the gradient direction of the vector and is marked as the +.>The +.>Vectors of the individual pixels; get->Vectors of all pixels in the neighborhood of the pixel, will be +.>The sum of vectors of all pixels in the neighborhood of the pixel is taken as the firstVector of neighborhood range of each pixel point, will be +.>The direction of the vector of the neighborhood range of the individual pixel points as the +.>The neighborhood of each pixel point is integratedGradient direction.
It should be noted that, whenWhen the neighborhood range of each pixel point is positioned at the edge of the heart ultrasonic image, the +.>Gradient direction of each pixel point in neighborhood of each pixel point, and +.>The difference between gradient directions of the whole neighborhood range of each pixel point is large, and +.>The gray value difference of the neighborhood range pixel points of the pixel points is large; and when->When the neighborhood range of each pixel point is not positioned at the edge in the heart ultrasonic image, the +.>Gradient direction of each pixel point in neighborhood of each pixel point, and +.>The difference between gradient directions of the whole neighborhood range of each pixel point is small, and +.>The gray value difference of the neighborhood range pixel points of the pixel points is small; therefore, the +.>The neighborhood range of the individual pixels is located at the edge of the ultrasound image of the heart.
Specifically, first obtain the firstAll pixel points in neighborhood range of each pixel point are grayStandard deviation of the degree value and average value of gray values of all pixel points; according to->Gradient direction of the neighborhood of the individual pixel point as a whole, first +.>Gradient direction of each pixel point in neighborhood range of each pixel point, the +.>Obtaining standard deviation of gray values of all pixels in a neighborhood range of each pixel and average value of gray values of all pixels, and obtaining the +.>The specific calculation formula of the possibility that the neighborhood range of each pixel point is positioned at the edge in the heart ultrasonic image is as follows:
in the method, in the process of the invention,indicate->The possibility that the neighborhood range of each pixel point is positioned at the edge of the heart ultrasonic image; />Indicate->Standard deviation of gray values of all pixel points in a neighborhood range of each pixel point; />Indicate->Gray scale of all pixel points in neighborhood range of each pixel pointThe mean of the values; />Indicate->The +.>Gradient direction of each pixel point; />Indicate->Gradient direction of the whole neighborhood range of each pixel point; />Indicate->The number of pixels in the neighborhood range of the individual pixels; />An exponential function based on a natural constant; />Representing an absolute value operation.
It should be further noted that,indicating->Gray value difference of neighborhood range pixel points of the individual pixel points, thus +.>The greater the value of +.>The more likely the neighborhood of individual pixels is locatedEdges in the heart ultrasound image;indicate->Gradient direction of each pixel point in neighborhood of each pixel point, and +.>Differences between gradient directions of the neighborhood of individual pixels as a whole, therefore +.>The greater the value of +.>The more likely the neighborhood range of the pixel points is positioned at the edge of the heart ultrasonic image; thus->The greater the value of +.>The more likely the neighborhood of pixels is at the edge in the ultrasound image of the heart.
So far, the possibility that the neighborhood range of each pixel point is positioned at the edge of the heart ultrasonic image is obtained.
Step S003: and calculating the similarity between the neighborhood ranges of the two pixels according to the distance between the two pixels in the heart ultrasonic image and the gray value of the pixel in the neighborhood range of the two pixels.
It should be noted that, the present embodiment is used as an intelligent extraction method for features of a cardiac ultrasound image, that is, by removing noise of the cardiac ultrasound image, the quality of the cardiac ultrasound image is improved, so as to achieve the purpose of extracting features in the cardiac ultrasound image more accurately. In order to remove noise in heart tissue and ensure that detailed features in heart tissue are not lost, noise in the heart ultrasound image is uniformly distributed in the heart ultrasound image, so that when a certain heart tissue is denoised, pixel points in the heart tissue are used for denoising the heart tissue as much as possible, and the loss of the detailed features in the heart tissue is avoided.
