An elbow joint rotation center axis identification method and system, an electronic device, and a medium

Abstract

The application discloses a method and system for identifying an elbow joint rotation center axis, electronic equipment and a medium, and relates to the field of elbow joint rotation center axis identification. The method comprises the following steps: acquiring a distal humerus CT image; constructing a three-dimensional surface model of the distal humerus; constructing an initial coordinate system of the distal humerus specified by ISB, and determining an initial rotation axis; setting the center of a circular arc with a circular arc angle for each side view of the distal humerus rotation area, and performing iterative fitting on the center of the circular arc and a straight line where a plurality of the centers are located to obtain a refined axis; determining the standard deviation of the distance between the corresponding intersection point and each point on the profile of each side view according to the change range of the optimization variable in the updated each side view of the distal humerus rotation area, the profile of each side view, and the intersection of each side view and the refined axis; and applying a genetic algorithm to optimize the refined axis to obtain an elbow joint rotation center axis. The application can accurately position the elbow joint rotation center axis.

Description

Technical Field

[0001] This invention relates to the field of elbow joint rotation center axis recognition, and in particular to a method, system, electronic device, and medium for elbow joint rotation center axis recognition. Background Technology

[0002] Automatic axis-fixing methods for the elbow joint can be divided into methods based on rotation function and methods based on medical imaging.

[0003] Rotation-based methods require patients to perform elbow flexion and extension movements, and external imaging devices are used to record the movement of optical sensors on the upper arm and forearm. The results are then obtained by circular arc fitting. This method is not suitable for patients who cannot flex or extend their elbows, and there are differences in each flexion and extension movement, resulting in poor repeatability of axis determination results.

[0004] Medical imaging-based methods often utilize medical anatomical knowledge, hence the term "elbow joint axis determination method based on anatomical structure." This method primarily relies on the clinician's subjective judgment, marking a point on the medial and lateral sides of the distal humerus, and using the line connecting these two points as the rotational axis of the elbow joint. There are two main ways to determine these two points: (1) considering the medial trochlear crest, lateral trochlear crest, and the maximum radius of the capitulum of the distal humerus as three concentric arcs, and finding the center of these three concentric circles; (2) using the line connecting the center of the arc of the trochlear groove of the humerus and the center of the fitting sphere of the capitulum of the humerus. Both methods require the humerus to be positioned in a standard lateral view. Both methods have the following drawbacks: ① They require a high level of experience from the doctor. The judgment of the standard lateral view is subjective, and inconsistencies may arise between different doctors or even from the same doctor at different times. Inaccurate judgment of the standard lateral view will lead to significant rotational errors. ② Both the patient and the doctor are exposed to significant X-ray radiation. Even for experienced doctors, multiple fluoroscopy sessions are required for accurate positioning. ③ When the patient has limited shoulder joint movement, especially limited abduction and internal rotation, it is often impossible to position the patient in the standard lateral view, thus making it impossible to identify concentric circles. ④ Due to congenital or acquired reasons, the anatomical landmarks are not obvious under fluoroscopy in some patients, making it impossible to determine the concentric circles. ⑤ Fractures or ligaments that have been repaired during surgery may shift or fail to be fixed during the placement of the surgical position. Clinically, Kirschner wires are usually used to maintain stability through the humeroulnar joint, but this increases the risk of injury to the patient. Furthermore, improper placement of the Kirschner wire in a single insertion can cause even greater damage.

[0005] The localization of the elbow joint rotation axis still requires significant manual intervention and has a low degree of automation. Furthermore, most axis-determining methods are highly subjective, relying heavily on the physician's experience and work status, thus affecting the reliability of the identification results. When manually locating the elbow joint rotation axis, inconsistencies in calibration results among physicians can lead to substantial rotational errors. Summary of the Invention

[0006] To address the aforementioned technical problems, this invention provides a method, system, electronic device, and medium for identifying the rotation center axis of the elbow joint, which can accurately locate the rotation center axis of the elbow joint.

[0007] To achieve the above objectives, the present invention provides the following solution:

[0008] A method for identifying the rotation center axis of an elbow joint, the method comprising:

[0009] Acquire CT images of the distal humerus;

[0010] Based on the CT images of the distal humerus, a three-dimensional surface model of the distal humerus is constructed.

[0011] The least squares method was applied to fit the distal humeral shaft axis of the three-dimensional surface model;

[0012] Based on the fitted distal humeral shaft axis, establish the ISB-specified initial coordinate system for the distal humerus and determine the initial rotation axis.

[0013] Based on the initial coordinate system of the distal humerus specified by the ISB and the initial rotation axis, the medial rib of the trochlear surface of the three-dimensional surface model is determined;

[0014] Based on the initial rotation axis, the portion between the outer surface of the humeral head and the medial ridge of the trochlea in the three-dimensional surface model is taken as the distal humeral focus rotation region. The center of the arc with a set arc angle is iteratively fitted to the side map of the distal humeral focus rotation region, and the straight lines containing multiple centers are iteratively fitted. When the fitted straight lines meet the set conditions, the fitted straight lines are taken as the thinning axis.

[0015] Based on the refinement axis and the preset angle range of the key distal humeral parts in the rotational region of concern for the distal humerus, a quadratic function is applied to fit the sagittal planes corresponding to the lateral views of the non-key distal humeral parts in the rotational region of concern for the distal humerus to obtain the rotation angle range of the non-key distal humeral parts. Then, based on the preset angle range of the key distal humeral parts and the rotation angle range of the non-key distal humeral parts, an initial region of concern for the distal humerus is obtained. The rotational region of concern for the distal humerus includes the key distal humeral parts and the non-key distal humeral parts. The key distal humeral parts include the trochlear region and the head of the humerus.

[0016] The initial distal region of interest is projected onto the humeral surface of the distal humeral region of interest rotation to obtain the updated distal humeral region of interest rotation.

[0017] The range of optimization variables corresponding to the first intersection point between the refinement axis and the medial trochlear ridge in the updated distal humeral focus rotation region is determined based on the radius of the medial trochlear ridge in the updated distal humeral focus rotation region.

[0018] The range of optimization variables corresponding to the second intersection point between the thinning axis and the lateral surface of the updated humeral head is determined based on the ball diameter of the humeral head.

[0019] Based on the rotation angle range and contour of each lateral view of the updated distal humeral focus rotation region, calculate the third intersection point of the thinning axis with each lateral view of the updated distal humeral focus rotation region, and the standard deviation of the distance between the point on the contour of the corresponding lateral view and the third intersection point.

[0020] Based on the variation range of the optimized variables corresponding to the first intersection point, the variation range of the optimized variables corresponding to the second intersection point, the rotation angle range of each side view of the updated distal humerus rotation region of interest, and the standard deviation, the genetic algorithm is applied to calculate the formula. The global optimal solution yields the elbow joint rotation center axis; where... and These are the boundary angle values ​​of the rotatable range of the i-th side view; N is the standard deviation of the distance values. S This represents the number of slice side maps.

[0021] Optionally, based on the CT images of the distal humerus, a three-dimensional surface model of the distal humerus is constructed, specifically including:

[0022] Based on the CT images of the distal humerus, a three-dimensional surface model of the distal humerus was constructed using the Marching Cube algorithm and the Laplace smoothing algorithm.

[0023] Optionally, the application of the least squares method to fit the distal humeral shaft axis of the three-dimensional surface model specifically includes:

[0024] The humeral shaft portion of the three-dimensional surface model is layered along the direction of the humeral shaft extension to obtain multiple humeral shaft layers.

