A medical image cardiothoracic ratio measurement method and system based on artificial intelligence

Abstract

The application discloses a kind of medical image cardiothoracic ratio measurement method and system based on artificial intelligence, method includes: obtaining the first picture that user stands in front of radiographic mainboard and is photographed by top camera, according to first picture, obtain the initial azimuth of user relative to radiographic mainboard;Judge whether the absolute value of initial azimuth exceeds first preset angle;Start chest radiography image device, collect the first chest medical image of user standing in front of radiographic mainboard;According to initial azimuth adjustment binary threshold, according to binary threshold to first chest medical image is carried out binary processing and obtains binary image;Gradient operator is used to extract the contour line of lung in binary image;The cardiothoracic ratio obtained by calculating heart transverse diameter divided by thoracic transverse diameter.The binary threshold of lung edge is adjusted in the application to improve the position accuracy of lung contour line, and then accurate cardiothoracic ratio is obtained.

Description

Technical Field

[0001] This invention relates to the field of medical imaging technology, and in particular to a method and system for measuring the cardiothoracic ratio in medical imaging based on artificial intelligence. Background Technology

[0002] The cardiothoracic ratio is the ratio of the transverse diameter of the heart to the transverse diameter of the thorax on an X-ray. The transverse diameter of the heart is the sum of the maximum distances from the left and right borders of the heart to the midline, while the transverse diameter of the thorax is the internal diameter of the thorax passing through the top of the right diaphragm. Cardiothoracic ratio measurement is one of the most commonly used methods for measuring the heart and is a frequently used indicator for assessing cardiac enlargement. Based on its value, the heart is classified as mild, moderate, or severe enlargement.

[0003] In existing medical imaging cardiothoracic ratio measurement systems, errors occur in the measurement method because the user does not fully contact the radiographic mainboard during a chest X-ray examination. Summary of the Invention

[0004] The applicant's research revealed that in existing technologies, users do not fully contact the radiographic panel during chest X-ray examinations, resulting in an angle between the user and the panel. Small angular deviations are difficult for physicians to detect. If there is a small angle between the user and the radiographic panel during imaging, the image of the lateral edge of the chest wall in the captured medical image will obscure the image of the lung edge. In other words, the lateral chest wall image range is larger than the image range captured at zero angle, while the lung image range is smaller than the image range captured at zero angle. Consequently, the dark area of ​​the lung in the captured medical image is reduced, while the change in the heart area is relatively small. This alters the result of measuring the cardiothoracic ratio based on the medical image, leading to inaccurate measurements.

[0005] In view of the aforementioned deficiencies of the prior art, the technical problem to be solved by the present invention is to provide a method and system for measuring the cardiothoracic ratio in medical images based on artificial intelligence, which aims to accurately measure the cardiothoracic ratio by adjusting the binarization threshold.

[0006] To achieve the above objectives, the first aspect of this invention discloses a method for measuring the cardiothoracic ratio in medical images based on artificial intelligence, the method comprising:

[0007] Step S1: Obtain a first image of the user standing in front of the radiating motherboard, captured by a camera on top. Based on the first image, obtain the initial azimuth angle of the user relative to the radiating motherboard. The initial azimuth angle is zero when the user is standing perpendicular to the radiating motherboard. The range of the initial azimuth angle is [-180°, 180°), where the initial azimuth angle is negative when the user rotates clockwise and positive otherwise.

[0008] Step S2: Determine whether the absolute value of the initial azimuth angle exceeds the first preset angle; if the initial azimuth angle exceeds the first preset angle, issue an adjustment reminder and return to step S1, otherwise proceed to step S3; the first preset angle is 5°;

[0009] Step S3: Turn on the chest X-ray imaging device and acquire the first chest medical image of the user standing in front of the radiographic panel;

[0010] Step S4: Adjust the binarization threshold G according to the initial azimuth angle. TH According to the binarization threshold G TH The first chest medical image is binarized to obtain a binarized image; the binarization threshold G TH Satisfying the first conversion curve G TH = g(θ), where the first conversion curve is obtained by fitting actual experiments, and θ is the initial azimuth angle;

[0011] Step S5: Use the gradient operator to extract the contour line of the lung in the binarized image and obtain the midline of the lung; measure the maximum distance of the right heart border from the midline and the maximum distance of the left heart border from the midline, and measure the transverse diameter of the thoracic cavity;

[0012] Step S6: Calculate the heart transverse diameter by the sum of the maximum distance from the right heart border to the midline and the maximum distance from the left heart border to the midline, and divide the heart transverse diameter by the thoracic transverse diameter to obtain the cardiothoracic ratio.

