Image sequence analysis method and system for postoperative displacement of orthopedic implant

Abstract

The invention discloses an image sequence analysis method and system for postoperative displacement of an orthopedic implant, and relates to the technical field of medical image analysis, and the method comprises the steps: obtaining a postoperative multi-angle two-dimensional image sequence, a follow-up visit period image sequence and preoperative three-dimensional CT data of a patient, building a standard coordinate system, carrying out the inverse transformation of a tilt matrix, correcting the coordinates of a screw, and carrying out the reconstruction of the screw. The influence of body position deviation is eliminated; a displacement vector set is generated based on the screw displacement difference value in the baseline period image and the follow-up period image, pseudo displacement and real displacement are judged through statistical analysis, screw equivalent coordinates are updated, the stability of the screw-rod system is further judged according to the screw-rod vector modulus length variation, and the stability level is output; according to the method, the real displacement and the pseudo displacement caused by the body position change can be accurately distinguished, the misjudgment caused by the body position deviation is reduced, and the accuracy of implant displacement detection is improved.

Description

Technical Field

[0001] This invention relates to the field of medical image analysis technology, specifically to a method and system for analyzing image sequences of postoperative displacement of orthopedic implants. Background Technology

[0002] With the advancement of orthopedic surgical techniques, orthopedic implants are increasingly used in spinal and pelvic surgeries. To ensure the effectiveness of implants, postoperative stability assessment is a crucial step. Traditional imaging methods, such as X-rays or CT scans, have become important means of assessing implant stability. In particular, during postoperative follow-up, two-dimensional imaging sequences are widely used to analyze changes in screw position in rod-and-screw systems to determine whether implant displacement has occurred.

[0003] Traditional methods determine implant displacement and system stability by comparing screw positions in baseline and follow-up images. However, the accuracy of these methods is often affected by image tilt, screw angle changes, and body positioning errors. During imaging, pelvic tilt can cause overlapping or distortion of screw projections, contaminating screw angle changes with positional shifts and making it difficult to accurately distinguish between true and false displacements. This is especially true when the pelvic tilt angle is large, where projection geometric distortion is more pronounced, leading to significant errors in displacement measurement and severely impacting the assessment of the stability of the screw-rod system.

[0004] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a method and system for analyzing imaging sequences of postoperative displacement of orthopedic implants.

[0006] To achieve the above objectives, the technical solution of the present invention is as follows:

[0007] In a first aspect, the present invention discloses an image sequence analysis method for postoperative displacement of orthopedic implants, comprising the following steps:

[0008] The patient's postoperative baseline two-dimensional image sequence, follow-up two-dimensional image sequence at the same angle and position, and preoperative three-dimensional CT data were acquired; the two-dimensional image sequence included the coordinates of the screw head center point, the screw connection point, and the coordinates of pelvic anatomical landmarks of the screw-rod system.

[0009] A standard coordinate system for the pelvis is established based on the preoperative three-dimensional CT data, and the pelvic tilt matrix at the current shooting angle is calculated based on the coordinates of pelvic anatomical landmarks in each two-dimensional image.

[0010] Perform an inverse tilt matrix transformation on the coordinates of the center point of the screw head to obtain the corrected screw coordinates in the standard body position;

[0011] Based on the difference in the coordinates of the correction screws at the same angle between the baseline period and the follow-up period, a set of displacement vectors for each screw at multiple angles is generated.

[0012] If the divergence of the displacement vector set exceeds a preset threshold, it is marked as a suspected pseudo displacement screw; otherwise, it is marked as a real displacement screw, and the average value of its displacement vector set is taken as the real displacement vector.

[0013] The equivalent coordinates of the screw during the follow-up period are updated based on the actual displacement vector. The change in the magnitude of the nail rod vector is calculated based on the coordinates of the nail rod connection point and the equivalent coordinates of the screw. The stability level of the nail rod system is determined based on the change in the magnitude of the nail rod vector.

[0014] Secondly, this invention discloses an image sequence analysis system for postoperative displacement of orthopedic implants, comprising:

[0015] The image data acquisition module is used to acquire the patient's postoperative baseline multi-angle two-dimensional image sequence, the follow-up period two-dimensional image sequence with the same angle and position, and the preoperative three-dimensional CT data. The two-dimensional image sequence includes the coordinates of the center point of the screw head of the screw-rod system, the coordinates of the screw-rod connection point, and the coordinates of the pelvic anatomical landmarks.

[0016] The tilt matrix calculation module is used to establish a standard coordinate system for the pelvis based on preoperative 3D CT data and to calculate the pelvic tilt matrix at the current shooting angle based on the coordinates of pelvic anatomical landmarks in the 2D image.

[0017] The calibration coordinate calculation module is used to perform an inverse tilt matrix transformation on the coordinates of the center point of the screw head to obtain the calibration screw coordinates under the standard body position.

[0018] The displacement vector generation module is used to generate a set of displacement vectors for each screw at multiple angles based on the difference in the coordinates of the correction screws at the same angle during the baseline period and the follow-up period.

[0019] The displacement judgment module is used to determine whether the divergence of the displacement vector set exceeds a preset threshold. If it does, it is marked as a suspected pseudo displacement screw; otherwise, it is marked as a real displacement screw, and the mean of the displacement vector set is used as the real displacement vector.

[0020] The stability assessment module is used to update the equivalent coordinates of the screws during the follow-up period based on the actual displacement vector, calculate the change in the magnitude of the screw vector based on the coordinates of the screw-bar connection point and the equivalent coordinates of the screw, and determine the stability level of the screw-bar system based on the change in the magnitude of the screw vector.

[0021] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0022] 1. By establishing a standard coordinate system for the pelvis and performing an inverse tilt matrix transformation on the coordinates of the center point of the screw head, the coordinate projection error caused by pelvic tilt can be effectively eliminated, improving the accuracy of screw position measurement and ensuring the reliability of displacement analysis results under multi-angle images.