[0025] Calculate the centroid of each of the described humeral shaft layers;

[0026] The centroids of each of the described humeral shaft layers are fitted with straight lines to obtain the humeral shaft axis;

[0027] By applying contour tracking and using the density of the contour region as the tracking termination condition, the direction of the humeral shaft axis is corrected to obtain the humeral shaft axis after correction.

[0028] The least squares method is applied to fit the humeral shaft axis after the correction direction to obtain the fitted humeral shaft axis.

[0029] Optionally, it specifically includes:

[0030] Establish multiple planes that are parallel to each other and perpendicular to the fitted humeral shaft axis;

[0031] By applying a contour tracking algorithm, the two points with the largest distance in each plane are determined;

[0032] Select the two points corresponding to the planes with the largest distance between the two points in each plane as the first single point on the medial epicondyle of the humerus and the second single point on the lateral epicondyle of the humerus;

[0033] Based on the first single point on the medial epicondyle of the humerus and the second single point on the lateral epicondyle of the humerus, the intersection of the fitted humeral shaft axis and the proximal surface of the three-dimensional surface model is taken as the proximal endpoint of the humeral shaft. An initial coordinate system for the distal humerus as defined by the ISB is established, and the initial rotation axis is determined. The initial coordinate system for the distal humerus as defined by the ISB includes point O, X-axis, Y-axis, and Z-axis. Point O is the midpoint between the first single point on the medial epicondyle of the humerus and the second single point on the lateral epicondyle of the humerus. The Y-axis is the straight line between the proximal endpoint of the humeral shaft and the midpoint. The X-axis is a straight line perpendicular to the plane containing the proximal endpoint of the humeral shaft, the first single point on the medial epicondyle of the humerus, and the second single point on the lateral epicondyle of the humerus. The Z-axis is a straight line perpendicular to the initial X-axis and the initial Y-axis. The initial rotation axis is the Z-axis.

[0034] Optionally, the medial trochlear ridge of the three-dimensional surface model is determined according to the initial coordinate system of the distal humerus specified by the ISB and the initial rotation axis, specifically including:

[0035] Using the plane containing the Y-axis, point O, and the Z-axis as the coronal plane of the distal humerus, and taking a preset first angle as the step size, the coronal plane is rotated around the Z-axis by a preset second angle to obtain multiple rotational plane systems;

[0036] The point furthest from the Z-axis at a preset location among the intersections of each of the rotating surface systems and the humeral surface of the three-dimensional surface model is used as the marker point;

[0037] By fitting multiple of the aforementioned landmarks, the medial ridge of the distal trochlea of ​​the humerus is obtained.

[0038] Optionally, based on the initial rotation axis, the portion between the lateral surface of the humeral head and the medial ridge of the trochlea in the three-dimensional surface model is taken as the distal humeral region of interest for rotation. The centers of the arcs with set arc angles on each side view of the distal humeral region of interest for rotation are iteratively fitted, and the straight lines containing multiple such centers are iteratively fitted. When the fitted straight lines meet set conditions, the fitted straight lines are used as the refinement axis, specifically including:

[0039] Based on the initial rotation axis, the portion between the outer surface of the humeral head and the medial ridge of the trochlea in the three-dimensional surface model is taken as the distal humeral focus rotation region. The center of the arc with a set arc angle is fitted to the side view of the distal humeral focus rotation region to obtain the center of the side view of the distal humeral focus rotation region.

[0040] A temporary axis of rotation is obtained by fitting the straight line containing the center of the circle in each side view of the region of interest in the distal humerus.

[0041] Using the temporary rotation axis as the current initial rotation axis, return to the execution step "Based on the initial rotation axis, take the part between the outer surface of the humeral head and the medial ridge of the trochlea in the three-dimensional surface model as the humeral distal focus rotation area, and fit the center of the arc of the arc angle set for each side map of the humeral distal focus rotation area to obtain the center of the arc of each side map of the humeral distal focus rotation area", and obtain the updated rotation axis again;

[0042] Determine the coordinates of the first and second updated intersection points of the updated initial rotation axis with the humeral surface of the distal humeral region of interest.

[0043] Determine the coordinates of the first and second re-updated intersection points of the re-updated rotation axis with the humeral surface of the distal humeral region of interest.

[0044] Calculate the distance between the first updated intersection point coordinates and the first further updated intersection point coordinates to obtain the first distance;

[0045] Calculate the distance between the second updated intersection point coordinates and the second further updated intersection point coordinates to obtain the second distance;

[0046] When the first distance and the second distance meet the set threshold range, the updated rotation axis is used as the thinning axis.

[0047] Optionally, based on the rotation angle range and contour of each lateral view of the updated distal humeral region of interest, the standard deviation of the distance between the third intersection point of the thinning axis and each lateral view of the updated distal humeral region of interest, and the distance between the point on the contour of the corresponding lateral view and the third intersection point is calculated, specifically including:

[0048] Sagittal sections were performed on key parts of the distal humerus in the updated region of rotation to obtain a lateral view of the distal humerus.

[0049] Based on the rotation angle range and contour of each side view of the updated distal humerus focus rotation region, calculate the third intersection point of the thinning axis with each side view of the updated distal humerus focus rotation region;

[0050] Based on the rotation angle range and contour of each lateral view of the updated distal humerus focus rotation region, the distance between the point on the contour of each lateral view that satisfies the rotation angle range of the corresponding lateral view and the third intersection point is calculated to obtain multiple contour distances.

[0051] Calculate the weighted standard deviation of the multiple contour distances to obtain the standard deviation of the distance between the point on the contour of the side view and the third intersection point.

[0052] An elbow joint rotation center axis recognition system, applied to the above-mentioned elbow joint rotation center axis recognition method, the system comprising:

[0053] The acquisition module is used to acquire CT images of the distal humerus;

[0054] The first construction module is used to construct a three-dimensional surface model of the distal humerus based on the CT image of the distal humerus.

[0055] The first fitting module is used to fit the distal humeral shaft axis of the three-dimensional surface model using the least squares method.

[0056] The second construction module is used to establish the initial coordinate system of the distal humerus as specified by ISB and determine the initial rotation axis based on the fitted distal humeral shaft axis.

[0057] The trochlear medial ridge determination module is used to determine the trochlear medial ridge of the three-dimensional surface model according to the initial coordinate system of the distal humerus specified by the ISB and the initial rotation axis.

[0058] The second fitting module is used to iteratively fit the center of the arc with a set arc angle on each side map of the humeral distal tip rotation region according to the initial rotation axis, and iteratively fit the straight lines where multiple centers are located. When the fitted straight lines meet the set conditions, the fitted straight lines are used as the thinning axis.

[0059] The third fitting module is used to fit a quadratic function to the sagittal plane corresponding to each lateral view of the non-critical distal humeral parts in the rotational region of interest to the thinning axis and the preset angle range of the critical distal humeral parts, thereby obtaining the rotation angle range of the non-critical distal humeral parts. Based on the preset angle range of the critical distal humeral parts and the rotation angle range of the non-critical distal humeral parts, an initial region of interest for the distal humeral is obtained. The rotational region of interest for the distal humeral includes the critical distal humeral parts and the non-critical distal humeral parts. The critical distal humeral parts include the trochlear region and the head of the humerus.

[0060] The projection module is used to project the initial distal region of interest onto the humeral surface of the distal region of interest rotation of the humerus, so as to obtain an updated distal region of interest rotation of the humerus.

[0061] The first parameter determination module is used to determine the range of optimization variables corresponding to the first intersection point of the thinning axis and the medial trochlear ridge in the updated distal humeral focus rotation area based on the radius of the medial trochlear ridge in the updated distal humeral focus rotation area.

