Muscle-bone ultrasound artifact interference processing method and system

Abstract

The invention relates to the field of image processing, in particular to a muscle-bone ultrasound artifact interference processing method and system, and the method comprises the steps: recording an acoustic shadow artifact region as a target region; reducing the ultrasonic frequency and obtaining a first region; performing contrast enhancement on the target area, and matching the target area with pixel points in the first area to obtain an interference area; increasing the ultrasonic frequency and obtaining a second area; marking an area outside the target area in the second area as an acoustic shadow fading area; increasing the gray values of all the pixel points outside the interference area in the second area to obtain a brightness improvement area, performing contrast enhancement to obtain a third area, obtaining a non-interference area according to the sound shadow fading area, and obtaining the noise interference degree of the pixel points in the non-interference area. And adjusting the gray values of the pixel points in the interference region by using the noise interference degree to obtain a gray correction region, and obtaining the artifact-eliminated muscle-bone ultrasound image by using the gray correction region. According to the invention, interference of artifacts is removed.

Description

Technical Field

[0001] The present invention relates to the field of image processing, and particularly to a method and system for processing artifact interference in musculoskeletal ultrasound. Background Art

[0002] Musculoskeletal ultrasound is a technique that uses high-frequency ultrasonic waves to image muscles, bones, and related soft tissues. When sound waves encounter tissues of different densities, reflection, refraction, and scattering occur. The reflected sound waves are received by the probe and converted into ultrasound images. In musculoskeletal ultrasound examinations, acoustic shadow artifacts are a common type of ultrasound image artifact, usually appearing behind highly reflective tissue interfaces, such as bones or calcified areas. Acoustic shadow artifacts may obscure or distort the true characteristics of diseased tissues, affecting the accuracy of diagnosis.

[0003] Generally, the method for removing acoustic shadow artifacts is to directly enhance the image in the acoustic shadow artifact area, such as enhancing the contrast. However, this solution is easily affected by the noise interference in the artifact area, resulting in the inability to restore or reveal accurate musculoskeletal textures. Summary of the Invention

[0004] To solve the above problems, the present invention provides a method and system for processing artifact interference in musculoskeletal ultrasound.

[0005] The method and system for processing artifact interference in musculoskeletal ultrasound of the present invention adopt the following technical solutions: An embodiment of the present invention provides a method for processing artifact interference in musculoskeletal ultrasound, the method comprising the following steps: Denote the musculoskeletal ultrasound image with an acoustic shadow artifact area as the target image, and denote the acoustic shadow artifact area in the target image as the target area; reduce the ultrasonic frequency and obtain the first image; denote the area in the first image that is the same as the target area as the first area; enhance the contrast of the target area so that the enhanced target area has the smallest difference in gray-scale distribution from the first area, match the texture of the pixel points in the enhanced target area with the texture of the pixel points in the first area, and obtain the interference area within the target area; Increase the ultrasonic frequency and obtain the second image; denote the acoustic shadow artifact area in the second image as the second area; denote the area outside the target area within the second area as the acoustic shadow fading area; increase the gray-scale values of all pixel points outside the interference area within the second area to obtain the brightness enhancement area, and the brightness enhancement area has the smallest difference in gray-scale values from all pixel points in the target area; Enhance the contrast of the brightness enhancement area to obtain the third area, and the third area has the smallest difference in gray-scale distribution from the first; The texture of the pixel points in the shadow fading area of the target image is matched with the texture of the pixel points in the shadow fading area of the second image to obtain a non-interference area in the shadow fading area. The texture of the pixel points in the non-interference area of the third area is matched with the texture of the pixel points in the non-interference area of the first image to obtain the noise interference degree of each pixel point in the non-interference area of the third area. The gray value of the pixel points in the interference area of the target area is adjusted using the noise interference degree to obtain a gray correction area. After enhancing the contrast of the gray correction area, it is fused with the first area to obtain a musculoskeletal ultrasound image with artifact elimination.

[0006] Preferably, enhancing the contrast of the target area so that the enhanced target area has the smallest difference in gray distribution from the first area includes the following specific steps: Obtain the gray histogram formed by all the gray values of the pixel points in the first area, denoted as the first histogram; According to the first histogram, perform histogram specification on the gray values of all the pixel points in the target area to obtain the enhanced target area.

[0007] Preferably, matching the texture of the pixel points in the enhanced target area with the texture of the pixel points in the first area to obtain the interference area in the target area includes the following specific steps: Match the texture of the pixel points in the enhanced target area with the texture of the pixel points in the first area to obtain the noise interference degree of each pixel point in the enhanced target area; In the enhanced target area, the area composed of all the pixel points with a noise interference degree greater than the first preset threshold is denoted as the interference area; The texture of the pixel points in the shadow fading area of the target image is matched with the texture of the pixel points in the shadow fading area of the second image to obtain a non-interference area in the shadow fading area, including the following specific steps: Match the texture of the pixel points in the shadow fading area of the target image with the texture of the pixel points in the shadow fading area of the second image to obtain the noise interference degree of each pixel point in the shadow fading area of the target image; In the shadow fading area of the target image, the area composed of all the pixel points with a noise interference degree less than the second preset threshold is denoted as the non-interference area; The noise interference degree is calculated from the distribution texture similarity between pixel points.