It should be further noted that, the principle of acquiring the heart ultrasonic image is that, due to the difference of propagation speed and absorption of ultrasonic waves by different heart tissues, the ultrasonic waves will reflect when encountering tissues with different densities or different acoustic impedances; these reflected signals are captured by the receiver, forming echo signals; the received echo signals are subjected to signal processing to obtain a heart ultrasonic image, so that pixel points in the same tissue in the heart ultrasonic image are similar, and the similarity degree between the neighborhood ranges of the pixel points needs to be calculated.
Specifically, for the calculation ofNeighborhood range and +.>The degree of similarity between the neighborhood ranges of the individual pixels; first obtain +.>Pixel dot and->The Euclidean distance between the individual pixels is then obtained>Gray values of all pixels in the neighborhood of the pixel>Gray values of all pixel points in a neighborhood range of each pixel point; according to->Pixel dot and->Between pixelsEuropean distance, th->Gray values of all pixels in the neighborhood of the pixel>Gray values of all pixels in a neighborhood range of each pixel are calculated as +.>Neighborhood range and +.>The similarity degree between the neighborhood ranges of the pixel points is calculated according to the following specific formula:
in the method, in the process of the invention,indicate->Neighborhood range and +.>The degree of similarity between the neighborhood ranges of the individual pixels; />Indicate->The average value of gray values of all pixel points in the neighborhood range of each pixel point; />Indicate->Average value of gray values of all pixel points in neighborhood range of each pixel point；/>Indicate->The +.>Gray values of the individual pixels; />Indicate->The +.>Gray values of the individual pixels; />Representing the number of pixel points in the neighborhood range; />Indicate->Pixel dot and->The Euclidean distance between the individual pixel points; />An exponential function based on a natural constant is represented.
It should be noted that the number of the substrates,representing the overall similarity of the gray values of the pixels in the neighborhood of two pixels,/->The larger the value of (2), the neighborhood range of two pixelsThe more similar the enclosure, i.e. the more likely two pixels are in the same heart tissue; />The distance between two pixels is shown, and because the same heart tissue is continuously distributed, the closer the distance between the pixels is, the more likely the two are in the same heart tissue; therefore->The larger the value, the more similar the neighborhood of the two pixels, i.e. the more likely the neighborhood of the two pixels is in the same heart tissue.
So far, the similarity degree between the neighborhood ranges of the pixel points is obtained.
Step S004: obtaining correction weights between two pixel points according to the similarity between the neighborhood ranges of the two pixel points and the possibility that the neighborhood ranges of the two pixel points are positioned at the edges in the heart ultrasonic image; obtaining a local range of each pixel point; and calculating the gray value of the corrected pixel point according to the correction weight between two pixel points in the local range of the pixel point.
It should be noted that, the present embodiment is used as an intelligent extraction method for features of a cardiac ultrasound image, which aims to improve the quality of the cardiac ultrasound image by removing noise of the cardiac ultrasound image, so as to extract the features in the cardiac ultrasound image more accurately. After the step S002 and the step S003 are passed, the possibility that the neighborhood range of each pixel point is positioned at the edge in the heart ultrasonic image and the similarity degree between the neighborhood ranges of the pixel points are obtained; the correction weights among the pixels can be obtained according to the possibility that the neighborhood range of each pixel is positioned at the edge in the heart ultrasonic image and the similarity degree among the neighborhood ranges of the pixels, then the new gray value of the pixel is obtained according to the correction weights among the pixels, the correction of the pixel is completed, and the denoising of the heart ultrasonic image is completed by correcting all the pixels.
It should be further noted that, when the degree of similarity between the neighborhood regions of the pixel points is higher, the likelihood that the pixels are located in the same heart tissue is higher, and therefore, when the degree of similarity between the neighborhood regions of the pixel points is higher, the correction weight between the pixel points is greater. Because the difference between the pixel points at the edge and the pixel points not at the edge is large, when the difference of the possibility that the neighborhood range of the different pixel points is positioned at the edge in the heart ultrasonic image is larger, the correction weight among the pixel points is smaller; therefore, the correction weight between the pixels can be obtained based on the correction weight.
