Auxiliary puncture method for subacromion gap under ultrasonic guidance
By adaptively cutting the scale and using significance weights to screen the bone area for local image enhancement, the problem of insufficient bone area representation in traditional ultrasound-assisted methods was solved, and the accuracy and effect of subacromial space puncture were improved.
Patent Information
- Application Number
- CN202510678471.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-26
- Publication Date
- 2025-09-16
AI Technical Summary
The traditional ultrasound-assisted method has poor image enhancement effect in subacromial space puncture due to the lack of clear bone area representation, which affects the puncture accuracy and effect.
The needle puncture depth is determined by adaptive cutting scale, and the significance weight is calculated by combining bone grayscale and edge features. The bone area is screened for local image enhancement to improve the clarity of the bone area.
It improves the accuracy and effect of subacromial space puncture, enhances the image clarity of the bone area, and ensures the accuracy and success rate of puncture.
Smart Images
Figure CN120643282A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of image enhancement, and in particular to an auxiliary puncture method of the subacromial space under ultrasound guidance. Background Art
[0002] Ultrasound-guided assisted puncture of the subacromial space refers to the technology of puncturing the subacromial space area of the shoulder under the guidance of ultrasound images, using the visualization images of the shoulder soft tissue provided by ultrasound. The visualization images of the shoulder soft tissue provided by ultrasound images can help doctors to fully understand the structure of the shoulder and the pathological conditions of the shoulder.
[0003] The traditional ultrasound-assisted method is to obtain an ultrasound image of the shoulder area through an ultrasound scanner, and then enhance the entire ultrasound image to make the overall features of the ultrasound image clearer; however, the process of puncturing the subacromial space needs to avoid the bone area, so in order to better assist the puncture through ultrasound images, it is necessary to highlight the features of the bone area; and the penetration ability of ultrasound in bones is weak, and the corresponding ultrasound image does not show the bones clearly enough, resulting in poor effect of directly enhancing the entire shoulder ultrasound image, and poor effect of assisting puncture based on the enhanced shoulder ultrasound image. Summary of the Invention
[0004] The present application provides an ultrasound-guided assisted puncture method for the subacromial space. The method first takes into account that the deeper the puncture needle is inserted, the higher the control accuracy requirement for the needle will be. Therefore, the cutting scale is adaptively determined according to the needle puncture depth at different sampling moments; then the corresponding initial significance is determined based on the bone grayscale features and bone edge features represented by each cut-out local image area; further, the significance weight representing the segmentation effect of the segmentation scale is determined based on the significance deviation between each local image area and the adjacent area; thereby, the bone area to be enhanced is screened out based on the final significance determined by the initial significance and the significance weight, and local image enhancement is performed according to the bone area to obtain an enhanced shoulder ultrasound image with better enhancement effect, which solves the technical problem of poor enhancement effect of the existing technology of directly enhancing the entire shoulder ultrasound image, so that the real-time assisted puncture of the subacromial space based on the enhanced shoulder ultrasound image is better.
[0005] The first aspect of the present application provides an ultrasound-guided subacromial space auxiliary puncture method, comprising:
[0006] During the subacromial space puncture, an initial shoulder ultrasound image is acquired at each sampling moment; an adaptive cutting scale at each sampling moment is determined based on the muscle tissue puncture depth of the puncture needle in the initial shoulder ultrasound image; and the initial shoulder ultrasound image is cut into at least two local image regions using a CA significance detection algorithm according to the adaptive cutting scale;
[0007] At each sampling moment, the corresponding bone grayscale eigenvalue is determined based on the overall grayscale size and grayscale distribution uniformity of the local image area; the corresponding bone edge eigenvalue is determined based on the edge smoothness of the local image area; and the initial saliency of each local image area is determined based on the bone grayscale eigenvalue and the bone edge eigenvalue;
[0008] At each sampling moment, determining a saliency weight of each local image region based on an overall deviation of initial saliency between each local image region and each adjacent local image region; and determining a final saliency of each local image region based on the saliency weight and the initial saliency;
[0009] All bone regions at each sampling moment are screened out according to the final significance; local image enhancement processing is performed on all bone regions to determine an enhanced shoulder ultrasound image at each sampling moment; and real-time auxiliary puncture of the subacromial space is performed according to the enhanced shoulder ultrasound image.
