A method for locating sacral nerve-free channels based on CPD point cloud registration technology

Abstract

The present invention relates to the field of positioning technology and provides a method for locating sacral nerve-free channels based on CPD point cloud registration technology, comprising the following steps: Step 1, data acquisition; Step 2, three-dimensional image point cloud processing; Step 3, first sacral foramen position annotation; Step 4, point cloud registration; Step 5, calculation of nerve-free channels. The present invention has the beneficial effects of being able to process three-dimensional CT images, and the processed three-dimensional model can be fully matched with a human sacral model. The nerve-free channels can be comprehensively determined from three dimensions: coronal, sagittal, and transverse. Compared to methods based on X-ray images, this method is more accurate, and the channel conditions can be intuitively visualized from the three-dimensional model.

Description

Technical Field

[0001] The present invention relates to the field of positioning technology, and in particular to a positioning method for a sacrum without nerve channels based on CPD point cloud registration technology. Background Art

[0002] During sacral surgery or procedures, it is crucial to accurately locate the nerve-free pathways in the sacrum to effectively avoid nerve damage. The sacral region has a complex neural structure, with dense distribution of nerves such as the sacral plexus and sacral nerve roots. Once damaged, it can cause serious functional impairment.

[0003] Specifically, in the current sacral neuromodulation (SNM) procedure, it is necessary to puncture and locate the S3 sacral foramen of the sacrum. The main technology is the cross positioning method under X-ray fluoroscopy. That is, under the anteroposterior fluoroscopy of the X-ray image, the sacral midline is determined and marked, the sacral joint is determined, and then a line is drawn connecting the lower ends of the bilateral sacral joints. The intersection of the connecting line and the midline is then calculated to be about 2 cm to the left and right, which is the location of the S3 sacral nerve foramen.

[0004] The main shortcomings of this method for locating the sacral foramen include three aspects. First, this method is only applicable to two-dimensional X-ray images, but not to CT data. Second, the position of the sacral neural foramen in this method is located 2 cm to the left and right of the intersection of the connecting line and the midline. The sacrum varies greatly from person to person, and the 2 cm position cannot be applicable to all people, so the generalization of this method is poor. Third, this method can only locate the third sacral foramen. When positioning above the first sacral foramen and the upper edge of the sacrum is required, this method will not work.

[0005] Application No. 201710520171.0 provides a device and method for locating the S3 sacral nerve cross, effectively using manual palpation or X-ray equipment to determine the lower edge of the sacroiliac joint and the sacrum, and to locate the S3 sacral nerve foramen. However, this application is limited to locating the third sacral foramen and cannot be used to locate the nerve-free channels above the first sacral foramen or at the upper edge of the sacrum. Summary of the Invention

[0006] The current problem is that the sacrum has no nerve channels and can only be calculated manually based on experience, which has low accuracy.

[0007] The present invention provides a method for locating a sacral nerve-free channel based on CPD point cloud registration technology, comprising the following steps:

[0008] Step 1, data acquisition: obtaining a 3D CT scan image of the sacrum;

[0009] Step 2, 3D image point cloud processing: The acquired CT image is processed into point cloud data using CPD point cloud registration technology, and the pixel positions of the CT image are mapped to 3D coordinates;

[0010] Step 3, marking the first sacral foramen: mark the first sacral foramen of the sacrum, denoted as hole_target;

[0011] Step 4, point cloud registration: The sacral point cloud data with the marked sacral foramen is used as the registration target, denoted as target, and the sacral point cloud data without the marked sacral foramen is used as source. The CPD point cloud registration algorithm is used for point cloud registration to obtain the mapping relationship between the configured source and target, denoted as f(°). Then, the data of the first marked sacral foramen is input into the mapping relationship f(°) to obtain hole_source = f(hole_target). The coordinates of hole_source are the position of the first sacral foramen of the unmarked sacrum.