Specifically, for the calculation ofPixel dot and->Correction weights between the individual pixels are first obtained +.>The possibility that the neighborhood range of all the pixel points in the neighborhood range of each pixel point is positioned at the edge of the heart ultrasonic image; then, obtain->The possibility that the neighborhood range of all the pixel points in the neighborhood range of each pixel point is positioned at the edge of the heart ultrasonic image; next, get the->Neighborhood range and +.>Degree of similarity between neighborhood regions of individual pixels +.>Finally according to->Likelihood that the neighborhood range of all pixel points in the neighborhood range of each pixel point is positioned at the edge in the heart ultrasonic image/>Likelihood that the neighborhood range of all pixels in the neighborhood range of each pixel is located at the edge in the heart ultrasound image +.>Get->Pixel dot and->The specific calculation formula of the correction weight among the pixel points is as follows:
in the method, in the process of the invention,indicate->Pixel dot and->Correction weights among the pixel points; />Indicate->The +.>The possibility that the neighborhood range of each pixel point is positioned at the edge of the heart ultrasonic image; />Indicate->The +.>The possibility that the neighborhood range of each pixel point is positioned at the edge of the heart ultrasonic image; />Indicate->Neighborhood range and +.>The degree of similarity between the neighborhood ranges of the individual pixels; />Representing the number of pixel points in the neighborhood range; />An exponential function based on a natural constant is represented.
When the correction weight between two pixel points is larger, the two pixel points may be located in the same heart tissue, so that the correction of the pixel points by the correction weight can keep detailed characteristics to a great extent, and the same heart tissue is continuously distributed, so that when the pixel points are corrected, the correction of the pixel points can be completed by the pixel points in the local range of the pixel points.
Specifically, a local range side length is preset，/>The specific value of (2) can be set by combining with the actual situation, the hard requirement is not required in the embodiment, and +_ is adopted in the embodiment>To describe, for the->A pixel dot of +.>A pixel point is taken as the center, and a +.>Local range of sizes, denoted by>Local range of individual pixel points; />
Then, the first stepMarking all pixel points in the local range of each pixel point as target pixel points, and acquiring gray values and the +.>Correction weights between the individual pixel points and the target pixel point; according to gray values of all target pixels +.>Correction weight between each pixel point and the target pixel point is calculated and corrected +.>The gray value of each pixel point is calculated according to the following specific formula:
in the method, in the process of the invention,indicating post-correction->Gray values of the individual pixels; />Indicate->Pixel dot and->Correction weights among the target pixel points; />Indicate->Pixel dot and->Correction weights among the target pixel points; />Indicate->Gray values of the target pixel points; />Representing rounding-to-rounding functions.
And in the same way, all the pixel points are corrected to finish denoising the heart ultrasonic image, so that a new heart ultrasonic image is obtained, and as the gray value of each pixel point in the new heart ultrasonic image is obtained by the gray value of each pixel point in the local range of the pixel point and the correction weight, the noise is removed greatly in the new heart ultrasonic image, and meanwhile, the loss of detail features in the heart ultrasonic image is avoided.
After denoising the heart ultrasonic image is completed, a new heart ultrasonic image is obtained, at this time, the characteristics in the heart ultrasonic image can be extracted by using a traditional region growing method, so that intelligent extraction of the heart ultrasonic image characteristics is realized.
This embodiment is completed.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the invention, but any modifications, equivalent substitutions, improvements, etc. within the principles of the present invention should be included in the scope of the present invention.