[0062] The second parameter determination module is used to determine the range of optimization variables corresponding to the second intersection point between the thinning axis and the lateral surface of the updated humeral head based on the ball diameter of the humeral head.

[0063] The third parameter determination module is used to calculate the third intersection point between the thinning axis and each side view of the updated distal humeral focus rotation area, as well as the standard deviation of the distance between the point on the contour of the corresponding side view and the third intersection point, based on the rotation angle range and contour of each side view of the updated distal humeral focus rotation area.

[0064] The solution module is used to calculate the solution using a genetic algorithm based on the variation range of the optimized variable corresponding to the first intersection point, the variation range of the optimized variable corresponding to the second intersection point, the rotation angle range of each side view of the updated distal humeral focus rotation region, and the standard deviation. The global optimal solution yields the elbow joint rotation center axis; where... and These are the boundary angle values ​​of the rotatable range of the i-th side view; N is the standard deviation of the distance values. S This represents the number of slice side maps.

[0065] An electronic device includes a memory and a processor, the memory storing a computer program, and the processor running the computer program to enable the electronic device to perform the above-described elbow joint rotation center axis recognition method.

[0066] A computer-readable storage medium storing a computer program, which is executed by a processor, to perform the above-described elbow joint rotation center axis identification method.

[0067] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0068] This invention provides a method for identifying the rotation center axis of the elbow joint. The method involves acquiring a CT image of the distal humerus; constructing a three-dimensional surface model of the distal humerus based on the CT image; fitting the distal humeral shaft axis of the three-dimensional surface model using the least squares method; establishing an initial coordinate system for the distal humerus as specified by the ISB and determining the initial rotation axis based on the fitted distal humeral shaft axis; determining the medial trochlear crest of the three-dimensional surface model based on the initial coordinate system and initial rotation axis as specified by the ISB; and using the portion between the lateral surface of the humeral head and the medial trochlear crest in the three-dimensional surface model as the distal humeral shaft axis. The focus is on the rotation region of the distal humerus. Iterative fitting is performed on the centers of arcs with defined angles for each lateral view of the rotation region, and iterative fitting is also performed on the lines containing multiple centers. When the fitted line meets the set conditions, the fitted line is used as the thinning axis. Based on the thinning axis and the preset angle range of key distal humeral regions within the rotation region, a quadratic function is applied to the sagittal planes corresponding to non-key distal humeral regions within the rotation region to obtain the rotation angle range of non-key distal humeral regions. Finally, based on the preset angle range of key distal humeral regions and non-key distal humeral regions... The rotation angle range of key distal humeral regions is used to obtain the initial distal humeral region of interest. This initial region of interest is projected onto the humeral surface of the distal humeral region of interest rotation to obtain the updated distal humeral region of interest rotation. The range of optimization variables corresponding to the first intersection point of the thinning axis and the medial trochlear crest in the updated distal humeral region of interest rotation is determined based on the radius of the medial trochlear crest. The range of optimization variables corresponding to the second intersection point of the thinning axis and the lateral surface of the updated distal humeral head is determined based on the diameter of the humeral head. The range of optimization variables corresponding to the second intersection point of the thinning axis and the lateral surface of the updated distal humeral head of interest rotation is determined based on the lateral dimensions of the updated distal humeral region of interest rotation. The rotation angle range and contour of the image are calculated, and the third intersection point of the refined axis with each lateral view of the updated distal humeral region of interest, as well as the standard deviation of the distance between the point on the contour of the corresponding lateral view and the third intersection point, are calculated. Based on the variation range of the optimization variables corresponding to the first intersection point, the variation range of the optimization variables corresponding to the second intersection point, and the rotation angle range and standard deviation of each lateral view of the updated distal humeral region of interest, a genetic algorithm is applied to obtain the elbow joint rotation center axis. This realizes a multi-level gradual optimization process from coarse fitting of the rotation axis to fine fitting of the rotation axis, and then to optimization of the rotation center axis, thereby improving the positioning accuracy of the elbow joint rotation center axis. Attached Figure Description

[0069] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0070] Figure 1 Flowchart of the elbow joint rotation center axis identification method provided by the present invention;

[0071] Figure 2 A schematic diagram of the elbow joint rotation center axis identification process provided in a specific embodiment of the present invention;

[0072] Figure 3 A schematic diagram of the three-dimensional humeral surface file provided by the present invention;

[0073] Figure 4 A schematic diagram of humeral orientation correction and humeral shaft axis acquisition provided by the present invention;

[0074] Figure 5 A schematic diagram of the medial and lateral epicondyles of the humerus provided by this invention;

[0075] Figure 6 This is a schematic diagram of obtaining the medial crest of the humeral trochlea provided by the present invention;

[0076] Figure 7 This is a schematic diagram of the forward 80-degree rotation range provided by the present invention;

[0077] Figure 8 This is a schematic diagram of the region of interest in the distal humerus provided by the present invention;

[0078] Figure 9 A schematic diagram showing the location of the intersection point between the temporary rotation axis and the humeral surface provided by the present invention;

[0079] Figure 10 An inside view schematic diagram of the temporary rotation axis acquisition process provided by the present invention;

[0080] Figure 11 A schematic diagram from the outside of the temporary rotation axis acquisition process provided by the present invention;

[0081] Figure 12 A schematic diagram of the elbow joint articular surface provided by the present invention;

[0082] Figure 13 This is a schematic diagram of the range of rotation of the distal humerus provided by the present invention;

[0083] Figure 14 A schematic diagram illustrating the division of the sagittal region of the distal humerus provided by the present invention;

[0084] Figure 15 A schematic diagram illustrating the division of the sagittal plane of the distal humerus provided by this invention;

[0085] Figure 16 A schematic diagram of the polar coordinate representation of the elbow joint rotation axis provided by the present invention;

[0086] Figure 17 A schematic diagram of the sagittal plane containing the trochlear groove of the humerus, the sagittal plane containing the lateral trochlear crest of the humerus, and the humeral capitulum region fitted by the present invention.

[0087] Figure 18 A block diagram of the elbow joint rotation center axis recognition system provided by the present invention.

[0088] Explanation of reference numerals in the attached figures:

[0089] Acquisition Module—1, First Construction Module—2, First Fitting Module—3, Second Construction Module—4, Determination Module for Medial Trochlear Ridge—5, Second Fitting Module—6, Third Fitting Module—7, First Parameter Determination Module—8, Solution Module—9, Medial Epicanthus—13, Lateral Epicanthus—14, Head of Humerus—15, Lateral Trochlear Ridge—16, Trochlear Groove of Humerus—17, Medial Trochlear Ridge—18, Tip of Olecranon Notch of Ulna—19, Head of Radial—20, Projection Module—10, Second Parameter Determination Module—11, Third Parameter Determination Module—12. Detailed Implementation

[0090] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0091] The purpose of this invention is to provide a method, system, electronic device, and medium for identifying the rotation center axis of the elbow joint, which can accurately locate the rotation center axis of the elbow joint.

[0092] This invention primarily addresses the problem of automatic and precise positioning of the elbow joint rotation axis during surgeries such as elbow fixation and elbow replacement. By inputting the patient's three-dimensional humeral surface file, it can automatically perform a multi-level progressive optimization process of determining the rotation axis, from coarse to fine and then to optimized. Combined with prior medical knowledge, it achieves highly accurate automatic identification of the elbow joint rotation axis and can quantify the rotational accuracy. Therefore, this invention proposes a fully automatic method for determining the elbow joint rotation axis, applicable to the automatic identification of the humeral shaft axis in some humeral parts, and also provides a method for evaluating the accuracy of elbow joint rotation axis determination.