[0008] Preferably, the specific steps for obtaining the noise interference degree are as follows: The pixel points within the enhanced target region, the pixel points within the shadow fading region in the target image, or the pixel points within the non-interference region in the third region are denoted as the first set of pixel points, and the pixel points within the first region, the pixel points within the shadow fading region in the second image, or all the pixel points within the non-interference region on the first image are denoted as the second set of pixel points; The pixel points in the first set of pixel points are denoted as target pixel points, and the pixel points in the second set of pixel points are denoted as reference pixel points; For any one target pixel point, the reference pixel point with the greatest distribution texture similarity to this target pixel point is denoted as the matching reference pixel point of the target pixel point; For any one target pixel point A, the matching reference pixel point corresponding to the target pixel point A is denoted as B, and the number of target pixel points corresponding to the matching reference pixel point B is denoted as N. The distribution texture similarity between the i-th target pixel point among them and the matching reference pixel point B is denoted as , and is denoted as the noise interference degree of the target pixel point A, where a represents the distribution texture similarity between the target pixel point A and the matching reference target pixel point B; exp() represents the exponential function with the natural constant as the base.

[0009] Preferably, the specific steps for obtaining the distribution texture similarity between the pixel points are as follows: For any one pixel point, within the neighborhood of this pixel point, use the sobel operator to obtain the gradient directions of all the pixel points within the neighborhood. The gradient directions of all the pixel points form a gradient direction histogram, and this gradient direction histogram is denoted as the neighborhood texture distribution feature of this pixel point; the cosine similarity between the neighborhood texture distribution features of any two pixel points is denoted as the distribution texture similarity between the pixel points.

[0010] Preferably, the step of obtaining the brightness enhancement region by increasing the gray values of all the pixel points outside the interference region in the second region includes the following specific steps: Within the second region, obtain the average value A1 of the gray values of all the pixel points outside the interference region; within the target region, obtain the average value A2 of the gray values of all the pixel points outside the interference region; the difference between A2 and A1 is denoted as B, and add B to the gray values of all the pixel points outside the interference region in the second region to obtain the brightness enhancement region.

[0011] Preferably, using the noise interference degree to adjust the gray values of the pixel points within the interference region in the target region to obtain the gray correction region includes the following specific steps: Obtain the gray histogram of all the pixel points in the target region. This gray histogram represents a histogram curve, where the abscissa is the gray value and the ordinate is the frequency of the gray value appearing in the target region; Obtain any pixel point b in the interference area within the target area. The gray value of pixel point b and the frequency of occurrence of the gray value correspond to a point Qb on the histogram curve. Filter the point Qb on the histogram curve to obtain the filtered point Qb. The ordinate of the filtered point Qb is denoted as Fb. Obtain the pixel point within the target area with the smallest gray value difference from pixel point b and a gray value occurrence frequency equal to Fb, and use the gray value of this pixel point as the adjusted gray value of pixel point b. Adjust the gray values of all pixel points in the interference area within the target area. The target area after the gray value adjustment is denoted as the gray correction area. The size of the filter kernel used during the filtering is determined by the contrast enhancement error amount, and the contrast enhancement error amount is positively correlated with the noise interference degree of all pixel points in the non-interference area of the third area.

[0012] Preferably, the size of the filter kernel is , where w represents the contrast enhancement error amount, L0 represents the preset basic filter kernel size, m represents the noise interference degree of pixel point b within the target area, represents the ceiling symbol.

[0013] Preferably, the contrast enhancement error amount is equal to the average value of the noise interference degrees of all pixel points in the non-interference area of the third area.

[0014] In another embodiment of the present invention, a system for processing artifact interference in musculoskeletal ultrasound is provided. The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the above method for processing artifact interference in musculoskeletal ultrasound are implemented.

[0015] The beneficial effects of the technical solution of the present invention are: The present invention enhances the contrast of the target area, so that the enhanced target area has the smallest gray value distribution difference from the first area, matches the texture of the pixel points in the enhanced target area with the texture of the pixel points in the first area, and obtains the interference area within the target area; the obtained interference area can preliminarily describe the noise interference situation or noise distribution situation in the enhanced target area.

[0016] Increase the ultrasonic frequency and obtain a second image; the region of the acoustic shadow artifact in the second image is denoted as the second region; the region outside the target region within the second region is denoted as the acoustic shadow fading region; increase the gray values of all pixel points outside the interference region within the second region to obtain a brightness enhancement region, and the difference in gray values between the brightness enhancement region and all pixel points of the target region is minimized. During this process, the brightness of the brightness enhancement region is aligned with that of the target region. When performing contrast enhancement on the brightness enhancement region subsequently, it is possible to avoid the problem of significant differences in the contrast enhancement effect due to the different ultrasonic frequencies used for acquisition between the second region and the target region. The reason for not considering the pixel points within the interference region is that the pixel points within the interference region may be pixel points introduced by ultrasonic image noise, and the gray values of these pixel points interfere with the brightness enhancement process and the subsequent contrast enhancement process.