Claims (9)
1. The intelligent heart ultrasonic image feature extraction method is characterized by comprising the following steps of:
collecting a heart ultrasonic image;
acquiring a neighborhood range of each pixel point in the heart ultrasonic image; acquiring gradient directions and gradient values of pixel points in the heart ultrasonic image according to the heart ultrasonic image; acquiring the gradient direction of the whole neighborhood range of the pixel point according to the gradient direction and the gradient value of the pixel point in the heart ultrasonic image; acquiring the possibility that the neighborhood range of the pixel point is positioned at the edge of the heart ultrasonic image according to the dispersion degree of the gray value of the pixel point in the neighborhood range of the pixel point and the difference between the gradient direction of each pixel point in the neighborhood range of the pixel point and the gradient direction of the whole neighborhood range of the pixel point;
calculating the similarity between the neighborhood ranges of the two pixels according to the distance between the two pixels in the heart ultrasonic image and the gray value of the pixel in the neighborhood range of the two pixels;
obtaining correction weights between two pixel points according to the similarity between the neighborhood ranges of the two pixel points and the possibility that the neighborhood ranges of the two pixel points are positioned at the edges in the heart ultrasonic image; obtaining a local range of each pixel point; according to the correction weight between two pixel points in the local range of the pixel points, calculating the gray value of the corrected pixel points;
according to the similarity between the neighborhood ranges of the two pixel points and the possibility that the neighborhood ranges of the two pixel points are positioned at the edge in the heart ultrasonic image, the correction weight between the two pixel points is obtained, and the method comprises the following specific steps:
for calculation of the firstPixel dot and->Correction weights between the individual pixels are first obtained +.>The possibility that the neighborhood range of all the pixel points in the neighborhood range of each pixel point is positioned at the edge of the heart ultrasonic image; then, obtain->The possibility that the neighborhood range of all the pixel points in the neighborhood range of each pixel point is positioned at the edge of the heart ultrasonic image; next, get the->Neighborhood range and +.>Degree of similarity between neighborhood regions of individual pixels +.>Finally according to->Likelihood that the neighborhood range of all pixel points in the neighborhood range of each pixel point is positioned at the edge in the heart ultrasonic image, and +.>Likelihood that the neighborhood range of all pixels in the neighborhood range of each pixel is located at the edge in the heart ultrasound image +.>Get->The pixel point and the first/>The specific calculation formula of the correction weight among the pixel points is as follows:
in the method, in the process of the invention,indicate->Pixel dot and->Correction weights among the pixel points; />Indicate->The +.>The possibility that the neighborhood range of each pixel point is positioned at the edge of the heart ultrasonic image; />Represent the firstThe +.>The possibility that the neighborhood range of each pixel point is positioned at the edge of the heart ultrasonic image; />Indicate->Neighborhood range and +.>The degree of similarity between the neighborhood ranges of the individual pixels; />Representing the number of pixel points in the neighborhood range; />An exponential function based on a natural constant is represented.
2. The intelligent extraction method for heart ultrasonic image features according to claim 1, wherein the method for acquiring heart ultrasonic images comprises the following specific steps:
and acquiring a heart ultrasonic image through a heart ultrasonic instrument.
3. The intelligent extraction method of the heart ultrasonic image features according to claim 1, wherein the method for obtaining the neighborhood range of each pixel point in the heart ultrasonic image comprises the following specific steps:
for the first in the cardiac ultrasound imageA pixel point for adding the first pixel in the heart ultrasonic image>All pixels in eight neighborhoods of the pixel are marked as +.>Neighborhood range of individual pixels.
4. The intelligent extraction method of the features of the heart ultrasonic image according to claim 1, wherein the steps of obtaining the gradient direction and the gradient value of the pixel point in the heart ultrasonic image according to the heart ultrasonic image comprise the following specific steps:
by means ofOperator acquisition->Gradient values in the horizontal direction and gradient values in the vertical direction of all pixel points in the neighborhood range of each pixel point are respectively marked as +.>、/>Obtain->Gradient values in the horizontal direction and gradient values in the vertical direction of all pixel points in the neighborhood range of each pixel point are determined by the +.>Gradient values in the horizontal direction and gradient values in the vertical direction of all pixel points in the neighborhood range of each pixel point to obtain the +.>The specific calculation formulas of the gradient directions of all the pixel points and the gradient values of all the pixel points in the neighborhood range of each pixel point are as follows:
in the method, in the process of the invention,indicate->The +.>Gradient direction of each pixel point; />Indicate->The +.>Gradient values of the individual pixels; />Indicate->The +.>Gradient values in the vertical direction of the individual pixel points; />Indicate->The +.>Gradient values in the horizontal direction of the individual pixel points; />Representing an arctangent function.