[0023] 2. By analyzing the divergence of the displacement vector set and combining the standard deviation of the three-dimensional orientation angle and modulus, it is possible to accurately distinguish between the real displacement and the pseudo displacement caused by changes in body position, thereby reducing misjudgments caused by body position deviations and improving the accuracy of implant displacement detection.

[0024] 3. By performing reverse weighted reconstruction of the position-sensitive factors on the suspected pseudo-displacement screws, the artifact displacement components caused by position were eliminated, thereby accurately obtaining the true displacement vector, further improving the accuracy of displacement analysis, and providing more reliable data support for the stability assessment of the screw-bar system. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art 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.

[0026] Figure 1 This is an overall block diagram of the method in Embodiment 1 of the present invention;

[0027] Figure 2 This is a flowchart of the method according to Embodiment 1 of the present invention;

[0028] Figure 3 This is an overall block diagram of the system in Embodiment 2 of the present invention. Detailed Implementation

[0029] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the present invention without creative effort are within the scope of protection of the present invention.

[0030] Application Overview: In existing postoperative stability assessments of orthopedic implants, displacement analysis of two-dimensional X-ray imaging sequences is limited by projection geometric distortion caused by pelvic tilt. Due to the lack of effective correction for differences in pelvic spatial pose at different shooting angles, the displacement measurement of the screw head center point coordinates includes coupling noise introduced by positional offset. This error directly causes the displacement vector set to exhibit non-physical divergence in three-dimensional space, making it impossible to separate the true displacement from positional pseudo-displacement using conventional statistical methods. The change in the magnitude of the screw-rod vector output by the system exhibits systematic deviation due to coordinate reference drift, affecting the reliability of the mechanical stability assessment of the screw-rod system.

[0031] For example, in dual-angle follow-up imaging of the pelvis in both posteroanterior and oblique views, the difference in the iliac wing tilt angle causes the screw projection overlap area to expand to 3.8 mm. Direct comparison of the screw-rod connection point coordinates between the baseline and follow-up periods in the original image coordinate system introduces a position-related displacement error of 0.7-1.2 mm. When the actual screw displacement falls within this error range, the standard deviation of the three-dimensional orientation angle of the displacement vector set exceeds 25 degrees, and the standard deviation of the modulus reaches 0.9 mm, increasing the probability of false displacement misjudgment to 62%. This phenomenon is particularly pronounced in oblique images with pelvic rotation angles exceeding 15 degrees, causing the calculation error of the screw-rod vector modulus change to reach 1.8 times the original value.

[0032] To address the aforementioned issues, this application first considers how to eliminate the interference of patient tilt on screw displacement measurement. By analyzing the spatial correspondence between multi-angle image sequences and 3D CT data, it was found that establishing a standard coordinate system can unify the reference benchmark for different imaging angles. Therefore, this application proposes constructing a standard pelvic coordinate system using preoperative 3D data, and mapping the screw coordinates in images from various angles to a unified space through inverse matrix transformation, thereby eliminating the patient tilt component. Further research revealed that true displacement should exhibit vector consistency after multi-angle correction, while spurious displacement will exhibit divergent characteristics due to differences in projection geometry. Based on this, a method for distinguishing true from spurious displacement is proposed using the statistical characteristics of the displacement vector set. Finally, the system's stress state is inferred from the change in the magnitude of the screw-rod vector, establishing a stability assessment model.

[0033] Example 1:

[0034] like Figure 1-2 As shown, the image sequence analysis method for postoperative displacement of orthopedic implants includes the following steps: acquiring multi-angle two-dimensional image sequences at the patient's postoperative baseline, two-dimensional image sequences at the same angle and position during the follow-up period, and preoperative three-dimensional CT data; the two-dimensional image sequences include the coordinates of the center point of the screw head, the coordinates of the screw-rod connection point, and the coordinates of pelvic anatomical landmarks of the screw-rod system.

[0035] A standard coordinate system for the pelvis is established based on the preoperative three-dimensional CT data, and the pelvic tilt matrix at the current shooting angle is calculated based on the coordinates of pelvic anatomical landmarks in each two-dimensional image.

[0036] Perform an inverse tilt matrix transformation on the coordinates of the center point of the screw head to obtain the corrected screw coordinates in the standard body position;

[0037] Based on the difference in the coordinates of the correction screws at the same angle between the baseline period and the follow-up period, a set of displacement vectors for each screw at multiple angles is generated.

[0038] If the divergence of the displacement vector set exceeds a preset threshold, it is marked as a suspected pseudo displacement screw; otherwise, it is marked as a real displacement screw, and the average value of its displacement vector set is taken as the real displacement vector.

[0039] The equivalent coordinates of the screw during the follow-up period are updated based on the actual displacement vector. The change in the magnitude of the nail rod vector is calculated based on the coordinates of the nail rod connection point and the equivalent coordinates of the screw. The stability level of the nail rod system is determined based on the change in the magnitude of the nail rod vector.