[0017] Further, perform contrast enhancement on the brightness enhancement region to obtain a third region, and the third region has the smallest difference in gray distribution from the first; the texture of the pixel points within the acoustic shadow fading region in the target image is matched with the texture of the pixel points within the acoustic shadow fading region in the second image to obtain the non-interference region within the acoustic shadow fading region. During this process, compared with the target region, the brightness enhancement region is obtained by collecting at a higher ultrasonic frequency and then adjusting the brightness, and it has more detailed information and more noise interference. At the same time, the third region eliminates the influence of the interference region. Therefore, the presence of the noise-affected region in the third region is the same or similar to that in the enhanced target region. Additionally, the fading enhancement region is an artifact change region brought about by the increase in frequency. The musculoskeletal texture within the fading enhancement region on the target image is not interfered by the artifact. Therefore, the obtained non-interference region is relatively less affected by noise. The fading enhancement region on the second image is interfered by the artifact. Therefore, the non-interference region can be used subsequently to analyze and describe the interference brought about by the artifact. Therefore, the third region and the non-interference region can be used to correct the contrast enhancement method and reduce the influence of noise during the contrast enhancement process.

[0018] Further, the texture of the pixel points within the non-interference region in the third region is matched with the texture of the pixel points within the non-interference region in the first image to obtain the degree of noise interference for each pixel point within the non-interference region in the third region, and use the degree of noise interference to adjust the gray values of the pixel points within the interference region in the target region to obtain a gray value correction region. This process adjusts the gray values of the pixel points within the interference region in the target region to remove the characteristics (such as slope, maximum, minimum characteristics, etc.) exhibited by the pixel points on the histogram curve. These characteristics introduce interference during contrast enhancement, and finally restore and reveal a more accurate musculoskeletal texture. Description of the Drawings

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0020] Figure 1 It is a flowchart of the steps of a method for processing artifact interference in musculoskeletal ultrasound provided by an embodiment of the present invention. Detailed implementation manners

[0021] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following, in combination with the drawings and preferred embodiments, details the specific implementation manners, structures, features and effects of a method and system for processing artifact interference in musculoskeletal ultrasound proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs.

[0023] The following specifically describes the specific solutions of a method and system for processing artifact interference in musculoskeletal ultrasound provided by the present invention with reference to the drawings.

[0024] Please refer to Figure 1 , which shows a flowchart of the steps of a method for processing artifact interference in musculoskeletal ultrasound provided by an embodiment of the present invention. The method includes the following steps: Step S001: The musculoskeletal ultrasound image with a shadow artifact area is recorded as the target image, and the shadow artifact area in the target image is recorded as the target area.

[0025] Collect a musculoskeletal ultrasound image using an ultrasonic probe. The frequency of the ultrasonic wave is P0. In this embodiment, P0 is 10 MHz. In other embodiments, P0 can be set to other values, which is not specifically limited in this embodiment. Detect the shadow artifact area in the musculoskeletal ultrasound image.

[0026] The shadow artifact area is a common artifact in musculoskeletal ultrasound detection. The reason for its formation is that behind some strong reflecting objects (such as bones or calcifications), since the sound wave is largely reflected or absorbed and cannot be clearly imaged, a shadow area is formed.

[0027] As a preferred example, the method for detecting the shadow artifact area in the musculoskeletal ultrasound image includes: Obtain the shadow artifact region in the musculoskeletal ultrasound image using a semantic segmentation network. For example, use the DeeplabV3 semantic segmentation network, which is a well-known technology, and the specific principle will not be elaborated in this embodiment.

[0028] As another example, the method for detecting the shadow artifact region in the musculoskeletal ultrasound image includes: Perform binary segmentation on the musculoskeletal ultrasound image using a threshold segmentation algorithm to obtain all the shadow regions in the musculoskeletal ultrasound image; in this embodiment, the first segmentation threshold used for threshold segmentation here is 50, and the connected domain composed of the pixel points in the musculoskeletal ultrasound image with gray values less than the first segmentation threshold is denoted as the shadow region.

[0029] Furthermore, perform binary segmentation on the musculoskeletal ultrasound image again using the threshold segmentation algorithm to obtain all the highlighted regions in the musculoskeletal ultrasound image. In this embodiment, the second segmentation threshold used for threshold segmentation here is 180, and the connected domain composed of the pixel points in the musculoskeletal ultrasound image with gray values greater than the second segmentation threshold is denoted as the highlighted region.

[0030] In this embodiment, the shadow region located below the highlighted region and closest to the highlighted region is denoted as the candidate region, and the maximum height of each candidate region is obtained (that is, the distance between the two pixel points on the contour line of the candidate region that are farthest apart in the vertical direction). The candidate region with the maximum height less than the preset height threshold is denoted as the shadow artifact region.

[0031] In this embodiment, the preset height threshold is 20% of the maximum height of the highlighted region above the candidate region.

[0032] When there is a shadow artifact region in the musculoskeletal ultrasound image, the collected musculoskeletal ultrasound image is denoted as the target image, and the shadow artifact region on the target image is denoted as the target region.

[0033] Step S002: Reduce the ultrasonic frequency and obtain the first image; the region where the target region is located in the first image is denoted as the first region; enhance the contrast of the target region, and match the texture of the pixel points in the enhanced target region with the texture of the pixel points in the first region to obtain the interference region in the target region.

[0034] When there is a shadow artifact region in the musculoskeletal ultrasound image, further reduce the ultrasonic frequency, and reduce the ultrasonic frequency to P1. In this embodiment, P1 = (1 - p1×P0), where p1 represents the frequency reduction coefficient. In this embodiment, p1 = 10% is used as an example for description, and in other embodiments, it can be set to other values, which are not specifically limited in this embodiment.