[0013] Optionally, the sub-steps of step S1 include: a lateral vector confirmation step;

[0014] In the first image, identify the horizontal line where the user's coronal plane is located, and determine the first horizontal axis vector as the direction from the user's left to the right.

[0015] The second lateral vector is defined as the parallel pointing direction to the right from the left end of the radial motherboard, and the angle between the second lateral vector and the first lateral vector is the initial azimuth angle.

[0016] Optionally, after step S4, the method further includes: correcting the binarization threshold G. TH step;

[0017] The correction binarization threshold G TH The steps include:

[0018] Obtain the first three-dimensional distance L between the measured intersection point of the fitted extensions of the left and right clavicles and the axis where the user's vertebrae are located;

[0019] Based on the first chest medical image, the first position of the intersection of the extended fitting points of the left and right clavicles is obtained;

[0020] Based on the first chest medical image, the location of the user's vertebrae along the first axis is obtained;

[0021] Based on the first axis and the first position, obtain the first planar distance Δx between the intersection point of the left and right clavicle fitting extensions and the first axis;

[0022] The corrected azimuth angle is obtained by correcting the first 3D distance L, the first planar distance Δx, and the initial azimuth angle θ. The corrected azimuth angle

[0023] According to the corrected azimuth angle The binarization threshold G is obtained from the first conversion curve. TH .

[0024] Optionally, after step S4, the method further includes:

[0025] Step A: Input the first chest medical image, and use the binarization threshold to set the pixel value of all pixels in the first chest medical image whose pixel value is less than or equal to the binarization threshold to 0, and set the pixel value of all pixels in the first chest medical image whose pixel value is greater than the binarization threshold to 255, so as to obtain a binarized image;

[0026] Step B: Locate the non-extractable area outside the chest region, and extract the chest region; remove the pixels of the non-extractable area from the binarized image to obtain the binarized image with the extracted area pixels.

[0027] Optionally, the sub-steps of step S5 include:

[0028] Step S501: Extract the lung contour line, select the outer contour line of the left lung and the outer contour line of the right lung, and calculate the central axis line based on the outer edges of the left and right lungs;

[0029] Step S502: Find the highest point at the bottom of the right lung field contour line, draw a first horizontal line through this point, and calculate the distance between the intersection of the first horizontal line and the left edge of the left lung field contour and the right edge of the right lung field contour to obtain the transverse diameter of the thoracic cavity.

[0030] The second aspect of this invention discloses a medical imaging cardiothoracic ratio measurement system based on artificial intelligence. The system includes: a camera, an initial azimuth angle acquisition module, an angle judgment module, an acquisition module, a binarization threshold adjustment module, an extraction measurement module, and a cardiothoracic ratio calculation module.

[0031] The camera is positioned above the user being tested.

[0032] The initial azimuth angle acquisition module is used to acquire a first image taken by a top camera while the user is standing in front of the radiating motherboard, and to obtain the initial azimuth angle of the user relative to the radiating motherboard based on the first image; wherein, when the user is standing with his / her facing perpendicular to the radiating motherboard, the initial azimuth angle is zero; the range of the initial azimuth angle is [-180, 180), wherein the initial azimuth angle is negative when the user is rotating clockwise, and positive otherwise;

[0033] The angle determination module is used to determine whether the absolute value of the initial azimuth angle exceeds the first preset angle; if the initial azimuth angle exceeds the first preset angle, an adjustment reminder is issued and the process returns to step S1, otherwise the process proceeds to step S3; the first preset angle is 5°.

[0034] The acquisition module is used to turn on the chest X-ray imaging device and acquire the first chest medical image of the user standing in front of the radiographic motherboard.

[0035] The binarization threshold adjustment module is used to adjust the binarization threshold G according to the initial azimuth angle. TH According to the binarization threshold G TH The first chest medical image is binarized to obtain a binarized image; the binarization threshold G TH Satisfying the first conversion curve G TH = g(θ), where the first conversion curve is obtained by fitting actual experiments, and θ is the initial azimuth angle;

[0036] The extraction and measurement module is used to extract the contour line of the lung in the binary image using a gradient operator, obtain the midline of the lung; measure the maximum distance between the right heart border and the midline of the contour line, the maximum distance between the left heart border and the midline of the lung, and measure the transverse diameter of the thoracic cavity;

[0037] The cardiothoracic ratio calculation module is used to calculate the heart transverse diameter by the sum of the maximum distance of the right heart border from the midline and the maximum distance of the left heart border from the midline, and to obtain the cardiothoracic ratio by dividing the heart transverse diameter by the thoracic transverse diameter.