[0040] The postoperative baseline multi-angle two-dimensional imaging sequence refers to the set of two-dimensional image data acquired from different shooting angles in the early postoperative period. This can be achieved by using an X-ray machine at different rotation angles to establish an initial spatial distribution benchmark for screw positions. The follow-up period two-dimensional imaging sequence with the same angle and position refers to the image sequence acquired during subsequent follow-up examinations at the same shooting angle and patient position as at the baseline period. This can be achieved by controlling the spatial position of the X-ray source and detector using a robotic arm positioning system to ensure consistency in the two image acquisition conditions and eliminate measurement errors caused by positional differences. Preoperative three-dimensional CT data refers to a three-dimensional model of the patient's pelvis generated using computed tomography (CT) technology. This can be achieved using medical image reconstruction algorithms to provide anatomical structural benchmarks for establishing a standard coordinate system. The pelvic standard coordinate system is a unified spatial reference system established based on pelvic anatomical landmarks. This can be achieved by selecting the pubic symphysis and bilateral anterior superior iliac spines as reference points and defining orthogonal coordinate axes to eliminate coordinate system deviations caused by pelvic tilt at different shooting angles. The pelvic tilt matrix is ​​a mathematical matrix describing the spatial transformation between the actual and standard pelvic positions under two-dimensional imaging angles. It is specifically generated by fitting the projection deviation of anatomical landmarks onto a reference plane using the least squares method, and is used to transform coordinates in images from different angles to a unified reference system. The inverse tilt matrix transformation is a mathematical operation that maps the screw head center point coordinates from the tilted position to the standard position. This is achieved using homogeneous coordinate transformation and inverse matrix multiplication, and is used to correct screw projection position offsets caused by pelvic tilt. The corrected screw coordinates are the three-dimensional coordinates of the screw in the standard position obtained after the inverse transformation. This is generated through matrix multiplication and is used to establish spatial position comparability in images from different angles. The displacement vector set is the set of displacement vectors calculated for the same screw under different imaging angles. It is generated by the difference between the corrected coordinates at the same angle during the baseline and follow-up periods, and is used to analyze position-independent true displacement patterns. The divergence of the displacement vector set refers to the statistical dispersion of the displacement vector in terms of direction and magnitude. This is specifically achieved by calculating the weighted sum of the standard deviations of the three-dimensional direction angles and the standard deviations of the magnitude, and is used to identify spurious displacement data caused by residual errors in body position or projection overlap. The change in the magnitude of the nail-bar vector refers to the absolute value of the change in vector length between the nail-bar connection point and the equivalent coordinates of the screw. This is specifically achieved by constructing vectors for the baseline and follow-up periods and calculating the difference in magnitude, and is used to quantify the degree of deformation of the nail-bar system.

[0041] This application further proposes the following steps for calculating the pelvic tilt matrix: extracting the sagittal and coronal reference planes of the pelvis from preoperative three-dimensional CT data; mapping the anterior superior iliac spine and pubic symphysis anatomical landmarks in each two-dimensional image to the reference planes; and generating the tilt matrix by fitting the spatial angle between the current projection plane and the reference plane using the least squares method.

[0042] The extraction of the sagittal and coronal reference planes is achieved through the geometric features of the pelvic symmetry axis in 3D CT data. The mapping of anatomical landmarks such as the anterior superior iliac spine and the pubic symphysis is performed using 3D coordinate projection transformation. The least squares fitting process is completed by solving for the minimum sum of squared residuals of the angle between the normal vectors of the projection plane and the reference plane. The combination of multi-dimensional mapping of anatomical landmarks and mathematical optimization of plane fitting can effectively eliminate plane tilt errors caused by body position deviations.

[0043] Specifically, the extraction of the reference plane provides a geometric reference system for the subsequent calculation of the spatial relationship of the projection plane, and the mapping of anatomical landmarks of the anterior superior iliac spine and the pubic symphysis ensures the spatial consistency between 2D images and 3D CT data. By using the least squares method to optimize the coordinates of multiple landmarks as a whole, the impact of single landmark positioning errors on the tilt matrix can be reduced. For example, when the projected coordinates of the anterior superior iliac spine in a 2D image experience millimeter-level shifts due to body tilt, the least squares method, through weighted calculation of multiple landmarks, controls the overall plane fitting error to within 0.5 degrees. The resulting tilt matrix accurately reflects the spatial pose changes of the pelvis under the actual imaging angle, providing reliable geometric correction parameters for the subsequent inverse transformation of screw coordinates.

[0044] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0045] The sagittal and coronal reference planes of the pelvis were extracted from preoperative 3D CT data. Specifically, pelvic structures were identified using segmentation algorithms, and anatomical landmarks such as the anterior superior iliac spine and pubic symphysis were extracted. Based on these landmarks, the sagittal and coronal planes of the pelvis were fitted as the reference planes of the standard coordinate system.

[0046] Anatomical landmarks of the anterior superior iliac spine and pubic symphysis in each 2D image are mapped to a reference plane. Further, feature points are labeled for each 2D X-ray image to obtain the 2D coordinates of the anterior superior iliac spine and pubic symphysis. These 2D coordinates are then mapped onto a reference plane in 3D space through projection transformation.

[0047] The tilt matrix is ​​generated by fitting the spatial angle between the current projection plane and the reference plane using the least squares method. Specifically, the least squares optimization objective function is established using the spatial relationship between the mapped 3D coordinate points and their corresponding points on the reference plane. This leads to the solution of the rotation matrix, i.e., the tilt matrix, between the projection plane and the reference plane.

[0048] Through the above technical solution, this application can accurately calculate the tilt angle of the pelvis, thereby eliminating the influence of positional changes on screw displacement measurement. This improves the accuracy of stability assessment of the screw-rod system and provides reliable technical support for clinical follow-up.

[0049] This application further proposes that the coordinates of the correction screw be obtained by multiplying the inverse of the tilt matrix with the coordinates of the center point of the screw head, wherein the matrix multiplication adopts a homogeneous coordinate transformation.

[0050] Homogeneous coordinate transformation constructs a four-dimensional coordinate space by adding a fourth constant term to the three-dimensional coordinate system, unifying spatial rotation and translation operations into a four-dimensional matrix form. During matrix multiplication, the linear combination property of the four-dimensional homogeneous coordinates can completely preserve the geometric transformation relationships of the original coordinate system, eliminating the cumulative errors caused by the separate calculation of rotation and translation components in conventional three-dimensional matrix multiplication.