5. The intelligent extraction method of the features of the heart ultrasonic image according to claim 1, wherein the method for obtaining the gradient direction of the whole neighborhood range of the pixel point according to the gradient direction and the gradient value of the pixel point in the heart ultrasonic image comprises the following specific steps:
for the firstThe +.>A pixel dot of +.>The +.>Gradient values of the individual pixels are taken as the modulus of the vector, and the +.>The +.>The gradient of each pixel point is taken as the gradient direction of the vector and is marked as the +.>The +.>Vectors of the individual pixels; get->Vectors of all pixels in the neighborhood of the pixel, will be +.>The sum of vectors of all pixels in the neighborhood of the pixel as +.>Vector of neighborhood range of each pixel point, will be +.>The direction of the vector of the neighborhood range of the individual pixel points as the +.>Gradient direction of the whole neighborhood range of each pixel point.
6. The intelligent extraction method of cardiac ultrasound image features according to claim 1, wherein the obtaining the possibility that the neighborhood range of the pixel is located at the edge of the cardiac ultrasound image according to the degree of dispersion of the gray value of the pixel in the neighborhood range of the pixel, the difference between the gradient direction of each pixel in the neighborhood range of the pixel and the gradient direction of the whole neighborhood range of the pixel, comprises the following specific steps:
first obtain the firstStandard deviation of gray values of all pixel points in a neighborhood range of each pixel point and average value of gray values of all pixel points; according to->Gradient direction of the neighborhood of the individual pixel point as a whole, first +.>Gradient direction of each pixel point in neighborhood range of each pixel point, the +.>Obtaining standard deviation of gray values of all pixels in a neighborhood range of each pixel and average value of gray values of all pixels, and obtaining the +.>The specific calculation formula of the possibility that the neighborhood range of each pixel point is positioned at the edge in the heart ultrasonic image is as follows:
in the method, in the process of the invention,indicate->The possibility that the neighborhood range of each pixel point is positioned at the edge of the heart ultrasonic image; />Indicate->Standard deviation of gray values of all pixel points in a neighborhood range of each pixel point; />Indicate->The average value of gray values of all pixel points in the neighborhood range of each pixel point; />Indicate->The +.>Gradient direction of each pixel point;indicate->Gradient direction of the whole neighborhood range of each pixel point; />Indicate->The number of pixels in the neighborhood range of the individual pixels; />An exponential function based on a natural constant; />Representing an absolute value operation.
7. The intelligent extraction method of the features of the heart ultrasonic image according to claim 1, wherein the calculating the similarity between the neighborhood ranges of the two pixels according to the distance between the two pixels in the heart ultrasonic image and the gray value of the pixel in the neighborhood range of the two pixels comprises the following specific steps:
for calculation of the firstNeighborhood range and +.>The degree of similarity between the neighborhood ranges of the individual pixels; first obtain +.>Pixel dot and->The Euclidean distance between the individual pixels is then obtained>Gray values of all pixels in the neighborhood of the pixel>Gray values of all pixel points in a neighborhood range of each pixel point; according to->Pixel dot and->Euclidean distance between individual pixels, th->Gray values of all pixels in neighborhood range of each pixel and the first pixelGray values of all pixels in a neighborhood range of each pixel are calculated as +.>Neighborhood range and +.>The similarity degree between the neighborhood ranges of the pixel points is calculated according to the following specific formula:
in the method, in the process of the invention,indicate->Neighborhood range and +.>The degree of similarity between the neighborhood ranges of the individual pixels;indicate->The average value of gray values of all pixel points in the neighborhood range of each pixel point; />Indicate->The average value of gray values of all pixel points in the neighborhood range of each pixel point; />Indicate->The +.>Gray values of the individual pixels; />Indicate->The +.>Gray values of the individual pixels; />Representing the number of pixel points in the neighborhood range; />Indicate->Pixel dot and->The Euclidean distance between the individual pixel points; />An exponential function based on a natural constant is represented.