[0038] Optionally, the initial azimuth angle acquisition module includes: a lateral vector confirmation unit;

[0039] The horizontal vector confirmation unit is used to identify the horizontal line where the user's coronal plane is located in the first image, and determine the user's left side pointing to the right side as the first horizontal axis vector;

[0040] The second lateral vector is defined as the parallel pointing direction to the right from the left end of the radial motherboard, and the angle between the second lateral vector and the first lateral vector is the initial azimuth angle.

[0041] Optionally, the system further includes: a correction binarization threshold G. TH The module, the correction binarization threshold G TH The module operates after the binarization threshold adjustment module.

[0042] The correction binarization threshold G TH The module is configured as follows:

[0043] Obtain the first three-dimensional distance L between the measured intersection point of the fitted extensions of the left and right clavicles and the axis where the user's vertebrae are located;

[0044] Based on the first chest medical image, the first position of the intersection of the extended fitting points of the left and right clavicles is obtained;

[0045] Based on the first chest medical image, the location of the user's vertebrae along the first axis is obtained;

[0046] Based on the first axis and the first position, obtain the first planar distance Δx between the intersection point of the left and right clavicle fitting extensions and the first axis;

[0047] The corrected azimuth angle is obtained by correcting the first 3D distance L, the first planar distance Δx, and the initial azimuth angle θ. The corrected azimuth angle

[0048] According to the corrected azimuth angle The binarization threshold G is obtained from the first conversion curve. TH .

[0049] Optionally, the system further includes an image binarization unit and an image extraction unit, wherein the image binarization unit and the image extraction unit operate before the extraction measurement module operates;

[0050] The image binarization unit is used to input the first chest medical image, and use the binarization threshold to set the pixel value of all pixels in the first chest medical image whose pixel value is less than or equal to the binarization threshold to 0, and at the same time set the pixel value of all pixels in the first chest medical image whose pixel value is greater than the binarization threshold to 255, so as to obtain a binarized image.

[0051] The image extraction unit: locates the non-extraction area outside the chest region, and extracts the chest region; removes the pixels of the non-extraction area in the binarized image to obtain the binarized image with the pixels of the extraction area.

[0052] Optionally, the extraction measurement module includes:

[0053] Midline calculation unit: Extract the lung contour line, select the outer contour line of the left lung and the outer contour line of the right lung, and calculate the midline line based on the outer edges of the left and right lungs;

[0054] Thoracic transverse diameter calculation unit: Find the highest point at the bottom of the right lung field contour line, draw the first horizontal line through this point, and calculate the distance between the intersection of the first horizontal line and the left edge of the left lung field contour and the right edge of the right lung field contour to obtain the thoracic transverse diameter.

[0055] The beneficial effects of this invention are as follows: 1. This invention obtains the initial azimuth angle of the user relative to the radiation motherboard by acquiring a first image taken by a camera on top of the user standing in front of the radiation motherboard; determines whether the absolute value of the initial azimuth angle exceeds a first preset angle; if the initial azimuth angle exceeds the first preset angle, an adjustment reminder is issued and the process returns to the previous step; otherwise, the process proceeds to the next step; the chest X-ray imaging device is turned on to acquire the first chest medical image of the user standing in front of the radiation motherboard; and the binarization threshold G is adjusted according to the initial azimuth angle. TH According to the binarization threshold G TH The first chest medical image is binarized to obtain a binarized image. This invention adjusts the binarization threshold of the lung edge image based on the initial azimuth angle, solving the problem of the chest wall lateral edge image obscuring the lung edge image in medical images taken at smaller azimuth angles. Specifically, the chest wall lateral image range is larger than the image range taken at zero angle, while the lung image range is smaller than the image range taken at zero angle, resulting in a reduced black area edge of the lung in the captured medical image. This corrects the lung range and improves the accuracy of the lung region. 2. This invention obtains the first planar distance between the intersection point of the left and right clavicle fitting extensions and the first axis line on the first medical image by measuring the first three-dimensional distance between the intersection point of the left and right clavicle fitting extensions and the axis line of the user's vertebrae. This corrects the initial azimuth angle to obtain the corrected azimuth angle. Improving the accuracy of the azimuth angle, and thus adjusting the binarization threshold, yields a more accurate cardiothoracic ratio. In summary, this invention achieves the adjustment of the binarization threshold according to the initial azimuth angle, solving the problem that when the medical images of the user are captured at a small azimuth angle, the edge range of the chest wall is obscured by the edge range of the lung, i.e. the range of the chest wall side image is larger than the range of the image captured at zero angle and the range of the lung image is smaller than the range of the image captured at zero angle. By adjusting the binarization threshold of the lung edge, the positional accuracy of the lung contour line is improved, thereby obtaining an accurate cardiothoracic ratio. Attached Figure Description

[0056] Figure 1 This is a flowchart illustrating a method for measuring the cardiothoracic ratio in medical imaging based on artificial intelligence, provided in a specific embodiment of the present invention.