[0051] Specifically, during the inverse transformation of the tilt matrix, the coordinates of the screw head center point are expanded into a four-dimensional homogeneous coordinate form, with the fourth dimension set to a fixed value of 1. The inverse matrix of the tilt matrix is ​​constructed as a four-dimensional square matrix, with the first three rows and three columns containing rotation components, and the first three rows and fourth column containing translation components. After matrix multiplication, homogeneous coordinate normalization is used to restore the four-dimensional coordinates to three-dimensional physical coordinates. For example, when the original coordinates of the screw head center point are (x, y, z), its homogeneous coordinate form is (x, y, z, 1), which, after multiplying by the four-dimensional inverse matrix, yields (x', y', z', w), and the final corrected coordinates are (x' / w, y' / w, z' / w). This method can accurately compensate for coordinate projection distortion caused by the tilt of the shooting position, ensuring the spatial consistency of the corrected screw coordinates at different angles.

[0052] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0053] The process of inverse transformation of the tilt matrix includes: correcting screw coordinates = inverse of the tilt matrix × coordinates of the center point of the screw head; where the matrix multiplication adopts homogeneous coordinate transformation.

[0054] In practice, the inverse of the tilt matrix is ​​first obtained. The tilt matrix is ​​a 3x3 matrix, and its inverse is obtained through matrix inversion. Then, the coordinates of the screw head center point are represented in homogeneous coordinate form, that is, by adding a fourth component 1 to the original three-dimensional coordinates (x, y, z), resulting in (x, y, z, 1). Next, matrix multiplication is performed, multiplying the inverse of the tilt matrix by the homogeneous coordinates of the screw head center point. Finally, the first three components of the result are used as the corrected screw coordinates.

[0055] Through the above technical solution, this application achieves precise correction of the coordinates of the screw head center point. By using the inverse transformation of the tilt matrix, the projection error caused by pelvic tilt imaging is eliminated, and the corrected screw coordinates under standard body position are obtained. This correction method improves the accuracy of subsequent displacement analysis and lays the foundation for accurately evaluating the stability of the screw-rod system.

[0056] This application further proposes a step for judging the degree of divergence of the displacement vector set, including: calculating the standard deviation of the three-dimensional orientation angle and the standard deviation of the magnitude of the displacement vector set; when the sum of the product of the standard deviation of the angle and the weighting coefficient of the magnitude and the standard deviation of the magnitude is greater than a preset threshold, it is judged as divergence.

[0057] The standard deviation of the three-dimensional orientation angle is obtained by calculating the difference in orientation angles of each displacement vector in three-dimensional space, and is used to quantify directional dispersion; the standard deviation of the magnitude is used to measure the degree of fluctuation in the length of the displacement vector. The magnitude weighting coefficient is preset based on the contribution of orientation and magnitude to pseudo-displacement in clinical data, and is used to adjust the weight ratio of the two in the judgment. The preset threshold is determined by statistically analyzing the differences in the distribution of true displacement and pseudo-displacement in different cases.

[0058] Specifically, when calculating the standard deviation of the three-dimensional orientation angles of the displacement vector set, the orientation angle of each vector needs to be decomposed into azimuth and pitch angles, and the root mean square value is taken after calculating the standard deviation of each. The standard deviation of the modulus length is calculated by the deviation of the modulus length of each vector from the average modulus length. The standard deviations of the orientation angles and the standard deviations of the modulus length are linearly combined with weighted coefficients and compared with a preset threshold. For example, when the modulus length weighting coefficient is set to 0.5, the standard deviation of the orientation angles plays a dominant role, which is suitable for orientation-sensitive pseudo-displacement scenarios; when the weighting coefficient is adjusted to 0.8, the influence of modulus length changes is enhanced, which is suitable for modulus length abnormalities caused by screw loosening. By dynamically adjusting the weighting coefficients and thresholds, different clinical needs can be adapted to improve the accuracy of judgment.

[0059] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0060] The steps for determining the degree of divergence of a displacement vector set include calculating the standard deviation of the three-dimensional orientation angles and the standard deviation of the modulus. When (angular standard deviation + modulus weighting coefficient * modulus standard deviation) > a preset threshold, it is determined to be divergent.

[0061] Specifically, first, the three-dimensional orientation angles of each vector in the displacement vector set are calculated. For each vector, the angles with the x, y, and z axes are calculated, yielding three angle values. Then, the standard deviation of these angle values ​​is calculated, yielding the angular standard deviation. Simultaneously, the standard deviation of the magnitudes of each vector in the displacement vector set is calculated, yielding the magnitude standard deviation.

[0062] Furthermore, a modulus weighting coefficient α is introduced, which is used to balance the importance of the two indicators, angle and modulus. The value of α ranges from 0 to 1, and the closer α is to 1, the greater the influence of the modulus standard deviation.

[0063] Finally, the standard deviation of the angle is added to the weighted standard deviation of the modulus and compared with a preset threshold. If (standard deviation of angle + α * standard deviation of modulus) > preset threshold, the displacement vector set is determined to be divergent, and the corresponding screw is marked as a suspected pseudo-displacement screw. Otherwise, it is determined to be non-divergent, and the corresponding screw is marked as a true displacement screw.

[0064] For example, setting α=0.5 and β=10°, for a certain screw, the angular standard deviation of its displacement vector set is 8° and the modulus standard deviation is 5mm. Calculation shows that 8° + 0.5 * 5mm = 10.5° is greater than 10°, therefore the screw is judged to be a suspected pseudo-displacement screw.