[0093] The following explains some technical terms related to technical solutions:

[0094] Elbow joint: The elbow joint is a trochlear joint that connects the upper arm (humerus) and the forearm (radius and ulna). It includes the humeral joint and the humeroradial joint, and has an approximately hinge-like structure. It affects the forearm and wrist joints through bending and straightening, thereby realizing the motor function of the upper limb.

[0095] Elbow joint rotation axis: also known as the elbow joint flexion-extension axis, studies have shown that the elbow joint can be considered as a single rotation axis for flexion and extension movements.

[0096] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0097] Example 1

[0098] like Figure 1 and Figure 2 As shown, the present invention provides a method for identifying the rotation center axis of an elbow joint, the method comprising:

[0099] Step 101: Obtain CT images of the distal humerus.

[0100] Step 102: Construct a 3D surface model of the distal humerus based on CT images of the distal humerus. Specifically, based on the CT images of the distal humerus, the Marching Cube algorithm and the Laplace smoothing algorithm are applied to construct a 3D surface model of the distal humerus. In practical applications, CT images of the distal humerus are collected, and the structures of the distal humerus are accurately segmented. The Marching Cube algorithm and the Laplace smoothing algorithm are used to construct a 3D surface file (.stl file), such as... Figure 3 As shown.

[0101] Step 103: Apply the least squares method to fit the distal humeral shaft axis of the 3D surface model; specifically, when the imaging region only contains the distal humerus, due to the lack of the glenohumeral joint center point, a short cylindrical fitting method for a portion of the shaft segment is used to estimate the humeral shaft axis. Contour tracking is performed in the correction direction, with the density of the contour region as the tracking termination condition. Within this segment, the least squares fitting centroid method is used to fit the humeral shaft axis. To control the shaft axis angle error, an iterative verification method is used. Figure 4 As shown, the dark blue arrows indicate the coordinate direction of the CT scan, the green and orange arrows indicate the orientation of the humerus after initial correction, and the dark blue area is the relevant area used for fitting the humeral shaft axis.

[0102]

[0103]

[0104] Among them, HSi Let p be the backbone axis of the i-th iteration. i and l i This represents the vertex and direction vector of the axis, where M is the number of parallel contours contained in the skeleton region. For the corresponding centroid; θ t This represents the tolerance value for the backbone angle error. i and l i-1 Let L be the axis vector at the i-th iteration and the axis vector at the (i-1)-th iteration, respectively. error This represents the angle difference between the i-th axis vector and the (i-1)-th axis vector. If this value is less than a threshold... Then the axis is feasible; LS ine This represents least-squares linear fitting.

[0105] Step 103 specifically includes:

[0106] Step 1031: The humeral shaft portion of the three-dimensional surface model is layered along the direction of the humeral shaft extension to obtain multiple humeral shaft layers.

[0107] Step 1032: Calculate the centroid of each of the described humeral shaft layers.

[0108] Step 1033: Perform linear fitting on the centroid of each of the described humeral shaft layers to obtain the humeral shaft axis.

[0109] Step 1034: Apply contour tracking, using the density of the contour region as the tracking termination condition, to correct the direction of the humeral shaft axis, and obtain the corrected humeral shaft axis.

[0110] Step 1035: Apply the least squares method to fit the humeral shaft axis after the direction is corrected, and obtain the fitted humeral shaft axis.

[0111] Step 104: Based on the fitted distal humeral shaft axis, establish the initial coordinate system of the distal humerus as specified by the ISB and determine the initial rotation axis; specifically, establish a parallel plane system perpendicular to the humeral shaft axis, perform contour tracking, and find the point pairs corresponding to the maximum distance within the layer, denoted as (p ME ,p LE This refers to a single point on the medial epicondyle 13 and a single point on the lateral epicondyle 14 of the humerus. For example... Figure 5 As shown in Table 1, the intersection of the humeral shaft axis and the proximal plane of the humeral model is named the Humerus Proximal Point (HPP). An Initial Distal Humerus Coordinate System (IDHCS) as defined by the ISB is established. Where Z... initial The elbow joint rotation axis is approximately defined by the ISB and is referred to as the initial rotation axis.

[0112] Table 1. Explanation of the Initial Coordinate System for the Distal Humerus

[0113]

[0114] Step 104 specifically includes:

[0115] Step 1041: Establish multiple planes that are parallel to each other and perpendicular to the fitted humeral shaft axis.

[0116] Step 1042: Apply the contour tracking algorithm to determine the two points with the largest distance in each plane.

[0117] Step 1043: Select the two points corresponding to the planes with the largest distance between the two points in each plane as the first single point on the medial epicondyle 13 of the humerus and the second single point on the lateral epicondyle 14 of the humerus.

[0118] Step 1044: Based on the first single point on the medial epicondyle 13 and the second single point on the lateral epicondyle 14 of the humerus, and taking the intersection of the fitted humeral shaft axis and the proximal surface of the three-dimensional surface model as the proximal endpoint of the humeral shaft, establish the initial coordinate system of the distal humerus as specified by the ISB and determine the initial rotation axis; the initial coordinate system of the distal humerus as specified by the ISB includes point O, X-axis, Y-axis and Z-axis; point O is the midpoint of the first single point on the medial epicondyle 13 and the second single point on the lateral epicondyle 14; the Y-axis is the straight line between the proximal endpoint of the humeral shaft and the midpoint; the X-axis is the straight line perpendicular to the plane containing the proximal endpoint of the humeral shaft, the first single point on the medial epicondyle 13 and the second single point on the lateral epicondyle 14; the Z-axis is the straight line perpendicular to the initial X-axis and the initial Y-axis; the initial rotation axis is the Z-axis.

[0119] Step 105: Determine the medial trochlear ridge of the three-dimensional surface model according to the initial coordinate system and initial rotation axis of the distal humerus specified by the ISB.

[0120] Specifically, the humeral-ulnar joint is characterized by the matching of the trochlear and ulnar trochlear notch regions, while the humeral-radial joint is characterized by the matching of the capitol 15 region of the humerus and the radial head 20 region. Therefore, the range of articular surfaces involved in elbow joint rotation on the humerus can be considered as the region from the medial trochlear flange 18 to the lateral region of the capitol 15, hereinafter referred to as the distal humeral region of concern.

[0121] Due to the unique structures of the humeral-ulnar and humeral-radial joints, the rotation radius and angle range of the articular mating surfaces vary considerably across different lateral views. However, the distal humerus, in the coronal plane and within its anterior 80-degree range, exhibits a relatively stable arc shape in the lateral view, which can be approximated as a variable-radius helical axis structure.

[0122] Step 105 specifically includes:

[0123] Step 1051: Using the plane containing the Y-axis, point O, and the Z-axis as the coronal plane of the distal humerus, and taking a preset first angle as the step size, rotate the coronal plane around the Z-axis by a preset second angle to obtain multiple rotational plane systems.

[0124] As a specific implementation method, the initial coordinate system YOZ initial The plane is the initial coronal plane of the distal humerus, around the Z-axis. initial Rotate the axis 80° anteriorly toward the humerus in steps of 10° to obtain the rotation plane system {RS}. i},like Figure 6 As shown.

[0125] Step 1052: Use the point that is furthest from the Z-axis at a preset location among the intersections of each rotational surface system and the humeral surface of the three-dimensional surface model as the marker point.