[0057] Figure 2This is a schematic diagram of the structure of a medical imaging cardiothoracic ratio measurement system based on artificial intelligence, provided in a specific embodiment of the present invention;

[0058] Figure 3 This is a schematic diagram of the correction binarization threshold structure;

[0059] Figure 4 This is a schematic diagram of the transverse diameter of the heart and the transverse diameter of the thoracic cavity;

[0060] Figure 5 This is a schematic diagram of the distance to the first plane viewed from above. Detailed Implementation

[0061] This invention discloses a method and system for measuring the cardiothoracic ratio in medical images based on artificial intelligence. Those skilled in the art can refer to the content of this document and appropriately modify the technical details to achieve the desired implementation. It should be particularly noted that all similar substitutions and modifications are obvious to those skilled in the art and are considered to be included in this invention. The methods and applications of this invention have been described through preferred embodiments. Those skilled in the art can clearly modify or appropriately change and combine the methods and applications described herein without departing from the content, spirit, and scope of this invention to implement and apply the technology of this invention.

[0062] The applicant's research revealed that in existing technologies, users do not fully contact the radiographic panel during chest X-ray examinations, resulting in an angle between the user and the panel. Small angular deviations are difficult for physicians to detect. If there is a small angle between the user and the radiographic panel during imaging, the image of the lateral edge of the chest wall in the captured medical image will obscure the image of the lung edge. In other words, the lateral chest wall image range is larger than the image range captured at zero angle, while the lung image range is smaller than the image range captured at zero angle. Consequently, the dark area of ​​the lung in the captured medical image is reduced, while the change in the heart area is relatively small. This alters the result of measuring the cardiothoracic ratio based on the medical image, leading to inaccurate measurements.

[0063] Therefore, embodiments of the present invention provide a medical imaging method for measuring the cardiothoracic ratio based on artificial intelligence, such as... Figure 1 , Figure 3-5 As shown, the method includes:

[0064] Step S1: Obtain a first image of the user standing in front of the radiating motherboard and captured by the top camera. Based on the first image, obtain the initial azimuth angle of the user relative to the radiating motherboard. The initial azimuth angle is zero when the user is standing perpendicular to the radiating motherboard. The range of the initial azimuth angle is [-180, 180), where the initial azimuth angle is positive when the user rotates clockwise and negative otherwise.

[0065] Step S2: Determine whether the absolute value of the initial azimuth exceeds the first preset angle; if the initial azimuth exceeds the first preset angle, issue an adjustment reminder and return to step S1, otherwise proceed to step S3; the first preset angle is 5°.

[0066] Step S3: Turn on the chest X-ray imaging equipment and acquire the first chest medical image of the user standing in front of the radiographic panel;

[0067] Step S4: Adjust the binarization threshold G according to the initial azimuth angle. TH According to the binarization threshold G TH The first chest medical image is binarized to obtain a binarized image; the binarization threshold G TH Satisfying the first conversion curve G TH = g(θ), the first conversion curve is obtained by fitting actual experiments, and θ is the initial azimuth angle;

[0068] Step S5: Use the gradient operator to extract the contour line of the lung in the binarized image and obtain the midline T of the lung; measure the maximum distance L1 from the right heart edge of the contour line to the midline and the maximum distance L2 from the left heart edge to the midline, and measure the transverse diameter L3 of the thoracic cavity.

[0069] Step S6: Calculate the heart transverse diameter by the sum of the maximum distance L1 from the right heart border to the midline and the maximum distance L2 from the left heart border to the midline. Divide the heart transverse diameter by the thoracic transverse diameter L3 to obtain the cardiothoracic ratio.

[0070] Optionally, the sub-steps of step S1 include: a lateral vector confirmation step;

[0071] Identify the horizontal line where the user's coronal plane is located in the first image, and determine the first horizontal axis vector as the user's left side pointing to the right side;

[0072] The second lateral vector is the parallel direction pointing to the right from the left end of the radial motherboard. The angle between the second lateral vector and the first lateral vector is the initial azimuth angle θ.