[0065] Through the above technical solution, this application can effectively distinguish between true displacement and pseudo displacement caused by body position. Since true displacement exhibits consistent behavior at different angles, its displacement vector set shows minimal divergence. In contrast, pseudo displacement caused by body position exhibits inconsistent behavior at different angles, resulting in a greater divergence in its displacement vector set. By calculating the standard deviation of the angle and modulus of the displacement vector set and comparing it with a preset threshold, it is possible to accurately determine whether the screw has undergone true displacement. This method overcomes the shortcomings of traditional techniques in distinguishing between true and pseudo displacement, improving the accuracy of stability assessment for the screw-rod system. Furthermore, by introducing a modulus weighting coefficient, the importance of the angle and modulus indicators can be flexibly adjusted to adapt to the needs of different clinical scenarios, enhancing the applicability and reliability of the method.

[0066] This application further proposes that when a screw is marked as a suspected pseudo-displacement screw, a positional sensitivity factor is calculated based on the pelvic tilt angle corresponding to each angle, and the displacement vector set is reconstructed by reverse weighting based on the positional sensitivity factor to obtain the true displacement vector.

[0067] The body position sensitivity factor is obtained through eigenvalue decomposition of the tilt matrix, specifically represented as the absolute cosine of the pelvic tilt angle at the current shooting angle. The reverse weighted reconstruction process employs a weighted average algorithm, which uses the reciprocal of the body position sensitivity factor to perform a weighted summation of the displacement vector, suppressing the displacement component corresponding to the body position sensitivity angle.

[0068] Specifically, in the set of displacement vectors suspected of being spurious displacement screws, the pelvic tilt angle corresponding to each shooting angle is extracted through eigenvalue decomposition of the tilt matrix. The absolute value of the cosine of the tilt angle is used as the position sensitivity factor, which reflects the sensitivity of the body tilt to displacement measurement at different angles. For each displacement vector, the reciprocal of the position sensitivity factor is used as a weighting coefficient to calculate a weighted average of the displacement vectors at all angles. During the weighting process, displacement vectors corresponding to angles with smaller position sensitivity factors are given higher weights, thereby weakening the spurious displacement components introduced by projection deviation at position-sensitive angles. For example, when the pelvic tilt angle at a certain shooting angle is 30 degrees, its position sensitivity factor is cos(30°) = 0.866, and its reciprocal weight is 1.15; while when the tilt angle is 5 degrees, the position sensitivity factor is cos(5°) = 0.996, and its reciprocal weight is close to 1. Through this algorithm, the contribution of displacement data at larger body tilt angles to the final true displacement vector is effectively suppressed, achieving the separation and correction of position-related artifact displacements.

[0069] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0070] When a screw is marked as a suspected pseudo-displacement screw, a positional sensitivity factor is calculated based on the pelvic tilt angle corresponding to each angle. The positional sensitivity factor is defined as the absolute cosine of the pelvic tilt angle at the current shooting angle, where the pelvic tilt angle is obtained through eigenvalue decomposition of the tilt matrix. Based on the positional sensitivity factor, the displacement vector set is reconstructed by inverse weighting to obtain the true displacement vector.

[0071] Specifically, the pelvic tilt angle is first obtained through eigenvalue decomposition of the tilt matrix. For example, for a 3x3 tilt matrix R, three eigenvalues ​​λ1, λ2, and λ3 can be obtained by solving the characteristic equation |R-λI|=0. The eigenvectors corresponding to these eigenvalues ​​are the three principal axis directions of the pelvis. The pelvic tilt angle can be calculated by the angles between these eigenvectors and the axes of the standard coordinate system.

[0072] Furthermore, the position sensitivity factor α = |cos(θ)| is calculated, where θ is the pelvic tilt angle. The position sensitivity factor reflects the degree to which pelvic tilt affects screw displacement measurement at the current imaging angle. The closer the α value is to 1, the closer the imaging angle is to the standard position, and the more reliable the measurement results.

[0073] Therefore, for each shooting angle k, a body position sensitivity factor can be obtained. These factors are used to perform inverse weighted reconstruction of the displacement vector to calculate the true displacement vector. This weighting method assigns higher weight to measurement results from more reliable shooting angles (with larger α values), thereby effectively reducing the influence of body position deviation on displacement measurement.

[0074] Through the above technical solution, this application can effectively distinguish between true displacement and spurious displacement caused by body position. By introducing a body position-sensitive factor and performing inverse weighted reconstruction, the accuracy of screw displacement measurement is improved. This method overcomes the measurement error caused by pelvic tilt in traditional two-dimensional X-ray image analysis, providing more reliable data support for clinical assessment of postoperative stability of orthopedic implants. Furthermore, this method requires no additional equipment or radiation dose and can be directly applied to existing follow-up procedures, demonstrating good practicality.

[0075] In some of the solutions described above in this application, when a screw is marked as a suspected pseudo-displacement screw, although the influence of pelvic tilt on the displacement vector at different shooting angles has been quantified using a positional sensitivity factor, it remains difficult to effectively separate position-related components from the multi-angle displacement vector set. The positional sensitivity factor can only reflect the degree of tilt at a specific angle, while the coupling effect of positional changes on the displacement vector at different angles has nonlinear characteristics. Directly using linear weighting may lead to reconstruction errors in the true displacement vector.

[0076] This application further proposes a method for subtracting the position-related component from the displacement vector of suspected pseudo-displacement screws, using the following formula:

[0077]

[0078] Where k represents different shooting angles, and n represents the number of shooting angles. For the body position sensitive factor under shooting angle k, Let k be the displacement vector at the shooting angle k. This is the actual displacement vector.

[0079] The body position sensitivity factor is defined as the absolute cosine of the pelvic tilt angle at the current shooting angle, obtained through eigenvalue decomposition of the tilt matrix. During the reverse weighted reconstruction process, inverse weighting coefficients are applied to the displacement vector set. A weighted summation is performed, and the weighting coefficients are negatively correlated with the body position sensitivity factor. For example, when the body position sensitivity factor for a certain angle is 0.2, its inverse weighting coefficient is 5, which is significantly higher than the inverse weighting coefficient of 1.25 when the sensitivity factor is 0.8. Therefore, the displacement vector corresponding to the shooting angle with a larger body tilt angle is assigned a lower weight in the calculation, thereby suppressing the interference of body position deviation on the true displacement vector.

