A C-arm navigation and positioning system for orthopedic surgery

Abstract

The present invention relates to the field of image recognition technology, and in particular to a C-arm navigation and positioning system for orthopedic surgery. The system searches for the endpoint position of the spine in the upper part of the patient's torso image under X-ray, and can screen out the accurate lateral area of the top of the spine by grayscale value difference and pixel position in the horizontal direction, and then search downward to obtain the spinal region based on the grayscale value difference. According to the position distribution of pixel points on the spinal region, the scoliosis region can be screened out and the degree of scoliosis can be obtained. By quantifying the degree of scoliosis, the effective number of screws to be inserted and the screw position can be planned, thereby achieving accurate navigation and positioning of the C-arm. The present invention accurately identifies the spinal region, determines the scoliosis region and effectively quantifies the degree of scoliosis, thereby obtaining the appropriate number of screws to be inserted in the scoliosis region and determining the screw position for C-arm navigation.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition, and in particular to a C-arm navigation and positioning system for orthopedic surgery. Background Art

[0002] The C-arm is a medical device used in orthopedic scoliosis correction. Doctors use it to insert screws into the curved area of the spine, thereby correcting the patient's scoliosis. The C-arm can be combined with an X-ray machine to provide real-time imaging of the spinal region. Based on this real-time imaging, doctors can determine the location of the curve and install the screws.

[0003] To improve the intelligence of C-arm scoliosis correction, existing technologies use computer vision technology to automatically identify the scoliosis area and then navigate and locate the screw position. However, because X-rays obtain images of the patient's body, including information on multiple bones such as the cervical spine, spine, and ribs, the image information is relatively complex. Moreover, the formation of the scoliosis area may cause the spine to deviate from other areas such as the ribs, making it difficult to identify the spine and accurately determine the scoliosis area and navigate the C-arm. Summary of the Invention

[0004] In order to solve the technical problem in the prior art that the spine and scoliosis area cannot be accurately and effectively identified due to the complex torso image information under X-ray of scoliosis patients, which leads to inaccurate screw positioning and navigation of the C-arm, the purpose of the present invention is to provide a C-arm navigation and positioning system for orthopedic surgery. The technical solution adopted is as follows:

[0005] The present invention provides a C-arm navigation and positioning system for orthopedic surgery, the system comprising:

[0006] The spine endpoint recognition module is used to perform threshold segmentation on the patient's X-ray body image to obtain a binary body image; select a starting point from the foreground pixels at the top of the binary body image to search downward, select foreground pixels in a preset conditional direction as new traversal points, and traverse downward until the direction of the line connecting the traversal points meets the preset spine edge direction interval, and select the spine endpoint from the traversal points;

[0007] A spinal region identification module is configured to obtain a spinal top transverse region based on pixel positions and grayscale values in a transverse direction of the spinal endpoint in a torso image; and to search downward from the spinal top transverse region based on grayscale value differences to obtain a spinal region;

[0008] A scoliosis region analysis module is used to screen out scoliosis regions in the spinal region based on the position distribution of pixel points in the transverse direction; and to obtain the degree of scoliosis in the scoliosis region based on the proportion of the scoliosis region in the spinal region and the degree of positional deviation;

[0009] The position navigation and positioning module is used to plan the number of screws to be inserted and the screw positions according to the degree of scoliosis, and to navigate and position the C-arm according to the screw positions.

[0010] Furthermore, the method for screening spinal endpoints includes:

[0011] During the traversal process, if the angles between a preset number of consecutive traversal points are in the preset spine edge direction, the first traversal point among the preset number of traversal points is used as the spine endpoint and the traversal is stopped.

[0012] Furthermore, the method for obtaining the transverse region of the top of the spine includes:

[0013] For a pixel point on the horizontal side of the spinal endpoint, obtain a first grayscale value difference between the pixel point and the spinal endpoint; in the binary image of the body, use the center point of the row where the spinal endpoint is located as a reference point; use the lateral distance between the spinal endpoint and the reference point as a first distance, and use the distance between the pixel point and the reference point as a second distance, and obtain the positioning weight of the pixel point based on the difference between the first distance and the second distance; filter out pixel points belonging to the lateral area of the top of the spine based on the first grayscale value difference and the positioning weight.