[0126] As a specific implementation method, in each RS i Above, identify the point on the articular surface portion of the intersection of this surface and the humeral surface, closest to the medial epicondyle 13, that is the most convex point from the coarse fixed axis, and denote it as the landmark point MRP. i (Medial RidgePoint).

[0127] Step 1053: Fit multiple landmarks to obtain the medial ridge of the distal trochlea of ​​the humerus.

[0128] Specifically, by integrating the landmarks obtained from each plane in the rotational plane system, the medial ridge plane of the distal trochlea of ​​the humerus is obtained through fitting, such as... Figure 7 As shown.

[0129] Step 106: Based on the initial rotation axis, take the portion between the outer surface of the humeral head and the medial ridge of the trochlea in the three-dimensional surface model as the humeral distal focus rotation region. Iteratively fit the center of the arc with the set arc angle of each side map of the humeral distal focus rotation region and iteratively fit the straight lines where multiple centers are located. When the fitted straight lines meet the set conditions, the fitted straight lines are used as the thinning axis.

[0130] Step 106 specifically includes:

[0131] Step 1061: Based on the initial rotation axis, using the portion between the lateral surface of the humeral head and the medial ridge of the trochlea in the three-dimensional surface model as the distal humeral region of interest rotation, fit the center of the arc with the set arc angle in each side view of the distal humeral region of interest rotation to obtain the center of the arc in each side view of the distal humeral region of interest rotation. In practical applications, the arc angle considered in each side view is the same. In practice, -20 degrees to 120 degrees is generally good, while some samples require adjustment of the parameter to -30 degrees to 30 degrees.

[0132] Step 1062: Fit the straight line containing the center of the circle in each side view of the rotation region of interest in the distal humerus to obtain a temporary rotation axis.

[0133] Step 1063: Using the temporary rotation axis as the current initial rotation axis, return to step 1061 to obtain the updated rotation axis.

[0134] Step 1064: Determine the coordinates of the first and second updated intersection points of the updated initial rotation axis with the humeral surface of the distal humeral region of interest.

[0135] Step 1065: Determine the coordinates of the first and second re-updated intersection points of the re-updated rotation axis with the humeral surface of the distal humeral region of interest.

[0136] Step 1066: Calculate the distance between the first updated intersection point coordinates and the first updated intersection point coordinates to obtain the first distance.

[0137] Step 1067: Calculate the distance between the second updated intersection point coordinates and the second updated intersection point coordinates to obtain the second distance.

[0138] Step 1068: When the first distance and the second distance meet the set threshold range, the updated rotation axis is used as the thinning axis.

[0139] As a specific implementation, the distal lateral head of the humerus, during elbow joint rotation, engages with the proximal radius to form the humeroradial joint. However, a small portion of its outermost part has limited engagement with the radius and exhibits significant irregularity in its arc. Therefore, excluding this outermost irregular portion during axis determination does not affect the reliability of the calculation results. Since the purpose of refining the axis is to obtain a more stable axis that approximates the ideal rotation axis, it is acceptable to discard the outer 10% portion of the current axis. This leads to the relevant parts of the elbow joint rotation center axis refinement process, as follows: Figure 8 As shown.

[0140] Furthermore, such as Figure 9 , Figure 10 and Figure 11 As shown, in the lateral view region of interest for the elbow joint, alternating fitting of the arc center and straight-line fitting of the arc center are performed to achieve iteration of the temporary coordinate system of the distal humerus and update of the rotation axis. This process is iterated until the change in the position of the intersection point of the rotation axis and the humeral surface is less than a threshold, thereby obtaining the optimal axis of the spiral fan-shaped cylinder in the region of interest. The temporary coordinate system is shown in Table 2.

[0141]

[0142]

[0143] Among them, v j This indicates that the m-th lateral section P m The point located above and within the range of 0° to 80° anterior to the coronal surface, O m P represents m The center of the circle fitted to the circular arc profile on the cross section, p medial and p lateral LS represents the intersection points of the temporary axis with the medial and lateral sides of the distal humerus, respectively. circle This represents least-squares circle center fitting. Representing point v j The corresponding angle in the temporary coordinate system of the humerus; LS line This represents least-squares linear fitting; Represents the centers of the circles; RA i Let represent the temporary axis of rotation for the i-th time.

[0144] Table 2. Explanation of the Temporary Coordinate System for the Distal Humerus

[0145]

[0146]

[0147] The detailed steps of the iterative refinement process are as follows:

[0148] Step 1: Determine the lateral region of interest for the elbow joint, including the initial axis of rotation and the medial rib of the trochlea; use the range of the anterior rotation angle of the coronal plane as a constraint, and use the maximum number of iterations or the change in distance between the medial and lateral points of the temporary axis of rotation satisfying a threshold range as the iteration termination condition.

[0149] Step 2: Calculate the lateral view slices of the lateral region of focus on the elbow joint.

[0150] Step 3: Calculate the points within the rotation angle range in each side map.

[0151] Step 4: Fit the center of each sagittal plane. The sagittal plane is a slice of the lateral view.

[0152] Step 5: Fit the axis of the circle center.

[0153] Step 6: Calculate the intersection of the axis and the surface of the humerus.

[0154] Step 7: When the change in the distance between the intersection points is less than the distance change threshold, update the temporary local coordinate system. The axis calculated in Step 6 is the iteratively refined axis. When the change in the distance between the intersection points is not less than the distance change threshold, update the temporary local coordinate system to obtain the updated temporary axis, and return to Step 1.

[0155] Step 107: Based on the thinning axis and the preset angle range of the key distal humeral parts in the rotational region of concern for the distal humeral, a quadratic function is applied to fit the sagittal planes corresponding to the lateral views of the non-key distal humeral parts in the rotational region of concern for the distal humeral to obtain the rotation angle range of the non-key distal humeral parts. Based on the preset angle range of the key distal humeral parts and the rotation angle range of the non-key distal humeral parts, an initial region of concern for the distal humeral is obtained. The rotational region of concern for the distal humeral includes the key distal humeral parts and the non-key distal humeral parts. The key distal humeral parts include the trochlear region and the head of the humerus.

[0156] Step 108: Project the initial distal focus area onto the humeral surface of the distal focus rotation area of ​​the humerus to obtain the updated distal focus rotation area of ​​the humerus; the key distal humeral parts include the trochlear region and the humeral head 15; the trochlear region includes the medial trochlear crest 18, the trochlear groove 17, the lateral trochlear crest 16, and the portion between the medial trochlear crest 18 and the trochlear groove 17, and also includes the portion between the trochlear groove 17 and the lateral trochlear crest 16.

[0157] As a specific implementation, the range of rotational coordination angles of the humeral ulnar joint and the humeral radioulnar joint varies in different lateral views, such as... Figure 12 As shown. For a normal elbow joint, its range of rotation is generally around 140°. The trochlear groove 17 of the humerus corresponds to the tip 19 of the olecranon semilunar notch of the ulna. Since the olecranon semilunar notch contains a 190° articular surface, to maintain a normal range of rotation of 140°, it should have a 330° articular surface to match it. The articular matching requirements of the medial and lateral trochlear crests are lower, especially the lateral trochlear crest 16, which has almost no matching relationship with the ulna and radius. During elbow joint rotation, the main matching relationship of the humeral capitulum 15 is in the radial head 20 region. Since the radial head 20 only contains a 40° articular surface, the humeral capitulum region should contain a 180° rotational articular surface, and this rotational articular surface should have a significant anterior offset corresponding to the coronal plane of the humerus.