[0012] Step 5, calculate the nerve-free channel: solve the maximum value of the z-axis coordinate of the first sacral foramen, that is, max(hole_source(z)). After obtaining the coordinates of the upper edge of the first sacral foramen, make sections along the z-axis and y-axis based on the upper edge of the first sacral foramen; then make sections along the z-axis and y-axis based on the tangent line with the minimum y-axis value of the sacrum; then make sections along the z-axis based on the leftmost and rightmost tangent lines of the sacrum. The intersecting rectangular area formed by the six sections is the nerve-free area of ​​the sacrum; then calculate the coordinates of the center points of the leftmost and rightmost sections in the rectangular area, and draw a straight line connecting the center points of the two sections. This straight line is the nerve-free channel of the sacrum.

[0013] As a preferred solution, the mapping method is:

[0014] 1) Calculate the total number of CT image slices as z_max, and calculate the width of the horizontal axis of each slice in the CT image as w_max and the length of the vertical axis as h_max;

[0015] 2) Convert the pixel position of each slice in the CT image into coordinates in three-dimensional space. The original pixel coordinates are recorded as p_o=(x_i, y_j, z_k), where z_k represents the kth slice. The conversion formula is:

[0016] x_new = x_i-(1+w_max) / 2;

[0017] y_new = y_j-(1+h_max) / 2;

[0018] z_new = z_k-(1+z_max) / 2;

[0019] The converted coordinates p_new=(x_new, y_new, z_new) are the three-dimensional coordinates of each image in the original CT image;

[0020] 3) Create a new three-dimensional matrix point=(w_max,h_max,z_max), assign the point of p_new coordinate to 1, and assign the other points to 0, then you will get the discretized point cloud coordinates, that is, the point cloud data of the sacrum.

[0021] As a preferred solution, the three-dimensional matrix point=(w_max, h_max, z_max) is established in a python programming language environment.

[0022] As a preferred solution, the data acquisition in step 1 adopts Python to read the three-dimensional sacrum model, and the core command for reading the three-dimensional sacrum is ds = pydicom.dcmread("digu.dcm").

[0023] The beneficial effects of the present invention are:

[0024] 1. The present invention can process three-dimensional CT images. The processed three-dimensional model can be fully matched with the human sacrum model. The nerve-free channel can be comprehensively determined from the coronal, sagittal and transverse dimensions. Compared with the method based on X-ray images, this method is more accurate and the channel condition can be intuitively seen from the three-dimensional model.

[0025] 2. The proposed method for locating the first sacral foramen based on point cloud registration technology is more robust and generalizable. Compared with the method for locating the S3 sacral foramen in X-ray images, it does not require rigid settings such as a fixed 2 cm width. The shape and size of individual sacrums vary greatly, and rigid distance settings are not applicable to all sacral models. This method can locate the first sacral foramen for any sacral model. DETAILED DESCRIPTION

[0026] To illustrate the features of the present invention, the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0027] Example 1:

[0028] This embodiment provides a method for automatically calculating the sacral surgical screw channel based on CPD point cloud registration technology to locate the nerve-free channel of the sacrum;

[0029] Step 1, data acquisition: Get a 3D CT scan image of the sacrum and use Python to read the 3D sacral model. First, install pydicom (pip install pydicom). The core command for Python to read the 3D sacrum is ds = pydicom.dcmread("digu.dcm"), and the Python function to read the model is:

[0030] import pydicom

[0031] import os

[0032] def read_dicom_series(dicom_dir):

[0033] series = []

[0034] for dcm_file in os.listdir(dicom_dir):

[0035] if dcm_file.endswith('.dcm'):

[0036] file_path = os.path.join(dicom_dir, dcm_file)

[0037] dcm = pydicom.dcmread(file_path)

[0038] series.append(dcm)

[0039] return series;