[0014] Furthermore, the method for obtaining the spinal region includes:

[0015] For each pixel point in the transverse area of the top of the spine, if the grayscale value difference between the pixel point in the vertical downward direction and the pixel point in the transverse area of the top of the spine is within the preset difference range, the pixel point in the vertical downward direction is used as the lower-level spine pixel point; all lower-level spine pixel points are screened out, and the outermost lower-level spine pixel point is selected as the lower-level spine endpoint, and the lower-level spine area is obtained according to the acquisition method of the transverse area of the top of the spine; based on the lower-level spine area, continue to traverse downward to obtain all lower-level spine areas to constitute the spine area.

[0016] Furthermore, the method for screening the scoliosis area includes:

[0017] Select the center point of the transverse area at the top of the spine as the standard position point, and obtain the first transverse distance between the center point of each row of the spinal area and the standard position point; for each row of the spinal area, select the endpoint on the same side as the spinal endpoint as the comparison endpoint, and obtain the second transverse distance between the comparison endpoint and the spinal endpoint; obtain the degree of transverse position offset of each row according to the first transverse distance and the second transverse distance; screen out the scoliotic spinal rows that constitute the scoliosis area according to the degree of position offset, and continuous scoliotic spinal rows constitute a scoliosis area.

[0018] Furthermore, the method for obtaining the degree of scoliosis includes:

[0019] For each scoliosis area, the maximum position deviation degree and the average position deviation degree of the scoliotic spinal column are obtained; the scoliosis degree is obtained based on the maximum position deviation degree, the average position deviation degree, and the proportion of the scoliosis area in the spinal column area.

[0020] Furthermore, the method for obtaining the number of screws inserted includes:

[0021] The normalized scoliosis degree is multiplied by the preset maximum number of screws, and the result of rounding down the product is the number of screws inserted.

[0022] Furthermore, the method for obtaining the screw position includes:

[0023] The screw positions are evenly arranged in the scoliotic region according to the number of screws inserted.

[0024] Furthermore, the starting point is the first foreground pixel from left to right on the first row of pixels from top to bottom of the binary image of the body.

[0025] Furthermore, the preset condition direction is the leftmost direction in the eight neighborhoods.

[0026] The present invention has the following beneficial effects:

[0027] The present invention takes into account that the upper part of the torso image under X-ray mainly consists of the patient's head and cervical spine area, in which the cervical spine area is connected to the spine and will not show scoliosis. Therefore, the endpoint position of the spine can be found in the upper part, and the spinal region is determined by a strategy of first horizontally and then vertically. The present invention reduces information complexity through threshold segmentation technology, and then determines the spinal endpoint based on the direction between the vertical pixels in the binary image. Considering that the spine is located in the central range of the torso image, the accurate horizontal area of the top of the spine can be screened out in the horizontal direction by grayscale value difference and pixel position, and then searching downward, and the spinal region can be obtained based on the grayscale value difference. Then, based on the position distribution of the pixels in the spinal region, the scoliosis area can be screened and the degree of scoliosis can be obtained. By quantifying the degree of scoliosis, the effective number of screws to be inserted and the screw position can be planned, thereby achieving accurate navigation and positioning of the C-arm. The present invention accurately identifies the spinal region, determines the scoliosis area and effectively quantifies the degree of scoliosis, and then can obtain the appropriate number of screws to be inserted in the scoliosis area and determine the screw position for C-arm navigation. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] 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.

[0029] Figure 1 This is a structural block diagram of a C-arm navigation and positioning system for orthopedic surgery provided by one embodiment of the present invention;

[0030] Figure 2 A schematic diagram of a binary image of a body provided by one embodiment of the present invention. DETAILED DESCRIPTION

[0031] To further illustrate the technical means and effectiveness of the present invention in achieving its intended objectives, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail the specific implementation, structure, features, and effectiveness of a C-arm navigation and positioning system for orthopedic surgery proposed in accordance with the present invention. In the following description, references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0032] 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.