8. The intelligent extraction method of cardiac ultrasound image features according to claim 1, wherein the obtaining the local range of each pixel point comprises the following specific steps:
presetting a local range side lengthFor->A pixel dot of +.>A pixel point is taken as the center, and a +.>Local range of sizes, denoted by>Local extent of individual pixels.
9. The intelligent extraction method of cardiac ultrasound image features according to claim 1, wherein the calculating the gray value of the corrected pixel according to the correction weight between two pixels in the local range of the pixel comprises the following specific steps:
for the firstA pixel dot for adding->Marking all pixel points in the local range of each pixel point as target pixel points, and acquiring gray values and the +.>Correction weights between the individual pixel points and the target pixel point; according to gray values of all target pixels +.>Correction weight between each pixel point and the target pixel point is calculated and corrected +.>The gray value of each pixel point is calculated according to the following specific formula:
in the method, in the process of the invention,indicating post-correction->Gray values of the individual pixels; />Representing the number of pixels in the local range; />Indicate->Pixel dot and->Correction weights among the target pixel points; />Indicate->Pixel dot and->Correction weights among the target pixel points; />Indicate->Gray values of the target pixel points; />Representing rounding-to-rounding functions.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410063150.0A CN117576417B (en) | 2024-01-17 | 2024-01-17 | Intelligent heart ultrasonic image feature extraction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410063150.0A CN117576417B (en) | 2024-01-17 | 2024-01-17 | Intelligent heart ultrasonic image feature extraction method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117576417A CN117576417A (en) | 2024-02-20 |
| CN117576417B true CN117576417B (en) | 2024-03-26 |
Family
ID=89886574
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410063150.0A Active CN117576417B (en) | 2024-01-17 | 2024-01-17 | Intelligent heart ultrasonic image feature extraction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117576417B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103208117A (en) * | 2013-03-21 | 2013-07-17 | 袁景 | Intelligent multifunctional belt surface patch edge detection method |
| CN104200442A (en) * | 2014-09-19 | 2014-12-10 | 西安电子科技大学 | Improved canny edge detection based non-local means MRI (magnetic resonance image) denoising method |
| CN116205823A (en) * | 2023-05-05 | 2023-06-02 | 青岛市妇女儿童医院（青岛市妇幼保健院、青岛市残疾儿童医疗康复中心、青岛市新生儿疾病筛查中心） | Ultrasonic image denoising method based on spatial domain filtering |
| CN116758071A (en) * | 2023-08-17 | 2023-09-15 | 青岛冠宝林活性炭有限公司 | Intelligent detection method for carbon electrode dirt under visual assistance |
| CN117237591A (en) * | 2023-11-16 | 2023-12-15 | 西安道法数器信息科技有限公司 | Intelligent removal method for heart ultrasonic image artifacts |
| CN117408929A (en) * | 2023-12-12 | 2024-01-16 | 日照天一生物医疗科技有限公司 | Tumor CT image area dynamic enhancement method based on image characteristics |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101919230B (en) * | 2007-12-25 | 2013-02-13 | 梅迪奇视觉-脑科技有限公司 | Noise reduction of images |
| CA2748234A1 (en) * | 2008-12-25 | 2010-07-01 | Medic Vision - Imaging Solutions Ltd. | Denoising medical images |
-
2024
- 2024-01-17 CN CN202410063150.0A patent/CN117576417B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103208117A (en) * | 2013-03-21 | 2013-07-17 | 袁景 | Intelligent multifunctional belt surface patch edge detection method |