[0040] Step 2, 3D image point cloud processing: The acquired CT image is processed into point cloud data using CPD point cloud registration technology. In the read CT image, each image is a slice of 3D data. The pixel positions of the CT image are mapped to 3D coordinates. The mapping method in this embodiment is (taking two points as an example):

[0041] 1) 50 slices of CT image are denoted as z_max=50, the width of the horizontal axis of each slice in the CT image is w_max=20, and the length of the vertical axis is h_max=20;

[0042] 2) Convert the pixel position of each CT slice into three-dimensional coordinates. The coordinates of a point on the upper edge of the left foramen of the first sacral foramen are p1 = (5, 4, 30), and the coordinates of a point on the upper edge of the right foramen are p2 = (15, 4, 30). The conversion formula is (discard the decimal point):

[0043] x1_new =5-(1+20) / 2=-5

[0044] y1_new =4-(1+20) / 2=-6

[0045] z1_new =20-(1+50) / 2=4

[0046] x2_new =15-(1+20) / 2=4

[0047] y2_new =4-(1+20) / 2=-6

[0048] z2_new =30-(1+50) / 2=4

[0049] The converted coordinates p_new = (x_new, y_new, z_new) are the three-dimensional coordinates of each image in the original CT image (the center point of the three-dimensional model is the origin), that is, the coordinates of p1 will become p1' = (-5, -6, 4), and p2 will become p2' = (4, -6, 4);

[0050] 3) In the Python programming language environment, create a new three-dimensional matrix point=(w_max,h_max,z_max), assign the point with the p_new coordinate to 1 and the other points to 0, and the discretized point cloud coordinates are obtained;

[0051] 4) Save the 3D matrix in .stl format to obtain the point cloud data of the sacrum.

[0052] The Python toolkits required in this embodiment are pydicom, numpy, and tqdm, which can be installed using pip install pydicom, numpy, and tqdm.

[0053] Step 3: Mark the first sacral foramen: Use cloudcompareV to mark the first sacral foramen of the sacrum, denoted as hole_target;

[0054] Step 4, point cloud registration: The sacral point cloud data with the marked sacral foramen is used as the registration target, denoted as target, and the sacral point cloud data without the marked sacral foramen is used as source. The CPD point cloud registration algorithm is used for point cloud registration to obtain the mapping relationship between the configured source and target, denoted as f(°). Then, the data of the first marked sacral foramen is input into the mapping relationship f(°) to obtain hole_source = f(hole_target). The coordinates of hole_source are the position of the first sacral foramen of the unmarked sacrum.

[0055] Step 5, calculate the nerve-free channel: solve the maximum value of the first sacral foramen z-axis coordinate, that is, max(hole_source(z)). After obtaining the coordinates of the upper edge of the first sacral foramen, make sections along the z-axis and y-axis respectively according to the upper edge of the first sacral foramen; then make sections along the z-axis and y-axis according to the tangent line of the sacrum with the minimum value of the y-axis; then make sections along the z-axis according to the leftmost tangent line and the rightmost tangent line of the sacrum. Making sections along the z-axis means extending the line connecting the upper edge point p1' of the left hole in the first sacral foramen and the upper edge point p2' of the right hole in the first sacral foramen along the z-axis. Extend one step, Then the coordinate of the z-axis is increased by 1, that is, the coordinate of p1' will become (-5,-6,5), and the coordinate of p2' will become p2'=(4,-6,5). To make a section along the y-axis, the line connecting points p1' and p2' is extended along the y-axis. If it is extended one step, the coordinate of the y-axis is decreased by 1, that is, the coordinate of p1' will become (-5,-7,4), and the coordinate of p2' will become p2'=(4,-7,4). Similarly, a section is made along the z-axis and the y-axis based on the minimum y-axis tangent of the sacrum. Then a section is made along the z-axis based on the leftmost and rightmost tangents of the sacrum.