[0035] After reducing the ultrasonic frequency, collect the musculoskeletal ultrasound image at the same position and in the same probe direction again, and denote it as the first image.

[0036] On the first image, the area with the same range as the target area is denoted as the first area. Compared with the target image, since low-frequency ultrasonic waves have stronger penetration ability, but weaker ability to distinguish fine musculoskeletal textures, the resolution of the first image is lower and there are fewer musculoskeletal textures in the details. However, the (non-detail) musculoskeletal textures within the first area are relatively obvious and clear.

[0037] In this embodiment, the target area is subjected to image enhancement (such as contrast enhancement), and then matched and fused with the first area, so that the musculoskeletal textures that are not obvious or lost within the target area are revealed, and at the same time, certain musculoskeletal texture details are included, thereby reducing the artifact interference.

[0038] First, the contrast of the target area is enhanced so that the enhanced target area has the smallest difference in gray-scale distribution from the first area.

[0039] It should be noted that the target image is acquired at a relatively high ultrasonic frequency compared to the first image. The resolution in the target image is higher than that in the first image, but the noise in the target image is also relatively more, especially these noises cannot be ignored during the contrast enhancement process.

[0040] As an example, enhancing the contrast of the target area so that the enhanced target area has the smallest difference in gray-scale distribution from the first area includes the following method: Obtain the gray-scale histogram formed by all the gray-scale values of all the pixel points within the first area, denoted as the first histogram; According to the first histogram, perform histogram specification on the gray-scale values of all the pixel points within the target area to obtain the enhanced target area, and the enhanced target area has an approximate histogram with the first area.

[0041] The above histogram specification based on the first histogram is a well-known technology, and the specific principle thereof will not be elaborated in this embodiment.

[0042] Furthermore, match the enhanced target area with the first area to obtain the interference area within the target area. The interference area describes the area that is interfered by noise during the contrast enhancement process.

[0043] As an example, match the textures of the pixel points within the enhanced target area with the textures of the pixel points within the first area to obtain the interference area within the target area, including: For any pixel point in the enhanced target region or the first region, within the neighborhood of this pixel point (for example, within a 9×9 neighborhood centered on this pixel point), use the Sobel operator to obtain the gradient directions of all pixel points within the neighborhood. The gradient directions of all pixel points form a gradient direction histogram, which is used to describe the probability of each gradient direction appearing within the neighborhood. Denote this gradient direction histogram as the neighborhood texture distribution feature of this pixel point.

[0044] Arbitrarily obtain two pixel points from the enhanced target region and the first region respectively. Calculate the distribution texture similarity of these two pixel points based on the neighborhood texture distribution features of these two pixel points. The pixel point in the enhanced target region is denoted as the target pixel point, and the pixel point in the first region is denoted as the reference pixel point.

[0045] For any target pixel point, the reference pixel point with the maximum distribution texture similarity to this target pixel point is denoted as the matching reference pixel point of the target pixel point.

[0046] It should be noted that there may be a situation where multiple target pixel points all correspond to the same matching reference pixel point. For any target pixel point A, the matching reference pixel point corresponding to target pixel point A is denoted as B, and the number of target pixel points (including target pixel point A) corresponding to the matching reference pixel point B is denoted as N. The distribution texture similarity between the i-th target pixel point among them and the matching reference pixel point B is denoted as Denote as the noise interference degree after enhancement and comparison of target pixel point A (briefly denoted as the noise interference degree of target pixel point A), where a represents the distribution texture similarity between target pixel point A and the matching reference target pixel point B; exp() represents the exponential function with the natural constant as the base.

[0047] So far, the texture of the pixel points in the enhanced target region is matched with the texture of the pixel points in the first region, and the noise interference degree of each pixel point in the enhanced target region is obtained.

[0048] Among them, the smaller the above-mentioned noise interference degree, it indicates that is larger, that is, the musculoskeletal texture represented by target pixel point A can better match the musculoskeletal texture represented by the matching reference target pixel point B (or it can be said that they are the same musculoskeletal texture but at different resolutions), and at the same time, the musculoskeletal texture represented by the matching reference target pixel point B cannot be well matched with the musculoskeletal textures represented by other target pixel points (that is, target pixel points other than A). At this time, it indicates that the musculoskeletal texture represented by target pixel point A is less interfered by the noise within the target region during the contrast enhancement process.

[0049] On the contrary, the larger the noise interference degree, it indicates that The smaller it is, that is, the less able the musculoskeletal texture represented by the target pixel point A can match the musculoskeletal texture represented by the matching reference pixel point B. At the same time, the musculoskeletal texture represented by the matching reference pixel point B can better match the musculoskeletal textures represented by other target pixel points. At this time, it indicates that the musculoskeletal texture represented by the target pixel point A is more interfered by the noise in the target area during the contrast enhancement process. Even the target pixel point A is formed after the contrast enhancement due to the noise.

[0050] As an example, calculating the distribution texture similarity of two pixel points according to the neighborhood texture distribution characteristics of the two pixel points includes the method of: obtaining the cosine similarity of the neighborhood texture distribution characteristics of the two pixel points and using the cosine similarity as the distribution texture similarity.