[0073] Optionally, after step S4, the method further includes: correcting the binarization threshold G. TH step;

[0074] Correction binarization threshold G TH The steps include:

[0075] Obtain the first three-dimensional distance L between the measured intersection point of the fitted extension of the left and right clavicles and the axis where the user's vertebrae are located;

[0076] Based on the first chest medical image, the first location of the intersection point P of the fitted extension of the left and right clavicles is obtained;

[0077] Based on the first chest medical image, the location of the user's vertebrae along the first axis T1 is determined;

[0078] Based on the first axis and the first position, the first planar distance Δx between the intersection point P of the fitted extension of the left and right clavicles and the first axis T1 is obtained;

[0079] The corrected azimuth angle is obtained by correcting the first stereo distance L, the first planar distance Δx, and the initial azimuth angle θ. Corrected azimuth

[0080] According to the corrected azimuth angle And the binarization threshold G is obtained from the first conversion curve. TH .

[0081] Optionally, after step S4, the method further includes:

[0082] Step A: Input the first chest medical image, and use a binarization threshold to set the pixel value of all pixels in the first chest medical image whose pixel value is less than or equal to the binarization threshold to 0, and set the pixel value of all pixels in the first chest medical image whose pixel value is greater than the binarization threshold to 255, so as to obtain a binarized image;

[0083] Step B: Locate the non-extractable area outside the chest region, and extract the chest region; remove the pixels of the non-extractable area from the binarized image to obtain a binarized image with the pixels of the extracted area.

[0084] Optionally, the sub-steps of step S5 include:

[0085] Step S501: Extract the lung contour lines, select the outer contour lines of the left lung and the right lung, and calculate the midline based on the outer edges of the left and right lungs;

[0086] Step S502: Find the highest point P1 at the bottom of the right lung field contour line, draw the first horizontal line through this point, and calculate the distance between the intersection points P2 and P3 of the first horizontal line and the left edge of the left lung field contour and the right edge of the right lung field contour to obtain the transverse diameter L3 of the thoracic cavity.

[0087] In summary, the embodiments of the present invention can adjust the binarization threshold according to the initial azimuth angle, which solves the problem that when the medical image of the user is captured at a small azimuth angle, the edge range of the chest wall is obscured by the edge range of the lung. That is, the image range of the chest wall side is larger than the image range captured at zero angle and the image range of the lung is smaller than the image range captured at zero angle. Adjusting the binarization threshold of the lung edge improves the positional accuracy of the lung contour line, thereby obtaining an accurate cardiothoracic ratio.

[0088] Based on the above-mentioned AI-based medical image cardiothoracic ratio measurement method, this embodiment of the invention also provides an AI-based medical image cardiothoracic ratio measurement system, the system including: camera 101, initial azimuth angle acquisition module 102, angle judgment module 103, acquisition module 104, binarization threshold adjustment module 105, extraction measurement module 106, and cardiothoracic ratio calculation module 107.

[0089] Camera 101 is positioned above the user being tested;

[0090] The initial azimuth angle acquisition module 102 is used to acquire a first image taken by the top camera while the user is standing in front of the radiating motherboard, and to obtain the initial azimuth angle of the user relative to the radiating motherboard based on the first image; wherein, when the user is standing with his / her facing perpendicular to the radiating motherboard, the initial azimuth angle is zero; the range of the initial azimuth angle is [-180, 180), wherein the initial azimuth angle is positive when the user is rotating clockwise, and negative otherwise;

[0091] Angle determination module 103 is used to determine whether the absolute value of the initial azimuth angle exceeds the first preset angle; if the initial azimuth angle exceeds the first preset angle, an adjustment reminder is issued and the process returns to step S1, otherwise the process proceeds to step S3; the first preset angle is 5°.

[0092] Acquisition module 104 is used to turn on the chest X-ray imaging equipment and acquire the first chest medical image of the user standing in front of the radiographic mainboard.

[0093] Binarization threshold adjustment module 105 is used to adjust the binarization threshold G according to the initial azimuth angle. TH According to the binarization threshold G TH The first chest medical image is binarized to obtain a binarized image; the binarization threshold G TH Satisfying the first conversion curve G TH = g(θ), the first conversion curve is obtained by fitting actual experiments, and θ is the initial azimuth angle;

[0094] The extraction and measurement module 106 is used to extract the contour line of the lung in the binary image using a gradient operator, obtain the midline of the lung, measure the maximum distance between the right heart border and the midline of the contour line, the maximum distance between the left heart border and the midline of the contour line, and measure the transverse diameter of the thoracic cavity.