[0010] Furthermore, the process of obtaining the adaptive cutting scale includes:
[0011] The pixel point of the puncture needle in the initial shoulder ultrasound image when the needle punctures the muscle tissue in the initial state is used as the needle insertion point; the Euclidean distance between the pixel point at the puncture needle position at each sampling moment and the needle insertion point is normalized to determine the muscle tissue puncture depth of the puncture needle at each sampling moment;
[0012] The product of the negative correlation mapping value of the muscle tissue puncture depth and the preset cutting scale is used as the adaptive cutting scale at each sampling moment.
[0013] Furthermore, the process of obtaining the bone grayscale eigenvalue includes:
[0014] At each sampling moment, the mean grayscale gradient value of all pixels in each local image area is used as the corresponding grayscale change feature value;
[0015] Taking the mean grayscale value of all pixels in the initial shoulder ultrasound image as the reference grayscale value;
[0016] The difference between the mean grayscale value of all pixels in each local image area and the reference grayscale value is used as the overall grayscale feature value of each local image area;
[0017] According to the grayscale change characteristic value and the grayscale overall characteristic value, the bone grayscale characteristic value of each local image area at each sampling moment is determined; the grayscale change characteristic value is negatively correlated with the bone grayscale characteristic value; the grayscale overall characteristic value is positively correlated with the bone grayscale characteristic value.
[0018] Furthermore, the process of determining the bone grayscale eigenvalue of each local image area at each sampling moment according to the grayscale change eigenvalue and the grayscale overall eigenvalue includes:
[0019] The product of the negative correlation mapping value of the grayscale change eigenvalue and the grayscale overall eigenvalue is normalized to determine the bone grayscale eigenvalue of each local image area at each sampling moment.
[0020] Furthermore, the process of obtaining the bone edge feature value includes:
[0021] At each sampling moment, perform canny edge detection on each local image region to determine the edge line in each local image region; calculate the curvature radius of each edge pixel on the edge line; and use the average of the curvature radii corresponding to all edge pixels as the edge curvature feature value of each local image region;
[0022] Perform corner detection on each local image area and determine the number of corresponding corner points;
[0023] The product of the negative correlation mapping value of the number of corner points and the edge curvature characteristic value is normalized to determine the bone edge characteristic value of each local image area at each sampling moment.
[0024] Furthermore, the process of obtaining the initial saliency includes:
[0025] The product of the bone grayscale eigenvalue and the bone edge eigenvalue is normalized to determine the initial saliency of each local image region at each sampling moment.
[0026] Furthermore, the process of obtaining the significance weight includes:
[0027] Taking each local image region at each sampling moment as a target image region in turn; taking the local image region adjacent to the target image region as a corresponding adjacent image region;
[0028] The difference between the initial saliency of the target image region and the initial saliency of each adjacent image region is used as the target contrast saliency of each neighboring image region; the mean of the target contrast saliency of all adjacent image regions is normalized to determine the saliency weight of the target image region.
[0029] Furthermore, the process of obtaining the final saliency includes:
[0030] The final saliency of each local image region at each sampling moment is determined according to the product of the saliency weight and the initial saliency.
[0031] Furthermore, the process of obtaining the skeleton region includes:
[0032] At each sampling moment, the local image region corresponding to the final saliency greater than the preset saliency threshold is regarded as the skeleton region.
[0033] Furthermore, the process of acquiring the enhanced shoulder ultrasound image includes:
[0034] At each sampling moment, Gaussian enhancement is performed with the final saliency value of each bone region as the Gaussian kernel size to determine the enhanced image region of each bone region;
[0035] The corresponding bone regions in the initial shoulder ultrasound image are replaced with the respective enhanced image regions to determine an enhanced shoulder ultrasound image at each sampling moment.
[0036] This application has the following beneficial effects:
[0037] Considering that the purpose of this application is to perform local image enhancement on the bone area in the ultrasound image collected by the ultrasound imaging system, thereby improving the clarity of the shoulder ultrasound image during the subacromial space puncture, so that the auxiliary puncture of the subacromial space based on the ultrasound image is effective; therefore, it is necessary to analyze based on the characteristics of the bone area. First, it is necessary to consider that the deeper the puncture needle is inserted during the subacromial space puncture, the higher the control accuracy requirement for the needle will be. Therefore, in order to improve the effect of image adaptive enhancement, the cutting scale is adaptively determined according to the muscle tissue puncture depth of the puncture needle.
[0038] Further analysis of the characteristics of the bone region reveals that bone is primarily composed of hard tissue with a high degree of calcification. Ultrasound has weaker penetration in bone and stronger echoes. Therefore, bone is denser than other tissue structures and has a more uniform texture overall. Consequently, the grayscale characteristics of the bone region are higher and more uniform overall. Therefore, the bone grayscale feature value is first determined based on the overall grayscale size and grayscale distribution uniformity of the local image region.
[0039] In addition, the bones in the shoulder area have relatively smooth and soft edges due to factors such as the movement of bone joints, that is, the edges of the corresponding bone area are relatively smooth; the bones in the bone area are usually a whole, and the integrity of the area is good, that is, the edges of the bone area are usually long and smooth straight lines, and the sharp angles of the edges of the contour lines of the corresponding area are usually fewer, and the edge corners are usually arc-shaped; compared with other human tissue areas, the edge contour of the bone area is relatively smoother; therefore, the present application further determines the corresponding bone edge eigenvalue based on the edge smoothness of the local image area; and further combines the bone grayscale eigenvalue and the bone edge eigenvalue to determine the initial significance of the bone significance feature.
[0040] Then, in the initial shoulder ultrasound image, based on the characteristic that the initial significance of the bone region is greater than that of other tissue regions, that is, the initial significance of the bone region is greater than that of other adjacent tissue regions, the bone region is further screened by the significance weight obtained based on the overall deviation of the initial significance between each local image region and the adjacent local image regions, so that the final significance obtained according to the significance weight and the initial significance is more accurate.
[0041] Finally, after the required bone areas are screened out according to the final significance, image enhancement is performed only on all bone areas, making the boundaries of the bone areas in the enhanced shoulder ultrasound image clearer, and making the effect of real-time auxiliary puncture of the subacromial space based on the enhanced shoulder ultrasound image better. BRIEF DESCRIPTION OF THE DRAWINGS
[0042] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0043] Figure 1 The present invention provides a flowchart of an ultrasound-guided subacromial space auxiliary puncture method according to an embodiment of the present invention. DETAILED DESCRIPTION
[0044] In order to further illustrate the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, the following, in combination with the accompanying drawings and preferred embodiments, describes in detail the specific implementation method, structure, characteristics and effects of an ultrasound-guided subacromial space auxiliary puncture method proposed by the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment, and the specific features, structures or characteristics in one or more embodiments may be combined in any suitable form. In addition, the terms "first" and "second" are used for descriptive purposes only and cannot be understood as implying or suggesting relative importance or implicitly indicating the number of technical features indicated. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features.
[0045] 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.
[0046] The specific scheme of the ultrasound-guided subacromial space auxiliary puncture method provided by the present invention is described in detail below with reference to the accompanying drawings.
[0047] The present invention provides an ultrasound-guided subacromial space auxiliary puncture method. Figure 1 , which shows a flow chart of an ultrasound-guided subacromial space auxiliary puncture method provided by one embodiment of the present invention, the method comprising:
[0048] Step S101: During the subacromial space puncture process, an initial shoulder ultrasound image is collected at each sampling moment; the adaptive cutting scale at each sampling moment is determined by the muscle tissue puncture depth of the puncture needle in the initial shoulder ultrasound image; and the initial shoulder ultrasound image is cut into at least two local image areas according to the adaptive cutting scale using a CA significance detection algorithm.
[0049] Considering that the purpose of this application is to perform local image enhancement on the bone area in the ultrasound image collected by the ultrasound imaging system, thereby improving the clarity of the shoulder ultrasound image during the subacromial space puncture, so that the auxiliary puncture of the subacromial space based on the ultrasound image is effective; therefore, it is necessary to analyze based on the characteristics of the bone area. First, it is necessary to consider that the deeper the puncture needle is inserted during the subacromial space puncture, the higher the control accuracy requirement for the needle will be. Therefore, in order to improve the effect of image adaptive enhancement, the cutting scale is adaptively determined according to the muscle tissue puncture depth of the puncture needle.
[0050] In a specific implementation of an embodiment of the present invention, during the subacromial space puncture, the Mindray M6T ultrasound imaging system is used to collect the initial shoulder ultrasound image at each sampling moment, and simultaneously record the position of the puncture needle at each sampling moment; the sampling frequency is set to once per second.
[0051] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the adaptive cutting scale includes:
[0052] The pixel point of the needle in the initial shoulder ultrasound image when the puncture needle is inserted into the muscle tissue in the initial state is used as the needle insertion point; the Euclidean distance between the pixel point of the puncture needle position and the needle insertion point at each sampling moment is normalized to determine the muscle tissue penetration depth of the puncture needle at each sampling moment; the product between the negative correlation mapping value of the muscle tissue penetration depth and the preset cutting scale is used as the adaptive cutting scale at each sampling moment. The farther the needle is from the insertion point at each sampling moment, the deeper the penetration. Therefore, the greater the muscle tissue penetration depth, the smaller the cutting scale required based on higher attention. In a specific implementation of an embodiment of the present invention, the preset cutting scale is set to 50%; it should be noted that CA significance detection is a technical means well known to those skilled in the art, and the selection of the preset cutting scale is based on the conventional steps in CA significance detection and can be adjusted at will.
[0053] In a specific implementation of the embodiment of the present invention, the process of obtaining the adaptive cutting scale is expressed as follows: t =R ′ ×(1-Norm(L t )); where R t is the adaptive cutting scale at the t-th sampling moment; R ′ is the preset cutting scale; L t is the muscle tissue penetration depth of the puncture needle at the t-th sampling moment; Norm() is a linear normalization function. Further, based on the obtained adaptive cutting scale, the initial shoulder ultrasound image is cut using the CA significance detection algorithm to obtain at least two local image regions.
[0054] Step S102: At each sampling moment, the corresponding bone grayscale eigenvalue is determined according to the overall grayscale size and grayscale distribution uniformity of the local image area; the corresponding bone edge eigenvalue is determined according to the edge smoothness of the local image area; the initial saliency of each local image area is determined according to the bone grayscale eigenvalue and the bone edge eigenvalue.
[0055] Further analysis of the characteristics of the bone region reveals that bone is primarily composed of hard tissue with a high degree of calcification. Ultrasound has weaker penetration in bone and stronger echoes. Therefore, bone is denser than other tissue structures and has a more uniform texture overall. Consequently, the grayscale characteristics of the bone region are higher and more uniform overall. Therefore, the bone grayscale feature value is first determined based on the overall grayscale size and grayscale distribution uniformity of the local image region.
[0056] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the bone grayscale eigenvalue includes:
[0057] At each sampling moment, the mean grayscale gradient value of all pixels in each local image region is used as the corresponding grayscale change eigenvalue; the mean grayscale value of all pixels in the initial shoulder ultrasound image is used as the reference grayscale value; and the difference between the mean grayscale value of all pixels in each local image region and the reference grayscale value is used as the grayscale overall eigenvalue of each local image region. Since the grayscale of the bone region is relatively uniform overall, the corresponding grayscale change eigenvalue is usually smaller; since the grayscale of the bone region is higher overall, the grayscale value mean of the bone region is larger than that of other regions, that is, the corresponding grayscale overall eigenvalue is larger. Therefore, based on the grayscale change eigenvalue and the grayscale overall eigenvalue, the bone grayscale eigenvalue of each local image region at each sampling moment is further determined; the grayscale change eigenvalue is negatively correlated with the bone grayscale eigenvalue; the grayscale overall eigenvalue is positively correlated with the bone grayscale eigenvalue; so that the larger the grayscale overall eigenvalue and the smaller the grayscale change eigenvalue, the more the corresponding local image region conforms to the grayscale characteristics of the bone region, that is, the larger the bone grayscale eigenvalue.
[0058] Preferably, in some possible implementations of the embodiments of the present invention, the process of determining the bone grayscale eigenvalue of each local image region at each sampling moment according to the grayscale variation eigenvalue and the grayscale overall eigenvalue includes:
[0059] The product between the negative correlation mapping value of the grayscale change eigenvalue and the grayscale overall eigenvalue is normalized to determine the bone grayscale eigenvalue of each local image area at each sampling moment. In a specific implementation of an embodiment of the present invention, the process of obtaining the bone grayscale eigenvalue is expressed by the formula: Among them, G t,k is the bone grayscale eigenvalue of the kth local image region at the tth sampling moment; g′ t,k is the grayscale change characteristic value of the kth local image area at the tth sampling moment; g t,k is the mean grayscale value of all pixels in the kth local image area at the tth sampling time; is the mean grayscale value of all pixels in the initial shoulder ultrasound image at the t-th sampling moment, that is, the reference grayscale value; is the overall grayscale eigenvalue of the kth local image region at the tth sampling moment; exp() is an exponential function with a natural constant as the base; Norm() is a linear normalization function.
[0060] In addition, the bones in the shoulder area have relatively smooth and soft edges due to factors such as the movement of bone joints, that is, the edges of the corresponding bone area are relatively smooth; the bones in the bone area are usually a whole, and the integrity of the area is good, that is, the edges of the bone area are usually long and smooth straight lines, and the sharp angles of the edges of the contour lines of the corresponding area are usually fewer, and the edge corners are usually arc-shaped; compared with other human tissue areas, the edge contour of the bone area is relatively smoother; therefore, the present application further determines the corresponding bone edge eigenvalue based on the edge smoothness of the local image area; and further combines the bone grayscale eigenvalue and the bone edge eigenvalue to determine the initial significance of the bone significance feature.
[0061] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the bone edge feature value includes:
[0062] At each sampling moment, canny edge detection is performed on each local image area to determine the edge line of each local image area; the curvature radius of each edge pixel on the edge line is calculated; the average of the curvature radii corresponding to all edge pixels is used as the edge curvature eigenvalue of each local image area; corner point detection is performed on each local image area to determine the number of corresponding corner points; the product of the negative correlation mapping value of the number of corner points and the edge curvature eigenvalue is normalized to determine the skeletal edge eigenvalue of each local image area at each sampling moment.
[0063] For the edge lines of a local image area, the larger the overall curvature radius of each edge pixel point, the more the edge conforms to the characteristics of a long smooth straight line, and the more obvious the edge characteristics of the corresponding bone area are; and the fewer the number of corner points, the fewer sharp angles or textures on the contour edge, and the more obvious the edge characteristics of the corresponding bone area are; therefore, the bone edge feature value is determined based on the number of corner points and the edge curvature feature value. It should be noted that the calculation of the curvature radius and the detection of corner points are technical means well known to those skilled in the art and will not be further defined or elaborated here.
[0064] In a specific implementation of the embodiment of the present invention, the process of obtaining the bone edge feature value is expressed by the formula: Among them, B t,k is the bone edge feature value of the kth local image region at the tth sampling moment; Mt,k is the number of corner points in the kth local image region at the tth sampling moment; d t,k,a is the curvature radius of the ath edge pixel point on the edge line of the kth local image region at the tth sampling moment;
[0065] is the edge curvature eigenvalue of the kth local image region at the tth sampling moment; Norm() is the linear normalization function; exp() is the exponential function with a natural constant as the base.
[0066] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the initial saliency includes:
[0067] Since the larger the bone grayscale eigenvalue and the larger the bone edge eigenvalue, the more the corresponding local image area conforms to the characteristics of the bone area, that is, the greater the corresponding significance; therefore, the product between the bone grayscale eigenvalue and the bone edge eigenvalue is further normalized to determine the initial significance of each local image area at each sampling moment.
[0068] In a specific implementation of the embodiment of the present invention, the process of obtaining the initial saliency is expressed by the formula: E t,k =Norm(G t,k ×B t,k ); where E t,k is the initial saliency of the kth local image region at the tth sampling moment; G t,k is the bone grayscale eigenvalue of the kth local image region at the tth sampling moment; B t,k is the bone edge feature value of the kth local image region at the tth sampling moment; Norm() is the linear normalization function.
[0069] Step S103: At each sampling moment, the saliency weight of each local image region is determined based on the overall deviation of the initial saliency between each local image region and each adjacent local image region; and the final saliency of each local image region is determined based on the saliency weight and the initial saliency.
[0070] Then, in the initial shoulder ultrasound image, based on the characteristic that the initial significance of the bone region is greater than that of other tissue regions, that is, the initial significance of the bone region is greater than that of other adjacent tissue regions, the bone region is further screened by the significance weight obtained based on the overall deviation of the initial significance between each local image region and the adjacent local image regions, so that the final significance obtained according to the significance weight and the initial significance is more accurate.
[0071] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the significance weight includes:
[0072] Each local image region at each sampling moment is sequentially taken as the target image region; the local image regions adjacent to the target image region are taken as the corresponding adjacent image regions; the difference between the initial saliency of the target image region and the initial saliency of each adjacent image region is taken as the target contrast saliency of each adjacent image region; the mean of the target contrast saliency of all adjacent image regions is normalized to determine the saliency weight of the target image region. The greater the target contrast saliency corresponding to each adjacent image region, the greater the initial saliency of the target image region compared to the initial saliency of other adjacent tissue regions, and the greater the probability that the target image region belongs to the bone region, that is, the more obvious the characteristics of the bone region at the corresponding scale.
[0073] Preferably, in some possible implementations of the embodiments of the present invention, the process of obtaining the final saliency includes:
[0074] The final saliency of each local image region at each sampling moment is determined based on the product of the saliency weight and the initial saliency. In a specific implementation of an embodiment of the present invention, the process of obtaining the final saliency is expressed as follows: Among them, W t,k is the final saliency of the kth local image region at the tth sampling moment; V t,k is the number of adjacent image regions of the kth local image region at the tth sampling moment; E t,k is the initial saliency of the kth local image region at the tth sampling moment; E′ t,k,v is the initial saliency of the vth adjacent image region of the kth local image region at the tth sampling moment; (E t,k -E′ t,k,v ) is the target contrast saliency of the vth adjacent image region of the kth local image region at the tth sampling time; is the saliency weight of the kth local image region at the tth sampling moment.
[0075] Step S104: All bone regions at each sampling moment are screened out based on the final significance; local image enhancement processing is performed on all bone regions to determine the enhanced shoulder ultrasound image at each sampling moment; and auxiliary puncture of the subacromial space is performed in real time based on the enhanced shoulder ultrasound image.
[0076] Finally, after the required bone areas are screened out according to the final significance, image enhancement is performed only on all bone areas, making the boundaries of the bone areas in the enhanced shoulder ultrasound image clearer, and making the effect of real-time auxiliary puncture of the subacromial space based on the enhanced shoulder ultrasound image better.
[0077] Preferably, in some possible implementations of the embodiments of the present invention, the process of acquiring the skeleton region includes:
[0078] The greater the final saliency, the more pronounced the bone features in the local image region at the corresponding cut-off scale, and the more likely it is a bone region. Therefore, at each sampling moment, the local image region corresponding to the final saliency greater than the preset significance threshold is considered a bone region. After the bone region is determined, further local enhancement is performed to make the bone more prominent, improving the effectiveness of the assisted puncture.
[0079] Preferably, in some possible implementations of the embodiments of the present invention, the process of acquiring a strong shoulder ultrasound image includes:
[0080] At each sampling moment, Gaussian enhancement is performed with the final significance value of each bone region as the Gaussian kernel size to determine the enhanced image area of each bone region; each enhanced image area replaces the corresponding bone area in the initial shoulder ultrasound image to determine the enhanced shoulder ultrasound image at each sampling moment. It should be noted that Gaussian enhancement is a technical means well known to those skilled in the art and will not be further described here. By performing separate enhancement on the bone area detected at each sampling moment, an enhanced shoulder ultrasound image with more obvious bone features at different sampling moments can be determined in real time, so that the effect of performing real-time auxiliary puncture of the subacromial space based on the enhanced shoulder ultrasound image is better. In a specific implementation method of an embodiment of the present invention, the preset significance threshold is set to 0.4, which can be adjusted according to the specific implementation environment. Finally, real-time auxiliary puncture of the subacromial space is performed based on the enhanced shoulder ultrasound image at each sampling moment.
[0081] In summary, the ultrasound-guided assisted puncture method of the subacromial space proposed in the present application first takes into account that the deeper the puncture needle is inserted, the higher the control accuracy requirement for the needle will be. Therefore, the adaptive cutting scale is determined according to the needle puncture depth at different sampling times; then the corresponding initial significance is determined according to the bone grayscale features and bone edge features represented by each local image area cut out; further, the significance weight representing the segmentation effect of the segmentation scale is determined according to the significance deviation between each local image area and the adjacent area; thereby, the final significance determined based on the initial significance and the significance weight is screened out to select the bone area that needs to be enhanced, and local image enhancement is performed according to the bone area to obtain an enhanced shoulder ultrasound image with better enhancement effect, which solves the technical problem of poor enhancement effect of directly enhancing the entire shoulder ultrasound image in the existing technology, so that the effect of real-time assisted puncture of the subacromial space based on the enhanced shoulder ultrasound image is better.
[0082] It should be noted that the order in which the embodiments of the present invention are described above is for illustrative purposes only and does not necessarily represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or sequential order shown to achieve the desired results. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.
[0083] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on the differences from other embodiments.
Claims
1. An ultrasound-guided subacromial space assisted puncture method, characterized in that: The method comprises: During the subacromial space puncture, an initial shoulder ultrasound image is acquired at each sampling moment; an adaptive cutting scale at each sampling moment is determined based on the muscle tissue puncture depth of the puncture needle in the initial shoulder ultrasound image; and the initial shoulder ultrasound image is cut into at least two local image regions using a CA significance detection algorithm according to the adaptive cutting scale; At each sampling moment, the corresponding bone grayscale eigenvalue is determined based on the overall grayscale size and grayscale distribution uniformity of the local image area; the corresponding bone edge eigenvalue is determined based on the edge smoothness of the local image area; and the initial saliency of each local image area is determined based on the bone grayscale eigenvalue and the bone edge eigenvalue; At each sampling moment, determining a saliency weight of each local image region based on an overall deviation of initial saliency between each local image region and each adjacent local image region; and determining a final saliency of each local image region based on the saliency weight and the initial saliency; All bone regions at each sampling moment are screened out according to the final significance; local image enhancement processing is performed on all bone regions to determine an enhanced shoulder ultrasound image at each sampling moment; and real-time auxiliary puncture of the subacromial space is performed according to the enhanced shoulder ultrasound image.
2. The ultrasound-guided subacromial space assisted puncture method according to claim 1, characterized in that: The process of obtaining the adaptive cutting scale includes: The pixel point of the puncture needle in the initial shoulder ultrasound image when the needle punctures the muscle tissue in the initial state is used as the needle insertion point; the Euclidean distance between the pixel point at the puncture needle position at each sampling moment and the needle insertion point is normalized to determine the muscle tissue puncture depth of the puncture needle at each sampling moment; The product of the negative correlation mapping value of the muscle tissue puncture depth and the preset cutting scale is used as the adaptive cutting scale at each sampling moment.
3. The ultrasound-guided subacromial space assisted puncture method according to claim 1, characterized in that: The process of obtaining the bone grayscale eigenvalue includes: At each sampling moment, the mean grayscale gradient value of all pixels in each local image area is used as the corresponding grayscale change feature value; Taking the mean grayscale value of all pixels in the initial shoulder ultrasound image as the reference grayscale value; The difference between the mean grayscale value of all pixels in each local image area and the reference grayscale value is used as the overall grayscale feature value of each local image area; According to the grayscale change characteristic value and the grayscale overall characteristic value, the bone grayscale characteristic value of each local image area at each sampling moment is determined; the grayscale change characteristic value is negatively correlated with the bone grayscale characteristic value; the grayscale overall characteristic value is positively correlated with the bone grayscale characteristic value.
4. The ultrasound-guided subacromial space assisted puncture method according to claim 3, characterized in that: The process of determining the bone grayscale eigenvalue of each local image area at each sampling moment according to the grayscale change eigenvalue and the grayscale overall eigenvalue comprises: The product of the negative correlation mapping value of the grayscale change eigenvalue and the grayscale overall eigenvalue is normalized to determine the bone grayscale eigenvalue of each local image area at each sampling moment.
5. The ultrasound-guided subacromial space assisted puncture method according to claim 1, characterized in that: The process of obtaining the bone edge feature value includes: At each sampling moment, perform canny edge detection on each local image region to determine the edge line in each local image region; calculate the curvature radius of each edge pixel on the edge line; and use the average of the curvature radii corresponding to all edge pixels as the edge curvature feature value of each local image region; Perform corner detection on each local image area and determine the number of corresponding corner points; The product of the negative correlation mapping value of the number of corner points and the edge curvature characteristic value is normalized to determine the bone edge characteristic value of each local image area at each sampling moment.
6. The ultrasound-guided subacromial space assisted puncture method according to claim 1, characterized in that: The process of obtaining the initial saliency includes: The product of the bone grayscale eigenvalue and the bone edge eigenvalue is normalized to determine the initial saliency of each local image region at each sampling moment.
7. The ultrasound-guided subacromial space assisted puncture method according to claim 1, characterized in that: The process of obtaining the significance weight includes: Taking each local image region at each sampling moment as a target image region in turn; taking the local image region adjacent to the target image region as a corresponding adjacent image region; The difference between the initial saliency of the target image region and the initial saliency of each adjacent image region is used as the target contrast saliency of each neighboring image region; the mean of the target contrast saliency of all adjacent image regions is normalized to determine the saliency weight of the target image region.
8. The ultrasound-guided subacromial space assisted puncture method according to claim 1, characterized in that: The process of obtaining the final saliency includes: The final saliency of each local image region at each sampling moment is determined according to the product of the saliency weight and the initial saliency.
9. The ultrasound-guided subacromial space assisted puncture method according to claim 1, characterized in that: The process of obtaining the bone region includes: At each sampling moment, the local image region corresponding to the final saliency greater than the preset saliency threshold is regarded as the skeleton region.
10. The ultrasound-guided subacromial space assisted puncture method according to claim 1, characterized in that: The process of acquiring the enhanced shoulder ultrasound image includes: At each sampling moment, Gaussian enhancement is performed with the final saliency value of each bone region as the Gaussian kernel size to determine the enhanced image region of each bone region; The corresponding bone regions in the initial shoulder ultrasound image are replaced with the respective enhanced image regions to determine an enhanced shoulder ultrasound image at each sampling moment.