[0033] The specific solution of the C-arm navigation and positioning system for orthopedic surgery provided by the present invention will be described in detail below with reference to the accompanying drawings.

[0034] See also Figure 1 , which shows a block diagram of a C-arm navigation and positioning system for orthopedic surgery provided by an embodiment of the present invention. The system includes: a spinal endpoint recognition module 101, a spinal region recognition module 102, a scoliosis region analysis module 103 and a position navigation and positioning module 104.

[0035] The embodiment of the present invention uses an X-ray machine to collect the patient's body image, which can be processed by the spine endpoint recognition module 101 after pre-processing such as denoising. The spine endpoint recognition module 101 is used to identify the endpoints of the spine to facilitate the subsequent modules to identify the spine area. First, it is necessary to reduce the information complexity in the body image, and obtain the body binary image of the body image through the threshold segmentation algorithm. In the body binary image, the foreground is the segmented bone area with a pixel value of 255; the background is useless information with a pixel value of 0. Please refer to Figure 2 , which shows a schematic diagram of a binary image of a body provided by an embodiment of the present invention.

[0036] like Figure 2 As shown, the upper half of the body binary image and the body image are the patient's cervical vertebra and head area. The cervical vertebra is connected to the spine, and the upper part of the cervical vertebra and the spine generally do not produce scoliosis. In addition, the cervical vertebra and the spine have obvious vertical downward and rectangular shape features, which are significantly different from the curvilinear features of the head. Therefore, the spine endpoint recognition module 101 selects a starting point from the foreground pixels at the top of the body binary image to search downward, selects a foreground pixel in the preset condition direction as a new traversal point to traverse downward, and if the traversal point is a side endpoint of the cervical vertebra or the spine, then its next traversal point should be about 90 degrees with the traversal point. Therefore, during the traversal process, the spine endpoint in the traversal point can be determined by judging whether the direction of the traversal point connection line meets the preset spine edge direction interval. That is, the overall idea of the spine endpoint recognition module 101 can be regarded as searching for a boundary of the foreground in the body binary image, and determining the spine endpoint according to the trend direction on the boundary.

[0037] Preferably, in this embodiment of the present invention, the starting point is the first foreground pixel from left to right in the first row of pixels from top to bottom in the binary image of the body. In this embodiment of the present invention, the preset conditional direction is set to the leftmost side of the eight-neighborhood. That is, for each traversal point, the leftmost foreground pixel in the last row of neighboring pixels in the eight-neighborhood is selected as the new traversal point. The resulting line connecting the traversal points can be considered the left foreground boundary of the binary image.

[0038] Preferably, in an embodiment of the present invention, the method for screening spinal endpoints includes:

[0039] During the traversal process, if the angles between a predetermined number of consecutive traversal points are in the predetermined spine edge direction, the first traversal point among the predetermined number of traversal points is used as the spine endpoint and the traversal is stopped. In this embodiment of the present invention, the predetermined number is set to 5, and the spine edge direction is set to 88 to 92 degrees in the horizontal direction.

[0040] After the spinal endpoints are determined, the spinal region identification module 102 can be used to analyze the features of the pixel points in the horizontal and vertical directions to identify the spinal region. For the spine, it will maintain a fixed pixel value on the body image, and the overall area will be in the middle area of the image. Although scoliosis will occur, it will still be distributed near the center line of the image. Therefore, in the horizontal direction, by comparing the grayscale values of the pixel points with the spinal endpoints, as well as the positions of the pixel points, the top horizontal area of the spine passing through the spinal endpoints can be obtained, that is, the top horizontal area of the spine is the first row of the spinal region. Based on the first row, the search can be continued downward, and the spinal pixel points in the vertical direction can be determined according to the difference in grayscale values, so that the complete spinal region can be obtained.