[0095] The cardiothoracic ratio calculation module 107 is used to calculate the heart transverse diameter by the sum of the maximum distance between the right heart border and the midline and the maximum distance between the left heart border and the midline. The cardiothoracic ratio is obtained by dividing the heart transverse diameter by the thoracic diameter.

[0096] Optionally, the initial azimuth angle acquisition module 102 includes: a lateral vector confirmation module;

[0097] The horizontal vector confirmation module is used to identify the horizontal line where the user's coronal plane is located in the first image, and determine the first horizontal axis vector as the user's left side pointing to the right side;

[0098] The second lateral vector is defined as the parallel pointing direction to the right from the left end of the radial motherboard. The angle between the second lateral vector and the first lateral vector is the initial azimuth angle.

[0099] Optionally, the system also includes: a correction binarization threshold G. TH Module, correction binarization threshold G TH This module operates after the binarization threshold adjustment module 105.

[0100] Correction binarization threshold G TH The module is configured as follows:

[0101] Obtain the first three-dimensional distance L between the measured intersection point of the fitted extension of the left and right clavicles and the axis where the user's vertebrae are located;

[0102] Based on the first chest medical image, the first location of the intersection point P of the fitted extension of the left and right clavicles is obtained;

[0103] Based on the first chest medical image, the location of the user's vertebrae along the first axis T1 is determined;

[0104] Based on the first axis and the first position, the first planar distance Δx between the intersection point P of the fitted extension of the left and right clavicles and the first axis T1 is obtained;

[0105] The corrected azimuth angle is obtained by correcting the first stereo distance L, the first planar distance Δx, and the initial azimuth angle θ. Corrected azimuth

[0106] According to the corrected azimuth angle And the binarization threshold G is obtained from the first conversion curve. TH .

[0107] Optionally, the system also includes an image binarization unit and an image extraction unit, which operate before the extraction measurement module 106 operates.

[0108] Image Binary Unit: Used to input the first chest medical image. Using a binarization threshold, all pixels in the first chest medical image with pixel values ​​less than or equal to the binarization threshold are set to a pixel value of 0, while all pixels in the first chest medical image with pixel values ​​greater than the binarization threshold are set to a pixel value of 255, thus obtaining a binarized image.

[0109] Image extraction unit: The non-extraction area is located outside the chest region, and the extraction area is the chest region; the pixels of the non-extraction area in the binarized image are eliminated to obtain a binarized image with the pixels of the extraction area.

[0110] Optionally, the extraction measurement module 106 includes:

[0111] Midline calculation unit: Extract the lung contour line, select the outer contour line of the left lung and the outer contour line of the right lung, and calculate the midline line based on the outer edges of the left and right lungs;

[0112] Thoracic transverse diameter calculation unit: Find the highest point P1 at the bottom of the right lung field contour line, draw the first horizontal line through this point, calculate the distance between the intersection points P2 and P3 of the first horizontal line and the left edge of the left lung field contour and the right edge of the right lung field contour, and obtain the thoracic transverse diameter L3.

[0113] In conjunction with the above embodiments, this embodiment of the invention obtains the initial azimuth angle of the user relative to the radiation motherboard by acquiring a first image taken by the top camera 101 while the user is standing in front of the radiation motherboard; determines whether the absolute value of the initial azimuth angle exceeds a first preset angle; if the initial azimuth angle exceeds the first preset angle, an adjustment reminder is issued and the process returns to the previous step; otherwise, the process proceeds to the next step; the chest X-ray imaging device is turned on to acquire the first chest medical image of the user standing in front of the radiation motherboard; and the binarization threshold G is adjusted according to the initial azimuth angle. TH According to the binarization threshold G TH The first chest medical image is binarized to obtain a binarized image. This invention adjusts the binarization threshold of the lung edge image based on the initial azimuth angle, solving the problem of the chest wall lateral edge image obscuring the lung edge image in medical images taken at a smaller azimuth angle. Specifically, the chest wall lateral image range is larger than the image range taken at zero angle, while the lung image range is smaller than the image range taken at zero angle, resulting in a reduced black area edge of the lung in the captured medical image. This corrects the lung range and improves the accuracy of the lung region. In this embodiment, the first planar distance between the intersection point of the left and right clavicle fitting extensions and the first axis of the user's vertebrae is obtained on the first medical image by measuring the first three-dimensional distance between the intersection point of the left and right clavicle fitting extensions and the first axis. The corrected azimuth angle is then obtained by correcting the initial azimuth angle. This improves the accuracy of the azimuth angle, thereby adjusting the binarization threshold to obtain a precise cardiothoracic ratio. In summary, this invention achieves the adjustment of the binarization threshold according to the initial azimuth angle, solving the problem that when the medical images of the user are captured at a small azimuth angle, the edge range of the chest wall is obscured by the edge range of the lung, i.e. the range of the chest wall side image is larger than the range of the image captured at zero angle and the range of the lung image is smaller than the range of the image captured at zero angle. By adjusting the binarization threshold of the lung edge, the positional accuracy of the lung contour line is improved, thereby obtaining an accurate cardiothoracic ratio.