[0051] Further, in the enhanced target area, all pixel points with a noise interference degree greater than the first preset threshold th1 are obtained, and the area formed by these pixel points is denoted as the interference area. This interference area is within the enhanced target area and also within the target area.

[0052] In this embodiment, th1 = 0.57 is used as an example for description. In other embodiments, th1 can be set to other values, and this embodiment does not make specific limitations.

[0053] Step S003: Increase the ultrasonic frequency and obtain a second image; the shadow artifact area in the second image is denoted as the second area; the area outside the target area in the second area is denoted as the shadow disappearance area; increase the gray values of all pixel points outside the interference area in the second area to obtain a brightness enhancement area; perform contrast enhancement on the brightness enhancement area to obtain a third area.

[0054] Further, increase the ultrasonic frequency to P2. In this embodiment, P2 = (1 + p2×P0), where p2 represents the frequency increase coefficient. In this embodiment, p2 = 0.2×p1 is used as an example for description. In other embodiments, it can be set to other values, and this embodiment does not make specific limitations.

[0055] After increasing the ultrasonic frequency, collect the musculoskeletal ultrasound image at the same position and the same probe direction again, denoted as the second image. Detect the shadow artifact area in the second image, denoted as the second area.

[0056] Compared with the target image, for the second image, since high-frequency ultrasonic waves have a stronger ability to resolve fine musculoskeletal textures but a weaker penetration ability, the resolution of the second image is higher, there are more musculoskeletal textures in the details, but the overall gray level in the second region is darker. The musculoskeletal textures are relatively less obvious. Additionally, the second region contains the target region. The region within the second region of the second image and outside the target region is denoted as the shadow fading region, and the shadow fading region describes the newly added artifact interference region due to the increase in ultrasonic wave frequency and the weakening of penetration ability.

[0057] Increase the gray level values of all pixel points outside the interference region within the second region to obtain the brightness enhancement region, and the difference in gray level values between the brightness enhancement region and all pixel points in the target region is the smallest. The brightness enhancement region does not include the interference region.

[0058] As an example, increasing the gray level values of all pixel points outside the interference region within the second region to obtain the brightness enhancement region includes the following method: Within the second region, obtain the average value A1 of the gray level values of all pixel points outside the interference region; within the target region, obtain the average value A2 of the gray level values of all pixel points outside the interference region; denote the difference between A2 and A1 as B, and add B to the gray level values of all pixel points outside the interference region within the second region to obtain the brightness enhancement region.

[0059] The overall brightness (i.e., the average value of the overall gray level values) of the obtained brightness enhancement region has the smallest difference from the overall brightness of the target region. The purpose is to align the brightness of the brightness enhancement region with that of the target region, and when enhancing the contrast of the brightness enhancement region subsequently, avoid the problem of a large difference in the contrast enhancement effect due to the different ultrasonic wave frequencies used in collecting the second region and the target region. The reason for not considering the pixel points within the interference region is that the pixel points within the interference region may be pixel points introduced by ultrasonic image noise, and the gray level values of these pixel points interfere with the brightness enhancement process and the subsequent contrast enhancement process.

[0060] Enhance the contrast of the brightness enhancement region so that the enhanced brightness enhancement region has the smallest difference in gray level distribution from the first region. Denote the enhanced brightness enhancement region as the third region. The specific process is the same as that in step S002.

[0061] Furthermore, match the textures of the pixel points within the shadow fading region in the target image with the textures of the pixel points within the shadow fading region in the second image to obtain the noise interference degree of each pixel point within the shadow fading region in the target image. This process is the same as that in step S002.

[0062] Further, within the acoustic shadow fading region in the target image, all pixel points with a noise interference degree less than the second preset threshold th2 are acquired, and the region formed by these pixel points is denoted as the non-interference region.

[0063] The purposes of obtaining the third region and the non-interference region in the above process are as follows: After enhancing the target region in the above process, compared with the first region, although the enhanced target region has a higher resolution, it also has more noise-affected regions. The reason for the existence of the noise-affected regions is that the target region is acquired at a relatively high ultrasonic frequency and has a strong perception ability for both noise and musculoskeletal texture. Coupled with the interference of artifacts, the noise in the target region is relatively large. When the contrast is enhanced, the noise in the target region has a greater impact during the contrast enhancement process, resulting in non-negligible noise-affected regions in the enhanced target region. Compared with the target region, the above brightness-enhanced region is obtained by adjusting the brightness after being acquired at a relatively high ultrasonic frequency, and has further detailed information and more noise interference. However, since the difference between P2 and P0 is smaller than the difference between P1 and P0, and the third region removes the influence of the interference region, the existence situation of the noise-affected regions in the third region is the same as or similar to that in the enhanced target region. Additionally, the fading-enhanced region is an artifact change region caused by the increase in frequency. The musculoskeletal texture in the fading-enhanced region on the target image is not affected by artifacts. Therefore, the obtained non-interference region is relatively less affected by noise. The fading-enhanced region on the second image is affected by artifacts. Therefore, the non-interference region can be used to analyze and describe the interference caused by artifacts subsequently. Therefore, in this embodiment, the third region and the non-interference region are used to correct the contrast enhancement method to reduce the influence of noise during the contrast enhancement process.

[0064] Step S004: Use the third region and the non-interference region to correct the contrast enhancement method to obtain a gray-scale correction region.