[0041] Preferably, in an embodiment of the present invention, the method for obtaining the lateral region of the top of the spine includes:

[0042] For a pixel point on the horizontal side of the spine endpoint, the first grayscale value difference between the pixel point and the spine endpoint is obtained; because the spine endpoint has been determined to be a pixel point of the spine, the smaller the first grayscale value difference, the closer the grayscale feature of the pixel point is to the spine endpoint, and the more likely it is to be a spine pixel point.

[0043] In a binary image of the body, the center point of the row containing the spinal endpoint is used as the reference point; the lateral distance between the spinal endpoint and the reference point is used as the first distance, and the distance between the pixel and the reference point is used as the second distance. Because the lateral region at the top of the spine can be considered a row without scoliosis, the first distance between the spinal endpoint and the reference point can be used as a reference distance. If the pixel is a spinal pixel, the second distance must be smaller than the first distance. In other words, the smaller the difference between the first and second distances, the more likely the pixel is a spinal pixel, and the larger the difference, the further the pixel's position deviates from the spinal region. Therefore, the pixel's positioning weight is obtained based on the difference between the first and second distances.

[0044] Pixels belonging to the horizontal area of the top of the spine are screened out based on the first grayscale value difference and the positioning weight.

[0045] In an embodiment of the present invention, the first grayscale value difference is the absolute value of the difference between the two grayscale values. The positioning weight is the absolute value of the difference between the first distance and the second distance. The inverse of the product between the first grayscale value difference and the positioning weight is normalized to obtain the spine pixel probability. The probability threshold is set to 0.7, and the pixels whose spine pixel probability is greater than the probability threshold are used as the pixels of the horizontal area of the top of the spine, thereby forming the horizontal area of the top of the spine.

[0046] It should be noted that the normalization method in the embodiment of the present invention adopts range normalization, and those skilled in the art may also choose other normalization methods, which will not be limited or elaborated here.

[0047] Preferably, in an embodiment of the present invention, the method for acquiring the spinal region includes:

[0048] For each pixel point in the horizontal area of the top of the spine, if the grayscale value difference between the pixel point in the vertical downward direction and the pixel point in the horizontal area of the top of the spine is within the preset difference range, the pixel point in the vertical downward direction is used as the lower-level spine pixel point.

[0049] Furthermore, in order to avoid information overlap between the spine region and the ribs caused by scoliosis, a row consisting of all lower-level spine pixels cannot be directly used as the lower-level spine. In this embodiment of the present invention, all lower-level spine pixels are screened out, and the outermost lower-level spine pixel points are selected as the lower-level spine endpoints. The lower-level spine region is obtained according to the method for obtaining the transverse region of the top of the spine. That is, the relationship between position and pixel value is reconsidered to determine the lower-level spine region, thereby avoiding misidentification caused by confusion of pixel value information.

[0050] Based on the lower spine region, the lower spine regions are traversed downwards to obtain all the lower spine regions to form the spine region. In this embodiment of the present invention, the difference interval is set to 0 to 5, that is, if the absolute value of the difference between the gray values is between 0 and 5, it is determined to be a lower spine pixel.

[0051] When pedicle screws are inserted into the spine, the number and location of the inserted pedicle screws vary depending on the degree of scoliosis. Therefore, the scoliosis region analysis module 103 is used to filter out scoliosis regions within the spinal region based on the positional distribution of pixels in the transverse direction. The degree of scoliosis in the scoliosis region is determined based on the proportion of the scoliosis region within the spinal region and the degree of positional deviation. Based on the degree of scoliosis, the number and location of screws to be inserted can be determined in subsequent modules.

[0052] Preferably, in an embodiment of the present invention, the method for screening the scoliosis area includes:

[0053] Because the transverse region of the top of the spine is an area where scoliosis does not occur, the center point of the transverse region of the top of the spine is selected as the standard position point.

[0054] The first transverse distance between the center point of each row in the spinal region and the standard position point is obtained. That is, the larger the first transverse distance, the more likely scoliosis is to occur in that row.