[0114] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0115] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0116] The above are merely preferred embodiments of the present invention and are not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention are included within the scope of protection of the present invention.

Claims

1. An artificial intelligence-based medical image cardiothoracic ratio measurement method, characterized by, The method comprises: Step S1, obtaining a first picture of a user standing in front of a radiation main plate taken by a camera on the top, and obtaining an initial azimuth angle of the user relative to the radiation main plate according to the first picture; wherein the initial azimuth angle is zero when the user stands perpendicular to the radiation main plate; the initial azimuth angle ranges from -180 to 180, wherein the initial azimuth angle is negative when the user rotates clockwise, and vice versa; Step S2, judging whether the absolute value of the initial azimuth angle exceeds a first preset angle; if the initial azimuth angle exceeds the first preset angle, an adjustment reminder is sent and the step S1 is returned, otherwise the step S3 is entered; the first preset angle is 5°; Step S3, starting a chest radiography image device, and collecting a first chest medical image of the user standing in front of the radiation main plate; Step S4, adjusting a binarization threshold according to the initial azimuth angle , according to the binarization threshold binarizing the first chest medical image to obtain a binarized image; the binarization threshold satisfies a first conversion curve , the first conversion curve is obtained by actual experiment fitting, is the initial azimuth angle; Step S5, extracting a contour line of a lung in the binary image by using a gradient operator, and obtaining a central axis of the lung; measuring a maximum distance of a right heart border to the central axis, a maximum distance of a left heart border to the central axis, and a thoracic transverse diameter; Step S6, calculating a sum of the maximum distance of the right heart border to the central axis and the maximum distance of the left heart border to the central axis to obtain a cardiac transverse diameter, and dividing the cardiac transverse diameter by the thoracic transverse diameter to obtain a cardiothoracic ratio; After step S4 also comprises: correcting the binarization threshold Step; The correction binarization threshold The steps include: Obtaining a first three-dimensional distance L of a fitting extension intersection point of left and right clavicles to an axis of a vertebra of the user; Obtaining a first position of the fitting extension intersection point of the left and right clavicles according to the first chest medical image; Obtaining a first axis of the vertebra of the user according to the first chest medical image; obtaining a first plane distance between the left and right clavicle fitting extension intersection point and the first axis according to the first axis and the first position ; According to the first stereoscopic distance L, the first plane distance , the initial azimuth angle , the correction obtains a modified azimuth angle ; the modified azimuth angle ; According to the modified azimuth angle and the first conversion curve to obtain the binary threshold . 2.The AI-based medical image cardiothoracic ratio measurement method of claim 1, wherein, The sub-step of step S1 comprises a transverse vector confirmation step; Identifying a horizontal line where a coronal plane of the user is located in the first picture, and determining a first transverse vector pointing from the left side of the user to the right side as a first transverse vector; A second transverse vector is determined according to a parallel pointing direction of a left end point of the radiation main plate to the right, and an included angle between the second transverse vector and the first transverse vector is the initial azimuth angle. 3.The AI-based medical image cardiothoracic ratio measurement method of claim 1, wherein, After step S4, the method further comprises: Step A: inputting the first chest medical image, setting a pixel value of all pixel points with a pixel value less than or equal to a binary threshold value in the first chest medical image to 0, and setting a pixel value of all pixel points with a pixel value greater than the binary threshold value in the first chest medical image to 255, to obtain a binary image; Step B: positioning a non-extraction area outside a chest area, and an extraction area as the chest area; eliminating pixels in the non-extraction area in the binary image to obtain the binary image with extraction area pixels. 4.The AI-based medical image cardiothoracic ratio measurement method of claim 1, wherein, The sub-step of step S5 comprises: Step S501: extracting a lung contour line, selecting an outer contour line of a left lung and an outer contour line of a right lung, and calculating the central axis according to the outer edges of the left and right lungs; Step S502: finding the highest point at the bottom of the right lung field contour line, drawing a first horizontal line through the point, calculating a distance between the intersection points of the first horizontal line and the left edge of the left lung field contour and the right edge of the right lung field contour, and obtaining a thoracic transverse diameter.