[0158] The rotational ranges considered for the trochlear and capitulum regions of the humerus are shown in Table 3. The remaining rotational angles corresponding to the sagittal planes are fitted using quadratic functions to map the rotational ranges onto the distal humeral model, as shown below. Figure 13 As shown.

[0159] Table 3. Explanation of Rotation Range at Key Locations of the Distal Humerus

[0160]

[0161] The humeral head 15 can be considered as a partial sphere. Spherical fitting is used to predict the humeral head region, achieving separation of the humeral head 15 from the humeral trochlea. The trochlear groove 17 is defined as the point where the rotational arc of the humeral trochlea region is minimized. Figure 14 So and Figure 15 As shown; in Figure 15 In the diagram, L represents the span of the region of interest in the sagittal direction; d represents the cut criteria for the lateral surface of the humeral head 15.

[0162] The mathematical representations of the fitted surface of the trochlear groove of the humerus and the fitted surface of the head of the humerus are shown below:

[0163]

[0164]

[0165]

[0166]

[0167]

[0168] in, π is the direction vector of the current axis, π1 represents the plane of the inner ridge of the trochlea, π2 represents the cross section where the outer vertex of the temporary rotation axis is located, and π3 and π4 represent the inner and outer cross sections used to fit the humeral head region, respectively. t The radius threshold for fitting the humeral capitulum region is represented by x, y, and z; x, y, and z are the coordinates of points on each lateral view within the region of interest for the distal humerus; x1, y1, and z1 are the coordinates of points on the medial trochlear crest plane; x2, y2, and z2 are the coordinates of points on the cross section containing the lateral vertex of the temporary rotation axis; x3, y3, and z3 are the coordinates of points on the medial cross section of the fitted humeral capitulum region; and x4, y4, and z4 are the coordinates of points on the lateral cross section of the fitted humeral capitulum region. The fitting surface of the trochlear groove 17 includes the cross section containing the medial trochlear crest plane and the lateral vertex of the temporary rotation axis, while the fitting surface of the capitulum 15 includes the medial and lateral cross sections used to fit the capitulum region.

[0169] Step 109: Determine the range of optimization variables corresponding to the first intersection point of the refinement axis and the medial trochlear ridge in the updated distal humeral focus rotation region based on the radius of the medial trochlear ridge in the updated distal humeral focus rotation region.

[0170] Step 110: Determine the range of optimization variables corresponding to the second intersection point of the thinning axis and the lateral surface of the updated humeral head based on the ball diameter of the humeral head.

[0171] Step 111: Based on the rotation angle range and contour of each lateral view of the updated distal humeral focus rotation region, calculate the standard deviation of the distance between the third intersection point of the thinning axis and each lateral view of the updated distal humeral focus rotation region, as well as the distance between the point on the contour of the corresponding lateral view and the third intersection point.

[0172] Step 111 specifically includes:

[0173] Step 1111: Perform sagittal sections on key parts of the distal humerus in the updated region of rotation to obtain a lateral view of the distal humerus.

[0174] Step 1112: Based on the rotation angle range and contour of each side view of the updated distal humeral focus rotation region, calculate the third intersection point between the thinning axis and each side view of the updated distal humeral focus rotation region.

[0175] Step 1113: Based on the rotation angle range and contour of each lateral view of the updated distal humerus focus rotation region, calculate the distance between the point on the contour of each lateral view that satisfies the rotation angle range of the corresponding lateral view and the third intersection point, and obtain multiple contour distances.

[0176] Step 1114: Calculate the weighted standard deviation of the multiple contour distances to obtain the standard deviation of the distance between the point on the contour of the side view and the third intersection point.

[0177] As a specific implementation method, the iterative refinement process has obtained a relatively stable and accurate rotation center axis, which is used to refine the intersection of the axis with the medial ridge of the humeral trochlea and the preset sagittal section of the humeral capitulum. and To optimize the initial conditions, the optimization variable axes are obtained by moving the points on these two side views, and are described in polar coordinates, such as... Figure 16 As shown. The first intersection point is... The second intersection point is The optimization variables can be represented by the following formula.

[0178] var=[r (m) θ (m) r (l)θ (l) ]

[0179] The range of variation for optimization variables was set based on the estimated radius of the trochlear crest of the humerus and the diameter of the fitted sphere of the humeral capitulum. The sagittal plane containing the trochlear groove 17, the sagittal plane containing the lateral trochlear crest 16, and the fitted humeral capitulum region are shown below. Figure 17 As shown.

[0180] In practical applications, sagittal sections of the distal humerus are processed, denoted as {S}. i}, i = 1, 2, ..., N S Based on the data in Table 3, the distal humerus was divided into zones, and the rotation angle corresponding to each lateral view was calculated by interpolation. Simultaneously calculate the intersection point between the current temporary rotation axis and the sagittal plane. Subsequently, the correlation contours between each sagittal section and the humeral surface model are calculated. Then, a set of points within a rotation angle range is extracted from this contour, and its correlation with the model is calculated. Find the standard deviation of the distance.

[0181] Step 113: Based on the optimization variable variation range corresponding to the first intersection point, the optimization variable variation range corresponding to the second intersection point, and the rotation angle range and standard deviation of each side view of the updated distal humerus rotation region of interest, apply the genetic algorithm to calculate the formula. The global optimal solution yields the elbow joint rotation center axis; where... and These are the boundary angle values ​​of the rotatable range of the i-th side view; N is the standard deviation of the distance values. S This represents the number of slice side maps.

[0182] In practical applications, the points on the relevant contours of each sagittal slice and the humeral surface model are calculated separately. The initial distance, the points on the relevant contour of each sagittal slice and the humeral surface model. Each contour has multiple initial distances. The standard deviations of these initial distances are calculated, resulting in a single standard deviation for each sagittal slice. Multiple sagittal slices correspond to multiple standard deviations. The formula is then applied... The accuracy of the humeral rotation axis is estimated to obtain the accuracy of the elbow joint rotation center axis.

[0183] The elbow joint rotation center axis identification method provided by this invention has the following technical effects:

[0184] (1) Compared with methods based on rotation function, the method proposed in this invention does not require the patient to perform multiple rotation movements, and is especially suitable for patients whose elbow joint injury affects their mobility.

[0185] (2) Compared with the “three concentric axes” and the “line connecting the center of the trochlear groove of the humerus and the center of the capitulum of the humerus” method, the method in this paper considers a more complete articular surface area of ​​the distal elbow joint of the humerus, and considers different rotation angle ranges that should be focused on in different lateral views based on prior anatomical knowledge.

[0186] (3) Compared with existing fixed-axis methods, the method of the present invention has better stability and automation.

[0187] Example 2

[0188] To implement the method corresponding to Embodiment 1 above and achieve the corresponding functions and technical effects, an elbow joint rotation center axis recognition system is provided below, such as... Figure 18 As shown, the elbow joint rotation center axis recognition system includes:

[0189] Acquisition module 1 is used to acquire CT images of the distal humerus.

[0190] The first construction module 2 is used to construct a three-dimensional surface model of the distal humerus based on the CT image of the distal humerus.

[0191] The first fitting module 3 is used to fit the distal humeral shaft axis of the three-dimensional surface model using the least squares method.

[0192] The second construction module 4 is used to establish the initial coordinate system of the distal humerus as specified by ISB and determine the initial rotation axis based on the fitted distal humeral shaft axis.

[0193] The trochlear medial ridge determination module 5 is used to determine the trochlear medial ridge of the three-dimensional surface model according to the initial coordinate system of the distal humerus specified by the ISB and the initial rotation axis.