| CN104200442A (en) * | 2014-09-19 | 2014-12-10 | 西安电子科技大学 | Improved canny edge detection based non-local means MRI (magnetic resonance image) denoising method |
| CN116205823A (en) * | 2023-05-05 | 2023-06-02 | 青岛市妇女儿童医院（青岛市妇幼保健院、青岛市残疾儿童医疗康复中心、青岛市新生儿疾病筛查中心） | Ultrasonic image denoising method based on spatial domain filtering |
| CN116758071A (en) * | 2023-08-17 | 2023-09-15 | 青岛冠宝林活性炭有限公司 | Intelligent detection method for carbon electrode dirt under visual assistance |
| CN117237591A (en) * | 2023-11-16 | 2023-12-15 | 西安道法数器信息科技有限公司 | Intelligent removal method for heart ultrasonic image artifacts |
| CN117408929A (en) * | 2023-12-12 | 2024-01-16 | 日照天一生物医疗科技有限公司 | Tumor CT image area dynamic enhancement method based on image characteristics |
Non-Patent Citations (5)
| Title |
|---|
| Improved non-local self-similarity measures for effective speckle noise reduction in ultrasound images;Fuyuan Mei et al.;Computer Methods and Programs in Biomedicine;20200715;全文 * |
| 噪声检测自适应窗口滤波算法;吴粉侠;段群;李红;;航空计算技术;20130115(第01期);全文 * |
| 基于梯度方向的非局部均值图像去噪算法;苗璐;张权;侯红花;赵明;桂志国;;中北大学学报(自然科学版);20180215(第01期);全文 * |
| 基于超像素的高光谱图像半监督分类方法研究;谢小云;中国优秀硕士学位论文全文数据库工程科技Ⅱ辑;20210515(第05期);全文 * |
| 基于边缘提取的非局部均值图像去噪;王思涛;金聪;;电子测量技术;20180608(第11期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117576417A (en) | 2024-02-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2022052303A1 (en) | Method, apparatus and device for registering ultrasound image and ct image | |
| CN111667467B (en) | Clustering algorithm-based lower limb vascular calcification index multi-parameter accumulation calculation method | |
| WO2006044996A2 (en) | System and method for automated boundary detection of body structures | |
| CN107292835B (en) | Method and device for automatically vectorizing retinal blood vessels of fundus image | |
| CN110136088B (en) | Human embryo heart ultrasonic image denoising method | |
| CN106530236B (en) | Medical image processing method and system | |
| CN117422628B (en) | Optimized enhancement method for cardiac vascular ultrasonic examination data | |
| CN110163825B (en) | Human embryo heart ultrasonic image denoising and enhancing method | |
| WO2023151280A1 (en) | Three-dimensional fusion method and fusion system for dual-mode coronary artery blood vessel images | |
| CN117576417B (en) | Intelligent heart ultrasonic image feature extraction method | |
| CN117576123A (en) | Cardiovascular CT image data segmentation detection method | |
| CN115861132B (en) | Blood vessel image correction method, device, medium and equipment | |
| CN107169978B (en) | Ultrasonic image edge detection method and system | |
| CN116596810B (en) | Automatic enhancement method for spine endoscope image | |
| CN110276772B (en) | Automatic positioning method and system for structural elements in muscle tissue | |
| CN115222878A (en) | Scene reconstruction method applied to lung bronchoscope surgical robot | |
| CN115222651A (en) | Pulmonary nodule detection system based on improved Mask R-CNN | |
| Klein et al. | RF ultrasound distribution-based confidence maps | |
| CN117593192B (en) | Gynecological cervical image enhancement analysis method | |
| CN103892848B (en) | Calcification detection method for mammary gland X-ray image | |
| CN112561820A (en) | Self-adaptive weighted mixed total variation method suitable for ultrasonic image denoising | |
| CN117974692B (en) | Ophthalmic medical image processing method based on region growing | |
| Anwar et al. | Automatic Segmentation of Heart Cavity in Echocardiography Images: Two & Four-Chamber View Using Iterative Process Method | |
| CN116167949B (en) | Clinical auxiliary decision-making system based on medical image big data | |
| Monica et al. | Assessment of fetal biometry using ultrasound images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |