Auxiliary positioning system for implanting hollow screw

Abstract

The invention relates to the technical field of medical treatment, and particularly discloses an auxiliary positioning system for implanting a hollow screw, which is used for solving the problem that the accuracy of implanting the hollow screw is influenced by slight displacement or deformation of a fracture part in a transition period between positioning of a patient and implementation of an actual operation. According to the method, the first bone image picture used for preoperative positioning and the second bone image picture newly obtained before the operation are obtained, the bone edge feature points in the image pictures are obtained by performing multiple analysis processing on the bone image pictures, and whether bone changes exist or not is analyzed and judged according to the bone edge feature points; according to the method, the abnormal skeleton image is marked, and the edge feature points at different positions in the second skeleton image and the first skeleton image are marked in the second skeleton image, so that the skeleton change area can be accurately marked, intraoperative deviation is reduced, and the accuracy of hollow screw implantation is improved.

Description

Technical Field

[0001] The present invention relates to the field of medical technology, and more specifically, to an auxiliary positioning system for the implantation of hollow screws. Background Art

[0002] Hollow screws play an important role in the treatment of fractures, especially in the internal fixation treatment of femoral neck fractures. With the aggravation of the aging society, the incidence of femoral neck fractures has been increasing year by year. Especially in the elderly patient group with osteoporosis, the treatment of femoral neck fractures has become a clinical problem that urgently needs to be solved. The existing literature (Xing Baorui. Clinical and finite element study on the treatment of senile fragile pelvic fractures by intelligent robot combined with bone cement reinforcement technology [D]. Hebei Medical University, 2024. DOI: 10.27111 / d.cnki.ghyku.2024.000125.) studies the application of hollow screw implantation technology in elderly patients with fragile pelvic fractures and gives examples such as Figure 2 shown in a typical case of a 67-year-old male patient with FFPIVc type before, during and after surgery, and Figure 3 shown in the surgical process of closed reduction and blocking screw technology for a 70-year-old female patient with FFPIIIc type to fix the channel screw for the treatment of FFPIIIc type pelvic fractures. During the implantation process of hollow screws, the patient's body position must be accurately set. Inaccurate or unstable body position will lead to positioning errors in target areas such as the femoral neck and femoral head, which will in turn affect the implantation angle and direction of the screws. Especially in elderly patients or patients with limited mobility, it is difficult to adjust the body position, which may lead to a decrease in the accuracy of screw implantation, affect the surgical effect, and increase the risk of postoperative complications. However, during the transition period between the patient's positioning and the actual surgical implementation, especially in elderly patients, due to various factors, the fracture site of the patient undergoes slight displacement or deformation, and there may be slight differences between the latest image before surgery and the original positioning image, which will in turn affect the accuracy of hollow screw implantation. To solve the above problems, a technical solution is provided now. Summary of the Invention

[0003] In order to overcome the above-mentioned defects of the prior art, the present invention provides an auxiliary positioning system for the implantation of hollow screws, which obtains the bone edge feature points in the image picture through multiple analysis and processing to solve the problem that the slight displacement or deformation of the fracture site during the transition period between the patient's positioning and the actual surgical implementation affects the accuracy of hollow screw implantation, and improves the accuracy of hollow screw implantation to solve the problems raised in the above background art.

[0004] To achieve the above object, the present invention provides the following technical solutions:

[0005] An auxiliary positioning system for hollow screw implantation, comprising an imaging parameter acquisition module, a bone image registration module, and an abnormality display center; the imaging parameter acquisition module is used to obtain a first bone image picture for preoperative positioning and a second bone image picture newly obtained before surgery; the bone image registration module is used to obtain the bone edge feature points in the image picture through multiple analysis and processing of the bone image picture, and analyze and judge whether there is bone change according to the bone edge feature points; the bone image registration module includes a bone image registration unit and a bone difference detection and extraction unit; the bone image registration unit is used to extract the bone gray gradient edges in the first bone image picture and the second bone image picture through edge detection, and respectively mark the first edge feature points {p i} of the first bone image picture and the second edge feature points {q j} of the second bone image picture; the bone difference detection and extraction unit is used to calculate the bone image difference based on the first edge feature points and the second edge feature points, and judge whether there is bone change by classifying and analyzing the edge feature points at the same position and different positions: normalize the absolute value of the difference of all edge feature points at the same position, and divide it by the number of edge feature points at the same position, and at the same time consider the sum of the differences of the unique edge feature points in the first image and the sum of the differences of the unique edge feature points in the second image to form the final bone change judgment value;

[0006] Construct a bone change judgment model through the bone image difference, and the formula of the bone change judgment model is:

[0007]

[0008] In the formula: β G is the bone change value, H is the number of edge feature points at the same position, ΔI(x,y) h is the bone image difference of the hth edge feature point at the same position, A is the number of edge feature points in the first bone image picture but not in the second bone image picture, ΔI(x a ,y a ) is the bone image difference of the edge feature point at the position (x a ,y a ), B is the number of edge feature points in the second bone image picture but not in the first bone image picture, ΔI(x b ,y b ) is the bone image difference of the edge feature point at the position (x b ,x b ).

[0009] As a further solution of the present invention, the bone image registration unit is used to extract the bone gray-scale gradient edges in the first bone image and the second bone image through edge detection, and respectively mark the first edge feature points {p i} of the first bone image and the second edge feature points {q j} of the second bone image. The specific steps are as follows:

[0010] Use the edge detection operator to calculate the gradient amplitude G(x, y) of each pixel point in the first bone image and the second bone image, and regard all pixel points with non-zero gradient amplitude values as preliminary edge points;

[0011] The set first threshold range for the edge pixel points in the first bone image is The second threshold range for the edge pixel points in the second bone image is

[0012] If the gradient amplitude of the preliminary bone edge point in the first bone image is greater than or equal to Then mark this edge point as "p high "; if the gradient amplitude of the preliminary bone edge point in the first bone image is within range, then mark this bone edge point as "p low "; if the gradient amplitude of the preliminary bone edge point in the first bone image is less than Then discard this bone edge point;

[0013] If the gradient amplitude of the preliminary bone edge point in the second bone image is greater than or equal to Then mark this bone edge point as "q high "; if the gradient amplitude of the preliminary bone edge point in the second bone image is within range, then mark this bone edge point as "q low "; if the gradient amplitude of the preliminary bone edge point in the second bone image is less than Then discard this bone edge point.

[0014] The bone image registration module further includes an image processing unit; the image processing unit is used to perform denoising and normalization processing on the first bone image and the second bone image respectively; the formula for performing normalization processing on the first bone image is:

[0015]

[0016] Where: I p (X1, Y1) is the pixel value at the position (X1, Y1) in the first bone image after normalization processing, I p,w(X1, Y1) is the pixel value at position (X1, Y1) in the first skeletal image before normalization, and min(I p,w (X, Y)) is the minimum pixel value in the first skeletal image before normalization, and max(I p,w (X, Y)) is the maximum pixel value in the first skeletal image before normalization.

[0017] The formula for normalizing the second skeletal image is:

[0018]

[0019] In the formula: I q (X2, Y2) is the pixel value at position (X2, Y2) in the second skeletal image after normalization, and I q,w (X2, Y2) is the pixel value at position (X2, Y2) in the second skeletal image before normalization, and min(I q,w (X, Y)) is the minimum pixel value in the second skeletal image before normalization, and max(I q,w (X, Y)) is the maximum pixel value in the second skeletal image before normalization.

[0020] Through three steps of image denoising, precise matching, and difference analysis, doctors can judge the fracture displacement situation in real time, optimize the implantation path of hollow screws, improve the surgical accuracy, and reduce the error risk. The application of this technology will enhance the intelligence level of orthopedic surgeries, reduce surgical complications, and improve the postoperative recovery effect of patients.

[0021] As a further solution of the present invention, for the bone edge point "p low ", if there is at least one bone edge point "p high " in its 8-neighborhood, then retain this bone edge point; if there is no bone edge point "p high " in its 8-neighborhood, then discard this bone edge point;

[0022] For the bone edge point "q low ", if there is at least one bone edge point "q high " in its 8-neighborhood, then retain this bone edge point; if there is no bone edge point "q high " in its 8-neighborhood, then discard this bone edge point;

[0023] Count the remaining "p high ", "p low ", "q high " and "q low " after the above operations, and for "p high ", "p low”The first edge feature point {p classified in the first bone image i}, and the second edge feature point {q in the second bone image of "q high ", "q low "}. j}

[0024] As a further solution of the present invention, the set first threshold range for the edge pixel points in the first bone image is Among them, the lowest threshold in the first threshold range and the highest threshold are determined as follows:

[0025] The lowest threshold is calculated based on the relative difference between the gradient magnitudes of all pixel points in the first bone image and the minimum gradient magnitude. Divide the difference between the gradient magnitude of each pixel point and the minimum gradient magnitude in the image by the minimum gradient magnitude, take the absolute value, sum over all pixel points, and finally take the average to obtain the lowest threshold; the highest threshold is calculated by dividing the difference between the gradient magnitude of each pixel point in the first bone image and the maximum gradient magnitude by the maximum gradient magnitude, taking the absolute value, accumulating, and then taking the average.

[0026] The lowest threshold in the first threshold range and the highest threshold are determined by the following formula:

[0027]

[0028] In the formula: is the lowest threshold in the first threshold range, is the highest threshold in the first threshold range, P is the number of pixel points in the first bone image, G p (x, y) is the gradient magnitude of the p-th pixel point in the first bone image, is the minimum gradient magnitude in the first bone image, is the maximum gradient magnitude in the first bone image.

[0029] As a further solution of the present invention, the second threshold range for the edge pixel points in the second bone image is Among them, the lowest threshold in the second threshold range and the highest threshold are determined as follows:

[0030] The lowest threshold Calculated based on the relative difference between the gradient magnitude of all pixel points in the second bone image and the minimum gradient magnitude. Divide the difference between the gradient magnitude of each pixel point and the minimum gradient magnitude in the image by the minimum gradient magnitude, take the absolute value, sum over all pixel points, and finally take the average to obtain the lowest threshold; the highest threshold The calculation method of is to divide the difference between the gradient magnitude of each pixel point and the maximum gradient magnitude in the second bone image by the maximum gradient magnitude, take the absolute value, accumulate, and then take the average to obtain it.

[0031] The lowest threshold in the second threshold range And the highest threshold The determination formula is:[[]]

[0032]

[0033] Wherein: Is the lowest threshold in the second threshold range, Is the highest threshold in the second threshold range, Q is the number of pixel points in the second bone image, G q (x,y) is the gradient magnitude of the q-th pixel point in the second bone image, Is the minimum gradient magnitude in the second bone image, Is the maximum gradient magnitude in the second bone image.

[0034] As a further solution of the present invention, the bone difference detection and extraction unit is used to calculate the bone image difference based on the first edge feature points and the second edge feature points, and determine whether there is a bone change by classifying and analyzing the edge feature points at the same position and different positions. The calculation of the bone image difference for the edge feature points at the same position is:

[0035] ΔI(x,y) = I q (x,y) - I p (x,y);

[0036] Wherein: ΔI(x,y) is the bone image difference at the coordinate position (x,y), I p (x,y) is the pixel value of the bone image at the position (x,y) in the first bone image, I q (x,y) is the pixel value of the bone image at the position (x,y) in the second bone image;

[0037] When calculating the bone image difference for the edge feature points at different positions, if the edge feature point at the position (x a ,y a ) is in the first bone image but not in the second bone image, then there is a bone image difference ΔI(x a ,ya ) = I p (x a , y a ) - 0; If the edge feature point at position (x b , y b ) is in the second skeletal image but not in the first skeletal image, then the skeletal image difference ΔI(x b , y b ) = 0 - I q (x b , y b );

[0038] Based on the above operation steps, summarize the skeletal image differences, and construct a skeletal change judgment model through the skeletal image differences to judge whether the skeleton has changed.

[0039] The technical effects and advantages of an auxiliary positioning system for cannulated screw implantation according to the present invention: By obtaining a first skeletal image for preoperative positioning and a second skeletal image obtained latest before surgery, analyzing and processing the skeletal image multiple times to obtain the skeletal edge feature points in the image, and analyzing and judging whether there is skeletal change based on the skeletal edge feature points, marking the abnormal skeletal images, and marking the edge feature points at different positions in the second skeletal image from those in the first skeletal image, it can accurately mark the skeletal change area, reduce intraoperative deviation, improve the accuracy of cannulated screw implantation, reduce the risk of secondary repair, screw loosening or displacement after surgery, and through automated image analysis and marking, reduce the time for doctors' preoperative image evaluation and improve the surgical efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a schematic diagram of skeletal image feature matching and change analysis provided by the present invention;

[0041] Figure 2 It is a typical diagram of a 67-year-old male patient with FFPIVc type before, during and after surgery provided by the present invention;

[0042] Figure 3 It is a surgical process diagram of a 70-year-old female patient with FFPIIIc type treated by closed reduction and blocking screw technique for FFPIIIc type pelvic fracture and channel screw fixation provided by the present invention;

[0043] Figure 4 It is the first skeletal image provided by the present invention;

[0044] Figure 5 It is the second skeletal image provided by the present invention;

[0045] Figure 6Schematic structural diagram of an auxiliary positioning system for hollow screw implantation provided by the present invention. Detailed implementation manners

[0046] Next, the technical solutions in the present invention will be clearly and completely described in conjunction with the accompanying drawings in the present invention. Obviously, the described technical solutions are only a part of the present invention, rather than all of them. Based on the technical solutions in the present invention, all other technical solutions obtained by those of ordinary skill in the art without making creative efforts shall fall within the protection scope of the present invention.

[0047] Embodiment 1

[0048] Figure 6 Schematic structural diagram of an auxiliary positioning system for hollow screw implantation provided by the present invention, as Figure 6 shown, an auxiliary positioning system for hollow screw implantation includes an image parameter acquisition module, a bone image registration module, and an abnormality display center; the image parameter acquisition module is connected to the bone image registration module, and the bone image registration module is connected to the abnormality display center.

[0049] The image parameter acquisition module is used to obtain a first bone image picture for preoperative positioning and a second bone image picture obtained most recently before the operation.

[0050] The bone image registration module is used to obtain the bone edge feature points in the image picture through multiple analysis and processing of the bone image picture, and analyze and judge whether there is bone change according to the bone edge feature points.

[0051] The abnormality display center is used to mark the bone image picture with abnormalities and mark the edge feature points at different positions in the second bone image picture from those in the first bone image picture.

[0052] It should be noted that the first bone image picture and the second bone image picture are collected by high-resolution imaging devices before the operation, such as CT, MRI or high-quality X-ray films.

[0053] Specifically, the bone image registration module includes an image processing unit, a bone image registration unit, and a bone difference detection and extraction unit; the image processing unit is connected to the bone image registration unit, and the bone image registration unit is connected to the bone difference detection and extraction unit.

[0054] The image processing unit is used to perform denoising and normalization processing on the first bone image picture and the second bone image picture respectively.

[0055] The bone image registration unit is used to extract the bone gray - scale gradient edges in the first bone image and the second bone image through edge detection, and based on the bone gray - scale gradient edges, respectively mark the first edge feature points {p i} of the first bone image and the second edge feature points {q j} of the second bone image.

[0056] The bone difference detection and extraction unit is used to calculate the bone image difference based on the first edge feature points and the second edge feature points, and determine whether there is a bone change by classifying and analyzing the edge feature points at the same position and different positions.

[0057] Denoising and normalization processing make the first bone image and the second bone image consistent in brightness, contrast and scale, avoiding the influence of different imaging devices, changing shooting angles or noise on the image quality, and ensuring the accuracy of subsequent analysis; the bone image registration unit extracts the bone gray - scale gradient edges through edge detection, effectively retaining the structural features of the bone and removing background noise or other tissue interferences; the bone difference detection and extraction unit calculates the bone image difference based on the changes of the first and second edge feature points, accurately identifies slight displacements or deformations after fractures, and determines whether the bone has changed by classification analysis (comparing the feature points at the same position and different positions), and marks the abnormal areas to help doctors quickly lock in potential error areas; by accurately analyzing the bone changes, doctors can make dynamic adjustments before surgery to prevent the influence of implantation angle errors caused by slight offsets of the fracture position on the surgical effect.

[0058] Specifically, the bone image registration unit is used to extract the bone gray - scale gradient edges in the first bone image and the second bone image through edge detection, and based on the bone gray - scale gradient edges, respectively mark the first edge feature points {p i} of the first bone image and the second edge feature points {q j} of the second bone image. The specific steps are as follows:

[0059] Use the edge detection operator to calculate the gradient amplitude G(x, y) of each pixel point in the first bone image and the second bone image, and regard all pixel points with non - zero gradient amplitude values as preliminary edge points.

[0060] Set the first threshold range for the edge pixel points in the first bone image as The second threshold range for the edge pixel points in the second bone image as

[0061] If the gradient amplitude of the preliminary bone edge points in the first bone image is greater than or equal to Then mark this edge point as "p high”; if the gradient amplitude of the preliminary bone edge point in the first bone image is within the range, mark this bone edge point as "p low "; if the gradient amplitude of the preliminary bone edge point in the first bone image is less than then discard this bone edge point.

[0062] If the gradient amplitude of the preliminary bone edge point in the second bone image is greater than or equal to then mark this bone edge point as "q high "; if the gradient amplitude of the preliminary bone edge point in the second bone image is within the range, mark this bone edge point as "q low "; if the gradient amplitude of the preliminary bone edge point in the second bone image is less than then discard this bone edge point.

[0063] For the bone edge point "p low ", if there is at least one bone edge point "p high " in its 8-neighborhood, then retain this bone edge point; if there is no bone edge point "p high " in its 8-neighborhood, then discard this bone edge point.

[0064] For the bone edge point "q low ", if there is at least one bone edge point "q high " in its 8-neighborhood, then retain this bone edge point; if there is no bone edge point "q high " in its 8-neighborhood, then discard this bone edge point.

[0065] Count the remaining "p high ", "p low ", "q high " and "q low " after the above operations. Classify "p high ", "p low " as the first edge feature points {p i} of the first bone image, and classify "q high ", "q low " as the second edge feature points {q j} in the second bone image.

[0066] Edge detection using the gradient magnitude can effectively identify fracture sites, joint contours, and key bony structures, ensuring the integrity of important bone information during the registration process. By setting a threshold range, low-gradient noise points can be excluded, enhancing the stability of edge detection and making the bone structure clearer. Different threshold ranges are set for the first image and the second image, which can adapt to bone images under different imaging devices and exposure conditions, ensuring the consistency of the registration effect. Using 8-neighborhood screening ensures that low-gradient edge points (p_low, q_low) are only retained when there are high-gradient edge points (p_high, q_high) around them, thus removing discrete noise points and making the bone edges smoother and more accurate. After screening, only the final feature point sets of the first and second bone images are retained, ensuring that the feature points used in the registration process are all valid bone edge points, which helps to improve the matching accuracy of the bone image registration unit and reduce problems such as bone misalignment and deformation analysis deviation caused by incorrect feature points. During the hollow screw implantation process, the fracture site may undergo slight displacement or deformation. If the bone image registration is inaccurate, it will lead to implant angle deviation, unstable fixation, and even secondary injury. The above method can accurately align the bone images and mark the fracture change area, thus helping doctors to adjust the implantation direction in real time and ensuring that the screw implantation angle and depth are consistent with the surgical plan, improving the surgical success rate.

[0067] Specifically, the first threshold range for the edge pixel points in the first bone image picture is wherein, the lowest threshold in the first threshold range and the highest threshold are determined by the formula:

[0068]

[0069] In the formula: is the lowest threshold in the first threshold range, is the highest threshold in the first threshold range, P is the number of pixel points in the first bone image picture, G p (x,y) is the gradient magnitude of the p-th pixel point in the first bone image picture, is the minimum gradient magnitude in the first bone image picture, is the maximum gradient magnitude in the first bone image picture.

[0070] The second threshold range for the edge pixel points in the second bone image picture is wherein, the lowest threshold in the second threshold range and the highest threshold are determined by the formula:

[0071]

[0072] In the formula: is the lowest threshold in the second threshold range, is the highest threshold in the second threshold range, Q is the number of pixel points in the second skeletal image, and G q (x, y) is the gradient magnitude of the q-th pixel point in the second skeletal image, is the minimum gradient magnitude in the second skeletal image, is the maximum gradient magnitude in the second skeletal image.

[0073] By setting the threshold range of the edge pixel points of the first image and the second image, the accuracy of image matching can be effectively enhanced, and the ability to judge the changes in skeletal images can be improved; through adaptive threshold calculation (normalized calculation based on the global minimum gradient and maximum gradient), it is ensured that the gradient features between different images can be effectively compared, and the matching degree between images can be improved; by setting the dynamic threshold range, the fracture edge points can be better extracted, the accurate positioning of the fracture position can be improved, which is helpful for preoperative planning and postoperative recovery monitoring; through adaptive threshold setting, the matching effect of different modality images can be improved, so that the system can be widely applied to a variety of medical image analysis tasks.

[0074] Specifically, the skeletal difference detection and extraction unit is used to calculate the skeletal image difference based on the first edge feature points and the second edge feature points, and determine whether there is a skeletal change by classifying and analyzing the edge feature points at the same position and different positions. The calculation of the skeletal image difference for the edge feature points at the same position is:

[0075] ΔI(x, y) = I q (x, y) - I p (x, y);

[0076] In the formula: ΔI(x, y) is the skeletal image difference at the coordinate position (x, y), and I p (x), y) is the pixel value of the skeletal image at the position (x, y) in the first skeletal image, and I q (x, y) is the pixel value of the skeletal image at the position (x, y) in the second skeletal image.

[0077] For the calculation of the skeletal image difference for the edge feature points at different positions, if the edge feature point at the position (x a , y a ) is in the first skeletal image but not in the second skeletal image, then there is a skeletal image difference ΔI(x a , y a ) = I p (x a , y a ) - 0; if the position (x b,y b ) If the edge feature point at () is in the second skeletal image but not in the first skeletal image, then there is a difference in skeletal images ΔI(x b ,y b ) = 0 - I q (x b ,y b ).

[0078] Based on the above operation steps, summarize the differences in skeletal images, and construct a skeletal change judgment model through the differences in skeletal images to judge whether the skeleton has changed. The formula of the skeletal change judgment model is:

[0079]

[0080] In the formula: β G is the skeletal change value, H is the number of edge feature points at the same position, ΔI(x, y) h is the difference in skeletal images of the h-th edge feature point at the same position, A is the number of edge feature points that are in the first skeletal image but not in the second skeletal image, ΔI(x a ,y a ) is the difference in skeletal images of the edge feature point at the position (x a ,y a ), B is the number of edge feature points that are in the second skeletal image but not in the first skeletal image, ΔI(x b ,y b ) is the difference in skeletal images of the edge feature point at the position (x b ,y b ).

[0081] Compare the skeletal change value with the preset skeletal change threshold. If the skeletal change value is greater than or equal to the preset skeletal change threshold, then the skeleton is abnormal, that is, it has changed; if the skeletal change value is less than the preset skeletal change threshold, then the skeleton is normal, that is, it has not changed.

[0082] By calculating the differences in images of feature points at the same position, it is possible to accurately evaluate whether there are morphological changes in the skeleton, such as fracture displacement, bone loss, callus hyperplasia, etc.

[0083] Example 2

[0084] As Figure 4 shown is the first skeletal image of the present invention and Figure 5The second skeletal image picture shown, based on the analysis of the first skeletal image picture and the second skeletal image picture in the above figure, obtains the schematic diagram of skeletal image feature matching and change analysis as shown in Figure 1. Based on the schematic diagram of skeletal image feature matching and change analysis, a broken line graph of skeletal image difference distribution can be seen. The horizontal axis represents different skeletal parts (distal condyle tibia, upper part of the diaphysis, fracture site, lower part of the diaphysis, distal epiphysis), and the vertical axis represents the image intensity difference (pixel value change). The blue image in the broken line graph represents the change trend of pixel values in the first skeletal image picture, and the blue image in the broken line graph represents the change trend of pixel values in the second skeletal image picture; select the pixel values of the following 10 points as shown in the following table to calculate the skeletal change value:

[0085]

[0086]

[0087] Obtain the image difference of the edge feature points that are in the first skeletal image picture but not in the second skeletal image picture and the image difference of the edge feature points that are in the second skeletal image picture but not in the first skeletal image picture, and import them into the skeletal change judgment model to calculate the skeletal change value.

[0088] In the embodiment of the present invention, by obtaining the first skeletal image picture for preoperative positioning and the second skeletal image picture newly obtained before surgery, obtaining the skeletal edge feature points in the image picture through multiple analysis and processing of the skeletal image picture, and analyzing and judging whether there is skeletal change according to the skeletal edge feature points, marking the skeletal image picture with abnormalities, and marking the edge feature points at different positions in the second skeletal image picture from those in the first skeletal image picture, it is possible to accurately mark the skeletal change area, reduce intraoperative deviation, improve the accuracy of hollow screw implantation, reduce the risk of secondary repair, screw loosening or displacement after surgery, and reduce the time for doctors' preoperative image evaluation through automated image analysis and marking, thereby improving the surgical efficiency.

[0089] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claimed rights.

[0090] Finally: The above is only the preferred solution of the present invention and is not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An auxiliary positioning system for hollow screw implantation, comprising an imaging parameter acquisition module, a bone image registration module, and an abnormal display center; characterized in that, The image parameter acquisition module is used to obtain the first bone image picture for preoperative positioning and the second bone image picture newly obtained before surgery; the bone image registration module is used to obtain the bone edge feature points in the image picture through multiple analysis and processing of the bone image picture, and analyze and judge whether there is bone change according to the bone edge feature points; the bone image registration module includes a bone image registration unit and a bone difference detection and extraction unit; the bone image registration unit is used to extract the bone gray gradient edges in the first bone image picture and the second bone image picture through edge detection, and respectively mark the first edge feature points {p i} of the first bone image picture and the second edge feature points {q j} of the second bone image picture; the bone difference detection and extraction unit is used to calculate the bone image difference based on the first edge feature points and the second edge feature points, and analyze and judge whether there is bone change by classifying and analyzing the edge feature points at the same position and different positions: normalize the absolute value of the difference between all edge feature points at the same position, and divide it by the number of edge feature points at the same position, and at the same time consider the sum of the differences of the unique edge feature points in the first image and the sum of the differences of the unique edge feature points in the second image to form the final bone change judgment value.

2. The auxiliary positioning system for hollow screw implantation according to claim 1, wherein The bone image registration unit is used to extract the bone gray-scale gradient edges in the first bone image and the second bone image through edge detection, and respectively mark the first edge feature points {p i} of the first bone image and the second edge feature points {q j} of the second bone image. The specific steps are as follows: Calculate the gradient magnitude G(x, y) of each pixel in the first skeletal image and the second skeletal image using an edge detection operator, and regard all pixels with non-zero gradient magnitude values as preliminary edge points; The set first threshold range for the edge pixel points in the first skeletal image is The second threshold range for the edge pixel points in the second skeletal image is If the gradient amplitude of the preliminary bone edge point in the first bone image is greater than or equal to then mark this edge point as "p high "; if the gradient amplitude of the preliminary bone edge point in the first bone image is within then mark this bone edge point as "p low "; if the gradient amplitude of the preliminary bone edge point in the first bone image is less than then discard this bone edge point; If the gradient amplitude of the preliminary bone edge point in the second bone image is greater than or equal to then mark this bone edge point as "q high "; if the gradient amplitude of the preliminary bone edge point in the second bone image is within then mark this bone edge point as "q low "; if the gradient amplitude of the preliminary bone edge point in the second bone image is less than then discard this bone edge point.