[0080] Specifically, when calculating the true displacement vector, the displacement vector at each shooting angle is first obtained. and corresponding postural sensitivity factors Multiply each displacement vector by the inverse weighting coefficient at that angle. The weighted displacement vector components are obtained. The weighted components of all angles are summed and then divided by the sum of the inverse weighting coefficients to achieve normalization. This calculation process correlates position-sensitive factors with the spatial distribution characteristics of the displacement vector through mathematical transformation. The introduction of inverse weights reduces the contribution of angle data with large positional tilts to the final result. For example, when a certain angle causes a large deviation between the pelvic projection plane and the standard coordinate system due to the patient's positional tilt, its... When the value is small, the inverse weighting coefficient increases, thereby weakening the spurious displacement component caused by body position deviation at that angle during the weighting process. Through inverse weighted fusion of multi-angle data, the final output true displacement vector can effectively eliminate systematic errors caused by body position changes and improve the accuracy of stability assessment of the spike-and-bar system.

[0081] Through the above technical solution, this application can effectively eliminate the influence of spurious displacement caused by changes in body position, and improve the accuracy of screw true displacement measurement. This reliably distinguishes between true displacement and positional spurious displacement, providing more accurate data support for the stability assessment of the screw-bar system.

[0082] This application further proposes a process for updating the equivalent coordinates of the screw during the follow-up period, which includes: the equivalent coordinates of the screw equal to the baseline-corrected screw coordinates plus the actual displacement vector.

[0083] The baseline correction screw coordinates are obtained under standard body position through inverse tilt matrix transformation, while the true displacement vector is the mean of the displacement vector set. Since the baseline correction screw coordinates and the true displacement vector belong to the same coordinate system, they are directly added together. Using the baseline correction screw coordinates as a reference point and the true displacement vector as a rigid displacement, their superposition yields the equivalent coordinates of the follow-up screw under standard body position.

[0084] Specifically, the baseline screw coordinates are located using a standard pelvic coordinate system established from preoperative 3D CT data, and the true displacement vector is determined by the mean of the displacement vector set. In the standard coordinate system, when the true displacement vector is applied to the baseline screw coordinates, the equivalent screw coordinates during the follow-up period are directly obtained through vector addition. This process avoids introducing additional coordinate system transformation errors, ensuring that the equivalent coordinates and the screw-rod connection point coordinates are in the same spatial reference system. The calculation of the modulus change of the screw-rod vector constructed based on the equivalent coordinates and the screw-rod connection point coordinates directly reflects the degree of deformation of the screw-rod system, thus providing an accurate data basis for stability assessment.

[0085] As a preferred embodiment, the solution of this application is specifically implemented as follows:

[0086] The process of updating the equivalent coordinates of screws during the follow-up period involves calculating the equivalent coordinates. Specifically, the equivalent coordinates are obtained by adding the baseline-corrected screw coordinates to the actual displacement vector. For example, for a screw with baseline-corrected screw coordinates of (10, 20, 30) mm and actual displacement vector of (1, -2, 3) mm, the equivalent coordinates during the follow-up period can be calculated as (11, 18, 33) mm. Furthermore, this calculation process can be applied to every screw in the screw-bar system to obtain the equivalent coordinates of all screws during the follow-up period.

[0087] Through the above technical solution, this application can accurately update the spatial position information of the screws during the follow-up period. This eliminates the influence of changes in body position on screw position measurement and improves the accuracy of screw displacement analysis. Furthermore, this method provides a reliable data foundation for subsequent stability assessment of the screw-rod system, helping clinicians to more accurately determine the mechanical stability status of the implant.

[0088] This application further proposes a calculation process for the change in the magnitude of the nail rod vector, including: constructing the baseline period nail rod vector and the follow-up period nail rod vector based on the coordinates of the nail rod connection point and the equivalent coordinates of the screw; calculating the absolute difference between the magnitude of the baseline period nail rod vector and the magnitude of the follow-up period nail rod vector corresponding to the same nail rod connection point to obtain the change in the magnitude of the nail rod vector; the construction of the nail rod vector includes: baseline period nail rod vector = baseline period nail rod connection point coordinates - baseline period correction screw coordinates; follow-up period nail rod vector = follow-up period nail rod connection point coordinates - follow-up period screw equivalent coordinates.

[0089] Specifically, when constructing the baseline and follow-up period nail-bar vectors, the spatial position changes of the nail-bar connection points are transformed into vector relationships by calculating the difference between the coordinates of the nail-bar connection points and the equivalent coordinates of the screws. The absolute difference is calculated using algebraic operations on the vector magnitude to eliminate the interference of directional differences on the deformation. The construction of the baseline nail-bar vector is based on the corrected screw coordinates to ensure that the body position deviation has been eliminated. The follow-up period nail-bar vector uses the screw position updated with equivalent coordinates to reflect the nail-bar connection status after the actual displacement.

[0090] Specifically, in the construction of the screw-bar vectors during the baseline and follow-up periods, the baseline screw-bar vector is first generated by calculating the difference between the corrected screw coordinates and the screw-bar connection point coordinates during the baseline period. This vector represents the initial relative positional relationship between the screw head and the screw-bar connection point in the standard body position. The follow-up screw-bar vector is generated by calculating the difference between the follow-up screw-bar connection point coordinates and the updated equivalent screw coordinates. The equivalent screw coordinates have been superimposed with the actual displacement vector, thus eliminating the influence of body position offset. Subsequently, the modulus of the baseline and follow-up screw-bar vectors corresponding to the same screw-bar connection point is calculated, and the absolute difference is taken as the modulus change. This method avoids the interference of vector direction offset caused by screw angle changes on deformation calculation by comparing the absolute values ​​of the vector modulus, thereby accurately reflecting the axial expansion and contraction deformation of the screw-bar connection point. For example, when the baseline screw-bar vector modulus is 15.2 mm and the follow-up modulus is 14.8 mm, the modulus change is 0.4 mm, which directly represents the degree of axial deformation of the screw-bar connection point.