[0065] As an example, the method of using the third region and the non-interference region to correct the contrast enhancement method includes the following: All pixel points in the non-interference region of the third region are denoted as the first pixel distribution, all pixel points in the non-interference region of the first image are denoted as the second pixel distribution, the texture of the pixel points in the first pixel distribution is matched with the texture of the pixel points in the second pixel distribution to obtain the noise interference degree of each pixel point in the first pixel distribution. This process is the same as that in step S002.

[0066] When the noise interference level of each pixel point in the first pixel distribution is relatively high, it indicates that when performing contrast enhancement on the target area or the second area, the interference of noise during the contrast enhancement process is relatively large; further speaking, when performing contrast enhancement on the target area, the interference area also participates in this process, resulting in noise interference. The interference area participates in (or participates too much in) the contrast enhancement process of the target area, resulting in the possibility that the interference area actually obtained in step S002 may be unreliable. Furthermore, when increasing the brightness of the second area (i.e., obtaining the brightness increase area), there is still noise interference (that is, although the interference area is excluded when obtaining the brightness increase area, the interference area may be unreliable because the interference area participates in the contrast enhancement process of the target area), ultimately resulting in a relatively high noise interference level for each pixel point in the first pixel distribution.

[0067] Adjust the gray values of the pixel points in the interference area within the target area by using the noise interference level of each pixel point in the first pixel distribution to obtain a gray value correction area; note that the gray values of the pixel points outside the interference area within the target area are not adjusted.

[0068] As an example, adjusting the gray values of the pixel points in the interference area within the target area by using the noise interference level of all pixel points in the first pixel distribution includes the following method: Obtain the gray value histogram of all pixel points in the target area. This gray value histogram represents a histogram curve, where the abscissa is the gray value and the ordinate is the frequency of the occurrence of this gray value in the target area.

[0069] Obtain any pixel point b within the interference area of the target image, obtain the gray value of pixel point b and the frequency of the occurrence of this gray value. This gray value and this frequency correspond to a point Qb on the histogram curve. Filter the point Qb on the histogram curve to obtain the filtered point Qb. The ordinate of the filtered point Qb is denoted as Fb. Obtain the pixel point within the target area with the smallest gray value difference from pixel point b and the frequency of the gray value occurrence equal to Fb, and use the gray value of this pixel point as the gray value of pixel point b. The gray value difference mentioned here refers to the absolute value of the difference between gray values.

[0070] This embodiment uses the mean filtering method for filtering, and the size of the filtering kernel is , where w represents the contrast enhancement error amount. In this embodiment, w is set equal to the mean value of the noise interference levels of all pixel points in the first pixel distribution. The larger the mean value of this noise interference level, the more difficult it is to restore the muscle and bone texture included in the non-interference area of the third area during contrast enhancement, indicating that the error generated during the contrast enhancement process due to noise interference is relatively large.

[0071] L0 represents the basic filtering kernel size. In this embodiment, L0 = 3 is used as an example for description. Denotes the ceiling symbol. Additionally, it should be noted that when the filter kernel size is even, the filter kernel size will be automatically incremented by one to ensure that the filter kernel size is always odd.

[0072] The purpose of filtering and resetting the gray value of pixel point b in the above process is to remove the features exhibited by pixel point b on the histogram curve (such as slope, maximum, minimum features, etc.). The reason is that in this embodiment, contrast enhancement is mainly based on the distribution characteristics of the gray histogram. After removing the features exhibited by pixel point b on the histogram curve, interference introduced by pixel point b during the contrast enhancement process can be avoided. Among them, the greater the average value of the noise interference levels of all pixel points within the first pixel distribution, the more necessary it is to remove the features exhibited by pixel point b on the histogram curve.

[0073] Similarly, update the gray values of all pixel points within the interference region.

[0074] As another example, adjusting the gray values of pixel points within the interference region in the target region using the noise interference level of each pixel point within the first pixel distribution includes the following method: When the average value of the noise interference levels of all pixel points within the first pixel distribution is greater than the preset interference threshold (for example, greater than 0.35), perform mean filtering on the gray values of all pixel points within all interference regions. For example, use a 7×7 filter kernel for filtering, and the filtered gray value is the updated gray value of all pixel points within the interference region.

[0075] This example has a relatively fast calculation speed but a lower accuracy. The reason is that the interference effects of different pixel points within the interference region during the contrast enhancement process are different. Directly performing mean filtering on the gray values of all pixel points within the interference region results in incomplete removal of the influence brought by some pixel points, still introducing interference during the contrast enhancement process, or some pixel points contain certain non-noise muscle and bone information, and these information cannot be retained with more muscle and bone detail textures after being removed.

[0076] In some other embodiments, adjusting the gray values of pixel points within the interference region in the target region using the noise interference level of each pixel point within the first pixel distribution includes the following method: The filter kernel size is , where m represents the noise interference level of pixel point b obtained in step S002. The larger m is, the more likely pixel point b contains more noise. At this time, the more noise pixel point b contains, the more serious the interference of the noise during the contrast enhancement process (that is, the greater the error during the contrast enhancement process), and the more necessary it is to remove the interference of pixel point b.

[0077] For the target region after the above gray value update and adjustment, it is denoted as the gray correction region.