[0194] The second fitting module 6 is used to iteratively fit the center of the arc with a set arc angle on each side map of the humeral distal tip rotation region according to the initial rotation axis, taking the part between the outer surface of the humeral capillary and the inner lateral ridge of the trochlea in the three-dimensional surface model as the humeral distal tip rotation region of interest, and iteratively fit the straight lines where the multiple centers are located. When the fitted straight lines meet the set conditions, the fitted straight lines are used as the thinning axis.

[0195] The third fitting module 7 is used to fit a quadratic function to the sagittal plane corresponding to each lateral view of the non-critical distal humeral part in the rotational region of the humeral distal focus, based on the thinning axis and the preset angle range of the critical distal humeral part in the rotational region of the humeral distal focus, to obtain the rotation angle range of the non-critical distal humeral part, and to obtain the initial humeral distal focus region based on the preset angle range of the critical distal humeral part and the rotation angle range of the non-critical distal humeral part; the humeral distal focus region includes the critical distal humeral part and the non-critical distal humeral part; the critical distal humeral part includes the trochlear region and the head of the humerus.

[0196] Projection module 8 is used to project the initial distal region of interest onto the humeral surface of the distal region of interest rotation of the humerus, so as to obtain an updated distal region of interest rotation of the humerus.

[0197] The first parameter determination module 9 is used to determine the range of optimization variables corresponding to the first intersection point of the thinning axis and the medial trochlear ridge in the updated distal humeral focus rotation region based on the radius of the medial trochlear ridge in the updated distal humeral focus rotation region.

[0198] The second parameter determination module 10 is used to determine the range of optimization variables corresponding to the second intersection point of the thinning axis and the lateral surface of the updated humeral head based on the ball diameter of the humeral head.

[0199] The third parameter determination module 11 is used to calculate the third intersection point of the thinning axis with each side view of the updated distal humeral focus rotation area and the standard deviation of the distance between the point on the contour of the corresponding side view and the third intersection point, based on the rotation angle range and contour of each side view of the updated distal humeral focus rotation area.

[0200] The calculation module 12 is used to calculate the formula using a genetic algorithm based on the variation range of the optimized variable corresponding to the first intersection point, the variation range of the optimized variable corresponding to the second intersection point, the rotation angle range of each side view of the updated distal humeral focus rotation region, and the standard deviation. The global optimal solution yields the elbow joint rotation center axis; where... and These are the boundary angle values ​​of the rotatable range of the i-th side view; N is the standard deviation of the distance values. S This represents the number of slice side maps.

[0201] Example 3

[0202] This invention provides an electronic device, including a memory and a processor. The memory stores a computer program, and the processor runs the computer program to enable the electronic device to perform the elbow joint rotation center axis recognition method of Embodiment 1.

[0203] Alternatively, the aforementioned electronic device may be a server.

[0204] In addition, this embodiment of the invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the elbow joint rotation center axis recognition method of Embodiment 1.

[0205] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0206] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for identifying the rotation center axis of an elbow joint, characterized in that, The method includes: Acquire CT images of the distal humerus; Based on the CT images of the distal humerus, a three-dimensional surface model of the distal humerus is constructed. The least squares method was applied to fit the distal humeral shaft axis of the three-dimensional surface model; Based on the fitted distal humeral shaft axis, establish the ISB-specified initial coordinate system for the distal humerus and determine the initial rotation axis. Based on the initial coordinate system of the distal humerus specified by the ISB and the initial rotation axis, the medial rib of the trochlear surface of the three-dimensional surface model is determined; Based on the initial rotation axis, the portion between the outer surface of the humeral head and the medial ridge of the trochlea in the three-dimensional surface model is taken as the distal humeral focus rotation region. The center of the arc with a set arc angle is iteratively fitted to the side map of the distal humeral focus rotation region, and the straight lines containing multiple centers are iteratively fitted. When the fitted straight lines meet the set conditions, the fitted straight lines are taken as the thinning axis. Based on the refinement axis and the preset angle range of the key distal humeral parts in the rotational region of concern for the distal humerus, a quadratic function is applied to fit the sagittal planes corresponding to the lateral views of the non-key distal humeral parts in the rotational region of concern for the distal humerus to obtain the rotation angle range of the non-key distal humeral parts. Then, based on the preset angle range of the key distal humeral parts and the rotation angle range of the non-key distal humeral parts, an initial region of concern for the distal humerus is obtained. The rotational region of concern for the distal humerus includes the key distal humeral parts and the non-key distal humeral parts. The key distal humeral parts include the trochlear region and the head of the humerus. The initial distal region of interest is projected onto the humeral surface of the distal humeral region of interest rotation to obtain the updated distal humeral region of interest rotation. The range of optimization variables corresponding to the first intersection point between the refinement axis and the medial trochlear ridge in the updated distal humeral focus rotation region is determined based on the radius of the medial trochlear ridge in the updated distal humeral focus rotation region. The range of optimization variables corresponding to the second intersection point between the thinning axis and the lateral surface of the updated humeral head is determined based on the ball diameter of the humeral head. Based on the rotation angle range and contour of each lateral view of the updated distal humeral focus rotation region, calculate the third intersection point of the thinning axis with each lateral view of the updated distal humeral focus rotation region, and the standard deviation of the distance between the point on the contour of the corresponding lateral view and the third intersection point. Based on the variation range of the optimized variables corresponding to the first intersection point, the variation range of the optimized variables corresponding to the second intersection point, the rotation angle range of each side view of the updated distal humerus rotation region of interest, and the standard deviation, the genetic algorithm is applied to calculate the formula. The global optimal solution yields the elbow joint rotation center axis; where... and These are the boundary angle values ​​of the rotatable range of the i-th side view; The standard deviation of the distance values; This represents the number of slice side maps.

2. The elbow joint rotation center axis identification method according to claim 1, characterized in that, Based on the CT images of the distal humerus, a three-dimensional surface model of the distal humerus is constructed, specifically including: Based on the CT images of the distal humerus, a three-dimensional surface model of the distal humerus was constructed using the Marching Cube algorithm and the Laplace smoothing algorithm.

3. The elbow joint rotation center axis identification method according to claim 1, characterized in that, The application of the least squares method to fit the distal humeral shaft axis of the three-dimensional surface model specifically includes: The humeral shaft portion of the three-dimensional surface model is layered along the direction of the humeral shaft extension to obtain multiple humeral shaft layers. Calculate the centroid of each of the described humeral shaft layers; The centroids of each of the described humeral shaft layers are fitted with straight lines to obtain the humeral shaft axis; By applying contour tracking and using the density of the contour region as the tracking termination condition, the direction of the humeral shaft axis is corrected to obtain the humeral shaft axis after correction. The least squares method is applied to fit the humeral shaft axis after the correction direction to obtain the fitted humeral shaft axis.

4. The elbow joint rotation center axis identification method according to claim 1, characterized in that, Based on the fitted distal humeral shaft axis, establish the ISB-specified initial coordinate system for the distal humerus and determine the initial rotation axis, specifically including: Establish multiple planes that are parallel to each other and perpendicular to the fitted humeral shaft axis; By applying a contour tracking algorithm, the two points with the largest distance in each plane are determined; Select the two points corresponding to the planes with the largest distance between the two points in each plane as the first single point on the medial epicondyle of the humerus and the second single point on the lateral epicondyle of the humerus; Based on the first single point on the medial epicondyle of the humerus and the second single point on the lateral epicondyle of the humerus, the intersection of the fitted humeral shaft axis and the proximal surface of the three-dimensional surface model is taken as the proximal endpoint of the humeral shaft. An initial coordinate system for the distal humerus as defined by the ISB is established, and the initial rotation axis is determined. The initial coordinate system for the distal humerus as defined by the ISB includes point O, X-axis, Y-axis, and Z-axis. Point O is the midpoint between the first single point on the medial epicondyle of the humerus and the second single point on the lateral epicondyle of the humerus. The Y-axis is the straight line between the proximal endpoint of the humeral shaft and the midpoint. The X-axis is a straight line perpendicular to the plane containing the proximal endpoint of the humeral shaft, the first single point on the medial epicondyle of the humerus, and the second single point on the lateral epicondyle of the humerus. The Z-axis is a straight line perpendicular to the initial X-axis and the initial Y-axis. The initial rotation axis is the Z-axis.