3. The auxiliary positioning system for cannulated screw implantation according to claim 2, wherein, For the bone edge point "p low ", if there is at least one bone edge point "p high " within its 8-neighborhood, then retain this bone edge point; If there is no bone edge point "p" in its 8 - neighborhood high ", then discard this bone edge point; For the bone edge point "q low ", if there is at least one bone edge point "q high " within its 8-neighborhood, then retain this bone edge point; If there is no bone edge point "q" in its 8-neighborhood high ", then discard this bone edge point; Count the remaining "p" after the above operations high ", "p low ", "q high ", and "q low ". Classify "p high ", "p low " as the first edge feature points {p i} of the first skeletal image, and classify "q high ", "q low " as the second edge feature points {q j} of the second skeletal image.

4. The auxiliary positioning system for hollow screw implantation according to claim 2, wherein, Set the first threshold range for the edge pixel points in the first skeletal image as wherein, the lowest threshold in the first threshold range and the highest threshold are determined by the following method: Minimum threshold Calculated based on the relative difference between the gradient magnitudes of all pixel points in the first bone image and the minimum gradient magnitude. Divide the difference between the gradient magnitude of each pixel point and the minimum gradient magnitude in the image by the minimum gradient magnitude, take the absolute value, sum over all pixel points, and finally take the average to obtain the minimum threshold; Maximum threshold The calculation method is to divide the difference between the gradient magnitude of each pixel point and the maximum gradient magnitude in the first bone image by the maximum gradient magnitude, take the absolute value, accumulate, and then take the average to obtain it.

5. The auxiliary positioning system for hollow screw implantation according to claim 2, wherein The second threshold range of the edge pixel points in the second bone image is wherein, the lowest threshold in the second threshold range and the highest threshold are determined as follows: Minimum threshold Calculated based on the relative difference between the gradient magnitudes of all pixel points in the second skeletal image and the minimum gradient magnitude. Divide the difference between the gradient magnitude of each pixel point and the minimum gradient magnitude in the image by the minimum gradient magnitude, take the absolute value, sum over all pixel points, and finally take the average to obtain the minimum threshold; Maximum threshold The calculation method is to divide the difference between the gradient magnitude of each pixel point and the maximum gradient magnitude in the second skeletal image by the maximum gradient magnitude, take the absolute value, accumulate, and then take the average to obtain it.

6. The auxiliary positioning system for hollow screw implantation according to claim 1, wherein, The skeletal difference detection extraction unit is used to calculate the skeletal image difference based on the first edge feature points and the second edge feature points, and determine whether there is a skeletal change by classifying and analyzing the edge feature points at the same position and different positions. The calculation of the skeletal image difference for the edge feature points at the same position is as follows: ΔI(x,y) = I q (x,y) - I p (x,y); Where: ΔI(x,y) is the difference of the bone image at the coordinate position (x,y), and I p (x,y) is the pixel value of the bone image at the position (x,y) in the first bone image, and I q (x,y) is the pixel value of the bone image at the position (x,y) in the second bone image; Calculate the bone image difference for edge feature points at different positions. If the edge feature point at position (x a , y a ) is in the first bone image but not in the second bone image, then the bone image difference ΔI(x a , y a ) = I p (x a , y a ) - 0; if the edge feature point at position (x b , y b ) is in the second bone image but not in the first bone image, then the bone image difference ΔI(x b , y b ) = 0 - I q (x b , y b ). Summarize the skeletal image differences based on the above operation steps, and construct a skeletal change judgment model through the skeletal image differences to judge whether the skeleton has changed.

Images