5. An artificial intelligence based medical image cardiothoracic ratio measurement system characterized in that, The system comprises a camera, an initial azimuth angle acquisition module, an angle judgment module, a collection module, a binary threshold adjustment module, an extraction measurement module, a cardiothoracic ratio calculation module; The camera is arranged on the top of the user to be measured; The initial azimuth angle acquisition module is configured to acquire a first picture of the user standing in front of a radiographic panel and photographed by the camera on the top, and obtain an initial azimuth angle of the user relative to the radiographic panel according to the first picture; when the user stands perpendicular to the radiographic panel, the initial azimuth angle is zero; the range of the initial azimuth angle is [-180, 180], wherein the initial azimuth angle is negative when the user rotates clockwise, and vice versa; The angle judgment module is configured to judge whether the absolute value of the initial azimuth angle exceeds a first preset angle; if the initial azimuth angle exceeds the first preset angle, an adjustment reminder is sent and the step S1 is returned, otherwise, the step S3 is entered; the first preset angle is 5°; The collection module is configured to start a chest radiography image device, and collect a first chest medical image of the user standing in front of the radiographic panel; The binarization threshold adjustment module is configured to adjust a binarization threshold according to the initial azimuth angle , according to the binarization threshold , the first chest medical image is subjected to binarization processing and a binarized image is obtained; the binarization threshold satisfies a first conversion curve The first conversion curve is obtained by actual experiment fitting, is the initial azimuth angle; The extraction measurement module is configured to extract a contour line of a lung in a binary image by using a gradient operator, and obtain a central axis of the lung; measure a maximum distance of a right heart border from the central axis, a maximum distance of a left heart border from the central axis, and a thoracic transverse diameter; The cardiothoracic ratio calculation module is configured to calculate a cardiothoracic ratio by dividing a heart transverse diameter by the thoracic transverse diameter, wherein the heart transverse diameter is a sum of the maximum distance of the right heart border from the central axis and the maximum distance of the left heart border from the central axis; The system also includes a correction binarization threshold module, the correction binarization threshold module works after the binarization threshold adjustment module works; The correction binarization threshold The module is configured to: A first three-dimensional distance L of a fitting extension intersection point of left and right clavicles on an axis of a vertebra of the user is acquired; A first position of the fitting extension intersection point of the left and right clavicles is obtained according to the first chest medical image; A first axis of the vertebra of the user is obtained according to the first chest medical image; obtaining a first plane distance between the left and right clavicle fitting extension intersection point and the first axis according to the first axis and the first position ; According to the first stereoscopic distance L, the first plane distance , the initial azimuth angle , the correction obtains a modified azimuth angle ; the modified azimuth angle ; According to the modified azimuth angle and the first conversion curve to obtain the binary threshold .

6. The artificial intelligence based medical image cardiothoracic ratio measurement system as claimed in claim 5, wherein, The initial azimuth angle acquisition module comprises a transverse vector confirmation unit; The transverse vector confirmation unit is configured to identify a horizontal line where a coronal plane of the user is located in the first picture, and determine a first transverse vector pointing from the left side of the user to the right side as a first transverse vector; A second transverse vector is configured to be parallel to a direction pointing to the right of a left end point of the radiographic panel, and an included angle between the second transverse vector and the first transverse vector is the initial azimuth angle.

7. The artificial intelligence based medical image cardiothoracic ratio measurement system as claimed in claim 5, wherein, The system further comprises an image binarization unit and an extraction image unit, and the image binarization unit and the extraction image unit work before the extraction measurement module works; The image binarization unit is configured to input the first chest medical image, set a pixel value of all pixel points in the first chest medical image less than or equal to a binary threshold value to 0, and set a pixel value of all pixel points in the first chest medical image greater than the binary threshold value to 255, to obtain a binary image; The extraction image unit positions a non-extraction area outside a chest area, the extraction area being the chest area; and eliminates pixels of the non-extraction area in the binary image to obtain the binary image with pixels of the extraction area.

8. The artificial intelligence-based medical image cardiothoracic ratio measurement system according to claim 5, characterized by, The extraction measurement module comprises: The central axis calculation unit extracts a lung contour line, selects an outer contour line of a left lung and an outer contour line of a right lung, and calculates the central axis according to outer edges of the left and right lungs. The thoracic transverse diameter calculation unit finds the highest point at the bottom of the right lung field contour line, draws a first horizontal line through the point, calculates the distance between the intersection of the first horizontal line and the left edge of the left lung field contour and the right edge of the right lung field contour, and obtains the thoracic transverse diameter.

Images