[0078] Step S005: After enhancing the contrast of the grayscale correction region, fuse it with the first region to obtain a musculoskeletal ultrasound image with artifact elimination.

[0079] Enhance the contrast of the grayscale correction region so that the enhanced grayscale correction region has the smallest difference in grayscale distribution from the first region (similarly to step S002).

[0080] Further, fuse the enhanced grayscale correction region with the first region to obtain an artifact interference elimination region.

[0081] As an example, fusing the enhanced grayscale correction region with the first region includes the method of: using the Gaussian pyramid fusion algorithm to fuse the enhanced grayscale correction region with the first region to obtain a fused region. The Gaussian pyramid fusion algorithm is a well-known technology, and the principle thereof will not be elaborated in this embodiment.

[0082] The fused region includes both the high-resolution but artifact-interfered musculoskeletal texture details in the target image collected at the high-frequency ultrasonic wave frequency and the obvious and artifact-uninterfered lower-resolution musculoskeletal texture in the first image collected at the low-frequency ultrasonic wave frequency. On the target image, replace the target region with the fused region to obtain a musculoskeletal ultrasound image with artifact elimination, and display this musculoskeletal ultrasound image on the display.

[0083] In another embodiment, when there are no texture details in the target region obtained in step S001, for example, when the mean value of the grayscale values of all pixel points in the target region is less than 50, it indicates that there is extremely strong artifact interference when performing musculoskeletal detection at the ultrasonic wave frequency of P0.

[0084] At this time, one method is: directly replace the target region with the first region to obtain a musculoskeletal ultrasound image with artifact elimination, and display this musculoskeletal ultrasound image on the display. This method has a fast calculation speed, but does not retain or restore the musculoskeletal texture details under artifact interference.

[0085] Another method is: reduce the ultrasonic wave frequency by 1% (in other embodiments, it can be reduced to other values), record the reduced ultrasonic wave frequency as P0 again, record the obtained musculoskeletal ultrasound image as the target image at this time, record the acoustic shadow artifact region in the target image as the target region, and then obtain a musculoskeletal ultrasound image with artifact elimination according to the methods of the above steps S002 to S005, and display this musculoskeletal ultrasound image on the display. This method retains or restores the musculoskeletal texture details under artifact interference as much as possible.

[0086] In another embodiment, when there are no texture details in the second region obtained in step S003, for example, when the average value of the gray values of all pixel points in the second region is less than 30, it indicates that there is extremely strong artifact interference when performing musculoskeletal detection at the ultrasonic frequency of P2. At this time, increase P2 by 1% (in other embodiments, it can be increased to other values), record the increased ultrasonic frequency as P2 again, and use the methods of steps S003 to S005 to obtain a musculoskeletal ultrasound image with artifacts eliminated, and display the musculoskeletal ultrasound image on the display.

[0087] Similarly, when there are no obvious texture details in the first region obtained in step S002, for example, when the average value of the gray values of all pixel points in the first region is less than 80, reduce P1 by 1%, record the reduced ultrasonic frequency as P1 again, and then obtain a musculoskeletal ultrasound image with artifacts eliminated according to the methods of the above steps S002 to step S005, and display the musculoskeletal ultrasound image on the display.

[0088] In another embodiment of the present invention, a system for processing artifact interference in musculoskeletal ultrasound is provided. The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the above steps S001 to S005 are implemented.

[0089] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for processing artifact interference in musculoskeletal ultrasound, characterized in that, The method includes the following steps: The musculoskeletal ultrasound image with a shadow artifact area is denoted as the target image, and the shadow artifact area in the target image is denoted as the target area; reduce the ultrasonic frequency and obtain the first image; the area in the first image that is the same as the target area is denoted as the first area; enhance the contrast of the target area so that the enhanced target area has the smallest difference in gray-level distribution from the first area, match the texture of the pixel points in the enhanced target area with the texture of the pixel points in the first area, and obtain the interference area in the target area; Increase the ultrasonic frequency and obtain the second image; the shadow artifact area in the second image is denoted as the second area; the area outside the target area in the second area is denoted as the shadow disappearance area; increase the gray-level values of all pixel points outside the interference area in the second area to obtain the brightness enhancement area, and the difference in gray-level values between the brightness enhancement area and all pixel points in the target area is the smallest; Enhance the contrast of the brightness enhancement area to obtain the third area, and the third area has the smallest difference in gray-level distribution from the first; Match the texture of the pixel points in the shadow disappearance area in the target image with the texture of the pixel points in the shadow disappearance area in the second image to obtain the non-interference area in the shadow disappearance area, match the texture of the pixel points in the non-interference area in the third area with the texture of the pixel points in the non-interference area in the first image, obtain the noise interference degree of each pixel point in the non-interference area in the third area, adjust the gray-level values of the pixel points in the interference area in the target area using the noise interference degree to obtain the gray-level correction area, enhance the contrast of the gray-level correction area and then fuse it with the first area to obtain the musculoskeletal ultrasound image with artifacts eliminated.

2. The method for processing artifact interference of musculoskeletal ultrasound according to claim 1, wherein The step of enhancing the contrast of the target area so that the enhanced target area has the smallest difference in gray-level distribution from the first area includes the following specific steps: Obtain the gray-level histogram composed of all the gray-level values of all pixel points in the first area, denoted as the first histogram; According to the first histogram, perform histogram specification on the gray-level values of all pixel points in the target area to obtain the enhanced target area.