5. The elbow joint rotation center axis identification method according to claim 4, characterized in that, Based on the initial coordinate system of the distal humerus specified by the ISB and the initial axis of rotation, the medial trochlear ridge of the three-dimensional surface model is determined, specifically including: Using the plane containing the Y-axis, point O, and the Z-axis as the coronal plane of the distal humerus, and taking a preset first angle as the step size, the coronal plane is rotated around the Z-axis by a preset second angle to obtain multiple rotational plane systems; The point furthest from the Z-axis at a predetermined location among the intersections of each of the rotating surface systems and the humeral surface of the three-dimensional surface model is used as the marker point; By fitting multiple of the aforementioned landmarks, the medial ridge of the distal trochlea of ​​the humerus is obtained.

6. The elbow joint rotation center axis identification method according to claim 1, characterized in that, Based on the initial rotation axis, the portion between the lateral surface of the humeral head and the medial ridge of the trochlea in the three-dimensional surface model is taken as the distal humeral region of interest for rotation. The centers of the arcs with set arc angles for each side view of the distal humeral region of interest for rotation are iteratively fitted, and the straight lines containing multiple such centers are iteratively fitted. When the fitted straight lines meet the set conditions, the fitted straight lines are used as the refinement axis, specifically including: Based on the initial rotation axis, the portion between the outer surface of the humeral head and the medial ridge of the trochlea in the three-dimensional surface model is taken as the distal humeral focus rotation region. The center of the arc with a set arc angle is fitted to the side view of the distal humeral focus rotation region to obtain the center of the side view of the distal humeral focus rotation region. A temporary axis of rotation is obtained by fitting the straight line containing the center of the circle in each side view of the region of interest in the distal humerus. Using the temporary rotation axis as the current initial rotation axis, return to the execution step "Based on the initial rotation axis, take the portion between the lateral surface of the humeral head and the medial ridge of the trochlea in the three-dimensional surface model as the humeral distal focus rotation region, and fit the center of the arc of the arc angle set for each side map of the humeral distal focus rotation region to obtain the center of the arc of each side map of the humeral distal focus rotation region", and obtain the updated rotation axis again; Determine the coordinates of the first and second updated intersection points of the updated initial rotation axis with the humeral surface of the distal humeral region of interest. Determine the coordinates of the first and second re-updated intersection points of the re-updated rotation axis with the humeral surface of the distal humeral region of interest. Calculate the distance between the first updated intersection point coordinates and the first further updated intersection point coordinates to obtain the first distance; Calculate the distance between the second updated intersection point coordinates and the second further updated intersection point coordinates to obtain the second distance; When the first distance and the second distance meet the set threshold range, the updated rotation axis is used as the thinning axis.

7. The elbow joint rotation center axis identification method according to claim 1, characterized in that, Based on the rotation angle range and contour of each lateral view of the updated distal humeral region of interest, calculate the standard deviation of the distance between the third intersection point of the thinning axis and each lateral view of the updated distal humeral region of interest, and the distance between the point on the contour of the corresponding lateral view and the third intersection point. Specifically, this includes: Sagittal sections were performed on key parts of the distal humerus in the updated region of rotation to obtain a lateral view of the distal humerus. Based on the rotation angle range and contour of each side view of the updated distal humerus focus rotation region, calculate the third intersection point of the thinning axis with each side view of the updated distal humerus focus rotation region; Based on the rotation angle range and contour of each lateral view of the updated distal humerus focus rotation region, the distance between the point on the contour of each lateral view that satisfies the rotation angle range of the corresponding lateral view and the third intersection point is calculated to obtain multiple contour distances. Calculate the weighted standard deviation of the multiple contour distances to obtain the standard deviation of the distance between the point on the contour of the side view and the third intersection point.

8. A system for recognizing the rotation center axis of an elbow joint, characterized in that, The system includes: The acquisition module is used to acquire CT images of the distal humerus; The first construction module is used to construct a three-dimensional surface model of the distal humerus based on the CT image of the distal humerus. The first fitting module is used to fit the distal humeral shaft axis of the three-dimensional surface model using the least squares method. The second construction module is used to establish the initial coordinate system of the distal humerus as specified by ISB and determine the initial rotation axis based on the fitted distal humeral shaft axis. The trochlear medial ridge determination module is used to determine the trochlear medial ridge of the three-dimensional surface model according to the initial coordinate system of the distal humerus specified by the ISB and the initial rotation axis. The second fitting module is used to iteratively fit the center of the arc with a set arc angle on each side map of the humeral distal tip rotation region according to the initial rotation axis, and iteratively fit the straight lines where multiple centers are located. When the fitted straight lines meet the set conditions, the fitted straight lines are used as the thinning axis. The third fitting module is used to fit a quadratic function to the sagittal plane corresponding to each lateral view of the non-critical distal humeral parts in the rotational region of interest to the thinning axis and the preset angle range of the critical distal humeral parts, thereby obtaining the rotation angle range of the non-critical distal humeral parts. Based on the preset angle range of the critical distal humeral parts and the rotation angle range of the non-critical distal humeral parts, an initial region of interest for the distal humeral is obtained. The rotational region of interest for the distal humeral includes the critical distal humeral parts and the non-critical distal humeral parts. The critical distal humeral parts include the trochlear region and the head of the humerus. The projection module is used to project the initial distal region of interest onto the humeral surface of the distal region of interest rotation of the humerus, so as to obtain the updated distal region of interest rotation of the humerus. The first parameter determination module is used to determine the range of optimization variables corresponding to the first intersection point of the thinning axis and the medial trochlear ridge in the updated distal humeral focus rotation area based on the radius of the medial trochlear ridge in the updated distal humeral focus rotation area. The second parameter determination module is used to determine the range of optimization variables corresponding to the second intersection point between the thinning axis and the lateral surface of the updated humeral head based on the ball diameter of the humeral head. The third parameter determination module is used to calculate the third intersection point between the thinning axis and each side view of the updated distal humeral focus rotation area, as well as the standard deviation of the distance between the point on the contour of the corresponding side view and the third intersection point, based on the rotation angle range and contour of each side view of the updated distal humeral focus rotation area. The solution module is used to calculate the solution using a genetic algorithm based on the variation range of the optimized variable corresponding to the first intersection point, the variation range of the optimized variable corresponding to the second intersection point, the rotation angle range of each side view of the updated distal humeral focus rotation region, and the standard deviation. The global optimal solution yields the elbow joint rotation center axis; where... and These are the boundary angle values ​​of the rotatable range of the i-th side view; The standard deviation of the distance values; This represents the number of slice side maps.

9. An electronic device, characterized in that, The device includes a memory and a processor, the memory being used to store a computer program, and the processor running the computer program to cause the electronic device to perform the elbow joint rotation center axis recognition method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the elbow joint rotation center axis identification method as described in any one of claims 1 to 7.

Images