[0055] For each row in the spinal region, the endpoint on the same side as the spinal endpoint is selected as the comparison endpoint, and the second transverse distance between the comparison endpoint and the spinal endpoint is obtained. A larger second transverse distance indicates that scoliosis is more likely to occur in that row.

[0056] The degree of lateral positional deviation for each row is determined based on the first and second lateral distances. The degree of positional deviation is statistically integrated from two perspectives: endpoint-to-endpoint comparison and midpoint-to-midpoint comparison. A greater degree of positional deviation indicates that the row is more likely to correspond to the scoliotic region.

[0057] Scoliotic spinal rows that constitute a scoliotic region are selected based on the degree of positional deviation. Consecutive scoliotic spinal rows constitute a scoliotic region. In this embodiment of the present invention, the product of the first lateral distance and the second lateral distance is normalized to obtain the degree of positional deviation. A deviation threshold is set to 0.3. When the degree of positional deviation exceeds the deviation threshold, the corresponding row is determined to be a scoliotic spinal row.

[0058] Preferably, in an embodiment of the present invention, the method for obtaining the degree of scoliosis includes:

[0059] For each scoliosis area, the maximum positional deviation and the average positional deviation of the scoliotic spinal column are obtained; the degree of scoliosis is obtained based on the maximum positional deviation, the average positional deviation, and the proportion of the scoliotic area in the spinal column area. In an embodiment of the present invention, the product of the maximum positional deviation, the average positional deviation, and the proportion of the scoliotic area in the spinal column area is used as the degree of scoliosis. It should be noted that the maximum positional deviation is the maximum value of the positional deviations of all rows in the scoliosis area, and the average positional deviation is the average value of the positional deviations of all rows.

[0060] The position navigation module 104 is used to plan the number of screws to be inserted and the positions of the screws according to the degree of scoliosis, and to navigate and locate the position of the C-arm based on the screw positions.

[0061] Preferably, in an embodiment of the present invention, since a greater degree of scoliosis requires more pedicle screws, the degree of scoliosis is normalized and then multiplied by a preset maximum number of screws, and the product is rounded down to the nearest integer, which is the number of screws to be inserted. In an embodiment of the present invention, the maximum number of screws is set to 10.

[0062] In an embodiment of the present invention, screw positions are evenly distributed across the scoliotic region based on the number of screws to be inserted. Using existing navigation algorithms, the C-arm can be navigated and positioned, automatically moving to the position where the screws are to be inserted and awaiting further control by the surgeon. Once one screw is inserted, the C-arm automatically moves to the next screw position until the surgery is complete.

[0063] In summary, the embodiment of the present invention searches for the endpoint position of the spine in the upper part of the patient's torso image under X-ray, and can screen out the accurate lateral area of the top of the spine by grayscale value difference and pixel position in the horizontal direction, and then search downward, and the spinal region can be obtained based on the grayscale value difference. According to the position distribution of pixel points on the spinal region, the scoliosis region can be screened out and the degree of scoliosis can be obtained. By quantifying the degree of scoliosis, the effective number of screws to be inserted and the screw position can be planned, thereby achieving accurate navigation and positioning of the C-arm. The present invention accurately identifies the spinal region, determines the scoliosis region and effectively quantifies the degree of scoliosis, thereby obtaining the appropriate number of screws to be inserted in the scoliosis region and determining the screw position for C-arm navigation.

[0064] 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.