[0091] Through the above technical solution, this application effectively solves the problem of misjudgment of the stability of the nail-rod system caused by body tilt. By constructing a nail-rod vector based on corrected coordinates and calculating its modulus change, the relative displacement error between the equivalent coordinates of the screw and the nail-rod connection point can be eliminated, thereby accurately reflecting the actual deformation degree of the nail-rod connection structure. This solution, by quantifying the absolute difference in the modulus of the nail-rod vector, can directly identify mechanical failures at the nail-rod connection point, providing a reliable basis for clinical judgment of the stability of the nail-rod system.

[0092] This application further proposes a process for determining the stability level of a nail-bar system based on the change in the vector magnitude of the nail bar, including: determining whether the change in the vector magnitude of the nail bar exceeds a preset elastic deformation threshold; if it does, marking the nail bar connection point as failed; otherwise, marking it as stable; calculating the proportion of failed nail bar connection points to all connection points; and outputting the stability level of the nail-bar system and the topological location map of the failed connection points based on this proportion.

[0093] Among them, the elastic deformation threshold is set through material mechanics experimental data. For example, the elastic deformation threshold of the titanium alloy nail rod system can be set to 0.5 mm. The failure ratio statistics adopt discretized interval division. For example, when the failure ratio is less than 10%, it is marked as low risk level, 10%-30% is medium risk level, and more than 30% is high risk level. The topology location map maps the spatial distribution of failure connection points through a three-dimensional coordinate system and distinguishes the risk level by color coding.

[0094] Specifically, the change in the modulus of the rod-pin vector is calculated based on the absolute difference between the rod-pin vector at baseline and during follow-up. When the change in the modulus of the rod-pin vector during follow-up relative to baseline exceeds the elastic deformation threshold, it indicates that the connection point has undergone irreversible deformation or loosening and needs to be marked as a failure. By statistically analyzing the failure rate of all connection points, the mechanical stability level of the entire rod-pin system can be quantitatively assessed. Furthermore, a topological location map generated based on three-dimensional coordinates can visually display the spatial distribution of failed connection points. For example, failure points can be highlighted in red on a pelvic model to help clinicians quickly locate areas requiring intervention. This process, combining physical threshold screening with spatial topological analysis, effectively improves the accuracy and visualization of stability assessment.

[0095] Through the above technical solution, this application can accurately distinguish between the elastic deformation and mechanical failure state of the nail-rod connection point. By statistically analyzing the proportion of failure points and their spatial distribution, it can intuitively reflect the overall stability defect area of ​​the nail-rod system, avoid misjudgment caused by local artifact interference, and improve the reliability of postoperative follow-up evaluation.

[0096] In some of the solutions described above in this application, displacement analysis of the nail-bar system based on two-dimensional image sequences requires manual operation for coordinate transformation and displacement calculation, which has problems such as low processing efficiency and susceptibility to subjective errors, making it difficult to achieve automated processing and stability assessment of multi-angle image data.

[0097] Example 2:

[0098] like Figure 3 As shown, the imaging sequence analysis system for postoperative displacement of orthopedic implants includes:

[0099] The image data acquisition module is used to acquire the patient's postoperative baseline multi-angle two-dimensional image sequence, the follow-up period two-dimensional image sequence with the same angle and position, and the preoperative three-dimensional CT data. The two-dimensional image sequence includes the coordinates of the center point of the screw head of the screw-rod system, the coordinates of the screw-rod connection point, and the coordinates of the pelvic anatomical landmarks.

[0100] The tilt matrix calculation module is used to establish a standard coordinate system for the pelvis based on preoperative 3D CT data and to calculate the pelvic tilt matrix at the current shooting angle based on the coordinates of pelvic anatomical landmarks in the 2D image.

[0101] The calibration coordinate calculation module is used to perform an inverse tilt matrix transformation on the coordinates of the center point of the screw head to obtain the calibration screw coordinates under the standard body position.

[0102] The displacement vector generation module is used to generate a set of displacement vectors for each screw at multiple angles based on the difference in the coordinates of the correction screws at the same angle during the baseline period and the follow-up period.

[0103] The displacement judgment module is used to determine whether the divergence of the displacement vector set exceeds a preset threshold. If it does, it is marked as a suspected pseudo displacement screw; otherwise, it is marked as a real displacement screw, and the mean of the displacement vector set is used as the real displacement vector.

[0104] The stability assessment module is used to update the equivalent coordinates of the screws during the follow-up period based on the actual displacement vector, calculate the change in the magnitude of the screw vector based on the coordinates of the screw-bar connection point and the equivalent coordinates of the screw, and determine the stability level of the screw-bar system based on the change in the magnitude of the screw vector.

[0105] The above description is merely an example and illustration of the structure of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described, or use similar methods to replace them, as long as they do not deviate from the structure of the invention or exceed the scope defined in the claims, all of which should fall within the protection scope of the present invention.