[0056] The intersecting rectangular area formed by the six sections is the nerve-free area of ​​the sacrum. Then, the coordinates of the center points of the leftmost section and the rightmost section in the rectangular area are calculated respectively, and a straight line is drawn connecting the center points of the two sections. This straight line is the nerve-free channel of the sacrum.

[0057] The above embodiments and accompanying drawings are intended only to illustrate the technical solutions of the present invention and are not intended to limit the present invention. The present invention has been described in detail with reference to preferred embodiments. Those skilled in the art should understand that any changes, modifications, additions, or substitutions made by those skilled in the art within the spirit and scope of the present invention do not depart from the spirit of the present invention and are intended to fall within the scope of the claims. Other related technical structures not fully disclosed in the present invention constitute prior art in the art.

Claims

1. A method for locating sacral nerve-free channels based on CPD point cloud registration technology, characterized in that: The following steps are involved: Step 1, data acquisition: obtaining a 3D CT scan image of the sacrum; Step 2, 3D image point cloud processing: The acquired CT image is processed into point cloud data using CPD point cloud registration technology, and the pixel positions of the CT image are mapped to 3D coordinates; Step 3, marking the first sacral foramen: mark the first sacral foramen of the sacrum, denoted as hole_target; Step 4, point cloud registration: The sacral point cloud data with the marked sacral foramen is used as the registration target, denoted as target, and the sacral point cloud data without the marked sacral foramen is used as source. The CPD point cloud registration algorithm is used for point cloud registration to obtain the mapping relationship between the configured source and target, denoted as f(°). Then, the data of the first marked sacral foramen is input into the mapping relationship f(°) to obtain hole_source = f(hole_target). The coordinates of hole_source are the position of the first sacral foramen of the unmarked sacrum. Step 5, calculate the nerve-free channel: solve the maximum value of the z-axis coordinate of the first sacral foramen, that is, max(hole_source(z)). After obtaining the coordinates of the upper edge of the first sacral foramen, make sections along the z-axis and y-axis based on the upper edge of the first sacral foramen; then make sections along the z-axis and y-axis based on the tangent line with the minimum y-axis value of the sacrum; then make sections along the z-axis based on the leftmost and rightmost tangent lines of the sacrum. The intersecting rectangular area formed by the six sections is the nerve-free area of ​​the sacrum; then calculate the coordinates of the center points of the leftmost and rightmost sections in the rectangular area, and draw a straight line connecting the center points of the two sections. This straight line is the nerve-free channel of the sacrum.

2. The method for locating sacral nerve-free channels based on CPD point cloud registration technology according to claim 1, characterized in that: The mapping method is: 1) Calculate the total number of CT image slices as z_max, and calculate the width of the horizontal axis of each slice in the CT image as w_max and the length of the vertical axis as h_max; 2) Convert the pixel position of each slice in the CT image into coordinates in three-dimensional space. The original pixel coordinates are recorded as p_o=(x_i, y_j, z_k), where z_k represents the kth slice. The conversion formula is: x_new = x_i-(1+w_max) / 2 y_new = y_j-(1+h_max) / 2 z_new = z_k-(1+z_max) / 2 The converted coordinates p_new=(x_new, y_new, z_new) are the three-dimensional coordinates of each image in the original CT image; 3) Create a new three-dimensional matrix point=(w_max,h_max,z_max), assign the point of p_new coordinate to 1, and assign the other points to 0, then you will get the discretized point cloud coordinates, that is, the point cloud data of the sacrum.

3. The method for locating a sacral nerve-free channel based on CPD point cloud registration technology according to claim 2, characterized in that: The three-dimensional matrix point=(w_max, h_max, z_max) is established in the python programming language environment.

4. The method for locating sacral nerve-free channels based on CPD point cloud registration technology according to claim 1, characterized in that: The data acquisition described in step 1 uses Python to read the three-dimensional sacrum model. The core command for reading the three-dimensional sacrum is ds = pydicom.dcmread("digu.dcm").