[0065] 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. A C-arm navigation and positioning system for orthopedic surgery, characterized in that: The system comprises: The spine endpoint recognition module is used to perform threshold segmentation on the patient's X-ray body image to obtain a binary body image; select a starting point from the foreground pixels at the top of the binary body image to search downward, select foreground pixels in a preset conditional direction as new traversal points, and traverse downward until the direction of the line connecting the traversal points meets the preset spine edge direction interval, and select the spine endpoint from the traversal points; A spinal region identification module is configured to obtain a spinal top transverse region based on pixel positions and grayscale values in a transverse direction of the spinal endpoint in a torso image; and to search downward from the spinal top transverse region based on grayscale value differences to obtain a spinal region; A scoliosis region analysis module is used to screen out scoliosis regions in the spinal region based on the position distribution of pixel points in the transverse direction; and to obtain the degree of scoliosis in the scoliosis region based on the proportion of the scoliosis region in the spinal region and the degree of positional deviation; A position navigation and positioning module is used to plan the number of screws to be inserted and the positions of the screws according to the degree of scoliosis, and to navigate and position the C-arm according to the positions of the screws; The screening method for the scoliosis area includes: Select the center point of the transverse area of the top of the spine as the standard position point, and obtain the first transverse distance between the center point of each row of the spine area and the standard position point; for each row of the spine area, select the endpoint on the same side as the spine endpoint as the comparison endpoint, and obtain the second transverse distance between the comparison endpoint and the spine endpoint; obtain the degree of positional offset of each row in the transverse direction according to the first transverse distance and the second transverse distance; screen out the scoliotic spine rows that constitute the scoliosis area according to the degree of positional offset, and continuous scoliotic spine rows constitute a scoliosis area; The method for obtaining the transverse region of the top of the spine comprises: For a pixel point on the lateral side of the spinal endpoint, a first grayscale value difference between the pixel point and the spinal endpoint is obtained; in the binary image of the body, the center point of the row where the spinal endpoint is located is used as a reference point; the lateral distance between the spinal endpoint and the reference point is used as a first distance, and the distance between the pixel point and the reference point is used as a second distance, and the positioning weight of the pixel point is obtained according to the difference between the first distance and the second distance; the pixel points belonging to the lateral area of the top of the spine are screened out according to the first grayscale value difference and the positioning weight; wherein the inverse of the product of the first grayscale value difference and the positioning weight is normalized to obtain the probability of the spinal pixel point, and the pixel point whose spinal pixel point probability is greater than the probability threshold is used as the pixel point of the lateral area of the top of the spine; The method for obtaining the spinal region includes: For each pixel point in the transverse area of the top of the spine, if the grayscale value difference between the pixel point in the vertical downward direction and the pixel point in the transverse area of the top of the spine is within the preset difference range, the pixel point in the vertical downward direction is used as the lower-level spine pixel point; all lower-level spine pixel points are screened out, and the outermost lower-level spine pixel point is selected as the lower-level spine endpoint, and the lower-level spine area is obtained according to the acquisition method of the transverse area of the top of the spine; based on the lower-level spine area, continue to traverse downward to obtain all lower-level spine areas to constitute the spine area.

2. A C-arm navigation and positioning system for orthopedic surgery according to claim 1, characterized in that: The screening method for the spinal endpoints includes: During the traversal process, if the angles between a preset number of consecutive traversal points are in the preset spine edge direction, the first traversal point among the preset number of traversal points is used as the spine endpoint and the traversal is stopped.

3. The C-arm navigation and positioning system for orthopedic surgery according to claim 1, characterized in that: The method for obtaining the degree of scoliosis includes: For each scoliosis area, the maximum position deviation degree and the average position deviation degree of the scoliotic spinal column are obtained; the scoliosis degree is obtained based on the maximum position deviation degree, the average position deviation degree, and the proportion of the scoliosis area in the spinal column area.

4. The C-arm navigation and positioning system for orthopedic surgery according to claim 1, characterized in that: The method for obtaining the number of screws inserted includes: The normalized scoliosis degree is multiplied by the preset maximum number of screws, and the result of rounding down the product is the number of screws inserted.

5. The C-arm navigation and positioning system for orthopedic surgery according to claim 1, characterized in that: The method for obtaining the screw position includes: The screw positions are evenly arranged in the scoliotic region according to the number of screws inserted.

6. The C-arm navigation and positioning system for orthopedic surgery according to claim 1, characterized in that: The starting point is the first foreground pixel from left to right on the first row of pixels from top to bottom of the binary image of the body.

7. The C-arm navigation and positioning system for orthopedic surgery according to claim 6, characterized in that: The preset condition direction is the leftmost direction in the eight neighborhoods.

Images