3. The method for processing the artifact interference of musculoskeletal ultrasound according to claim 1, wherein The step of matching the texture of the pixel points in the enhanced target area with the texture of the pixel points in the first area to obtain the interference area in the target area includes the following specific steps: Match the texture of the pixel points in the enhanced target area with the texture of the pixel points in the first area to obtain the noise interference degree of each pixel point in the enhanced target area; In the enhanced target area, the area composed of all pixel points with a noise interference degree greater than the first preset threshold is denoted as the interference area; The step of matching the texture of the pixel points in the shadow disappearance area in the target image with the texture of the pixel points in the shadow disappearance area in the second image to obtain the non-interference area in the shadow disappearance area includes the following specific steps: Match the texture of the pixel points in the shadow disappearance area in the target image with the texture of the pixel points in the shadow disappearance area in the second image to obtain the noise interference degree of each pixel point in the shadow disappearance area in the target image; In the target image, the area composed of all pixel points with noise interference degree less than the second preset threshold within the acoustic shadow fading area is denoted as the non-interference area; The noise interference degree is calculated from the distribution texture similarity between pixel points.

4. The method for processing artifact interference of musculoskeletal ultrasound according to claim 1 or 3, characterized in that, The specific steps to obtain the noise interference degree are as follows: Denote the pixel points in the enhanced target area, the pixel points in the acoustic shadow fading area of the target image, or the pixel points in the non-interference area of the third area as the first pixel point set, and denote the pixel points in the first area, the pixel points in the acoustic shadow fading area of the second image, or all pixel points in the non-interference area of the first image as the second pixel point set; The pixel points in the first pixel point set are denoted as target pixel points, and the pixel points in the second pixel point set are denoted as reference pixel points; For any target pixel point, the reference pixel point with the maximum distribution texture similarity to this target pixel point is denoted as the matching reference pixel point of the target pixel point; For any target pixel point A, the corresponding matching reference pixel point of target pixel point A is denoted as B, the number of target pixel points corresponding to the matching reference pixel point B is denoted as N, and the distribution texture similarity between the i-th target pixel point among them and the matching reference pixel point B is denoted as , and is denoted as the noise interference degree of target pixel point A, where a represents the distribution texture similarity between target pixel point A and matching reference target pixel point B; exp() represents the exponential function with the natural constant as the base.

5. The method for processing artifact interference of musculoskeletal ultrasound according to claim 3, characterized in that The specific steps to obtain the distribution texture similarity between pixel points are as follows: For any pixel point, within the neighborhood of this pixel point, use the sobel operator to obtain the gradient directions of all pixel points in the neighborhood. The gradient directions of all pixel points form a gradient direction histogram, and denote this gradient direction histogram as the neighborhood texture distribution feature of this pixel point; The cosine similarity of the neighborhood texture distribution features of any two pixel points is denoted as the distribution texture similarity between pixel points.

6. The method for processing artifact interference of musculoskeletal ultrasound according to claim 1, wherein, Increasing the gray values of all pixel points outside the interference area in the second area to obtain the brightness enhancement area includes the following specific steps: In the second area, obtain the mean value A1 of the gray values of all pixel points outside the interference area; in the target area, obtain the mean value A2 of the gray values of all pixel points outside the interference area; denote the difference between A2 and A1 as B, and add B to the gray values of all pixel points outside the interference area in the second area to obtain the brightness enhancement area.

7. The artifact interference processing method of musculoskeletal ultrasound according to claim 1, characterized in that, Adjusting the gray values of the pixel points in the interference area of the target area using the noise interference degree to obtain the gray value correction area includes the following specific steps: Obtain the gray value histogram of all pixel points in the target area. This gray value histogram represents a histogram curve, where the abscissa is the gray value and the ordinate is the frequency of the gray value appearing in the target area; Obtain any pixel point b in the interference area of the target area. The gray value of pixel point b and the frequency of the gray value appearing correspond to a point Qb on the histogram curve. Filter the point Qb on the histogram curve to obtain the filtered point Qb. Denote the ordinate of the filtered point Qb as Fb; obtain the pixel point in the target area with the smallest gray value difference from pixel point b and the gray value appearance frequency equal to Fb, and use the gray value of this pixel point as the adjusted gray value of pixel point b; Adjust the gray values of all pixel points in the interference area of the target area. The target area with the adjusted gray values is denoted as the gray value correction area; The size of the filter kernel used during filtering is determined by the contrast enhancement error amount, and the contrast enhancement error amount is positively correlated with the noise interference degree of all pixel points in the non-interference area of the third area.

8. The method for processing artifact interference of musculoskeletal ultrasound according to claim 7, wherein The size of the filtering kernel is , where w represents the contrast enhancement error amount, L0 represents the preset basic filtering kernel size, m represents the noise interference degree of the pixel b in the target area, represents the ceiling symbol.

9. The artifact interference processing method for musculoskeletal ultrasound according to claim 7, wherein The contrast enhancement error amount is equal to the average of the noise interference degrees of all pixel points in the non-interference area of the third area.

10. An artifact interference processing system for musculoskeletal ultrasound, the system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method for processing artifact interference of musculoskeletal ultrasound according to any one of claims 1 to 9.

Images