[0106] In the description of this specification, references to terms such as "an embodiment," "example," "specific example," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0107] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for analyzing imaging sequences of postoperative displacement of orthopedic implants, characterized in that, Includes the following steps: The patient's postoperative baseline two-dimensional image sequence, follow-up two-dimensional image sequence at the same angle and position, and preoperative three-dimensional CT data were acquired; the two-dimensional image sequence included the coordinates of the screw head center point, the screw connection point, and the coordinates of pelvic anatomical landmarks of the screw-rod system. A standard coordinate system for the pelvis is established based on the preoperative three-dimensional CT data, and the pelvic tilt matrix at the current shooting angle is calculated based on the coordinates of pelvic anatomical landmarks in each two-dimensional image. Perform an inverse tilt matrix transformation on the coordinates of the center point of the screw head to obtain the corrected screw coordinates in the standard body position; Based on the difference in the coordinates of the correction screws at the same angle between the baseline period and the follow-up period, a set of displacement vectors for each screw at multiple angles is generated. If the divergence of the displacement vector set exceeds a preset threshold, it is marked as a suspected pseudo displacement screw; otherwise, it is marked as a real displacement screw, and the average value of its displacement vector set is taken as the real displacement vector. The equivalent coordinates of the screw during the follow-up period are updated based on the actual displacement vector. The change in the magnitude of the nail rod vector is calculated based on the coordinates of the nail rod connection point and the equivalent coordinates of the screw. The stability level of the nail rod system is determined based on the change in the magnitude of the nail rod vector.

2. The image sequence analysis method for postoperative displacement of orthopedic implants according to claim 1, characterized in that: The steps for calculating the pelvic tilt matrix include: The sagittal and coronal reference planes of the pelvis were extracted from preoperative 3D CT data; Map the anterior superior iliac spine and pubic symphysis anatomical landmarks in each two-dimensional image to the reference plane; The tilt matrix is ​​generated by fitting the spatial angle between the current projection plane and the reference plane using the least squares method.

3. The image sequence analysis method for postoperative displacement of orthopedic implants according to claim 1, characterized in that: The process of inverse transformation of the tilt matrix includes: Correction screw coordinates = inverse of tilt matrix × coordinates of screw head center point; The matrix multiplication uses homogeneous coordinate transformation.

4. The image sequence analysis method for postoperative displacement of orthopedic implants according to claim 1, characterized in that: The steps for determining the degree of divergence of the displacement vector set include: Calculate the standard deviation of the three-dimensional orientation angles and the standard deviation of the modulus of the displacement vector set; When (angular standard deviation + modulus weight coefficient * modulus standard deviation) > preset threshold, it is determined to be divergent.

5. The image sequence analysis method for postoperative displacement of orthopedic implants according to claim 1, characterized in that: Also includes: When a screw is marked as a suspected pseudo-displacement screw, the position sensitivity factor is calculated based on the pelvic tilt angle corresponding to each angle. The body position sensitivity factor is defined as the absolute value of the cosine of the pelvic tilt angle at the current shooting angle, wherein the pelvic tilt angle is obtained by eigenvalue decomposition of the tilt matrix. Based on the aforementioned position-sensitive factors, the displacement vector set is reconstructed by reverse weighting to obtain the true displacement vector.

6. The image sequence analysis method for postoperative displacement of orthopedic implants according to claim 5, characterized in that: The reverse weighted reconstruction process is as follows: For screws suspected of spurious displacement, subtract the position-related component from their displacement vector using the following formula: Where k represents different shooting angles, and n represents the number of shooting angles. For the body position sensitive factor under shooting angle k, Let k be the displacement vector at the shooting angle k. This is the actual displacement vector.

7. The image sequence analysis method for postoperative displacement of orthopedic implants according to claim 1, characterized in that: The process of updating the equivalent coordinates of the screws during the follow-up period includes: Screw equivalent coordinates = baseline period corrected screw coordinates + actual displacement vector.

8. The image sequence analysis method for postoperative displacement of orthopedic implants according to claim 1, characterized in that: The calculation process for the change in the vector length of the nail rod includes: Based on the coordinates of the nail-bar connection point and the equivalent coordinates of the screw, construct the baseline nail-bar vector and the follow-up nail-bar vector; Calculate the absolute difference between the baseline period spike vector magnitude and the follow-up period spike vector magnitude at the same spike connection point to obtain the change in spike vector magnitude; The construction of the spike vector includes: Baseline period nail vector = Baseline period nail connection point coordinates - Baseline period correction screw coordinates; Follow-up period nail vector = Follow-up period nail connection point coordinates - Follow-up period screw equivalent coordinates.

9. The image sequence analysis method for postoperative displacement of orthopedic implants according to claim 1, characterized in that: The process of determining the stability level of the spike system based on the change in the vector magnitude of the spike includes: Determine whether the change in the vector magnitude of the nail rod exceeds a preset elastic deformation threshold: if yes, mark the nail rod connection point as failed; otherwise, mark the nail rod connection point as stable. Calculate the proportion of failed nail rod connection points out of all connection points, and output the stability level of the nail rod system and the topology map of the failed connection points based on this proportion.

10. An image sequence analysis system for postoperative displacement of orthopedic implants, characterized in that: The method for analyzing postoperative displacement of orthopedic implants as described in any one of claims 1-9 includes: The image data acquisition module is used to acquire the patient's postoperative baseline multi-angle two-dimensional image sequence, the follow-up period two-dimensional image sequence with the same angle and position, and the preoperative three-dimensional CT data. The two-dimensional image sequence includes the coordinates of the center point of the screw head of the screw-rod system, the coordinates of the screw-rod connection point, and the coordinates of the pelvic anatomical landmarks. The tilt matrix calculation module is used to establish a standard coordinate system for the pelvis based on preoperative 3D CT data and to calculate the pelvic tilt matrix at the current shooting angle based on the coordinates of pelvic anatomical landmarks in the 2D image. The calibration coordinate calculation module is used to perform an inverse tilt matrix transformation on the coordinates of the center point of the screw head to obtain the calibration screw coordinates under the standard body position. The displacement vector generation module is used to generate a set of displacement vectors for each screw at multiple angles based on the difference in the coordinates of the correction screws at the same angle during the baseline period and the follow-up period. The displacement judgment module is used to determine whether the divergence of the displacement vector set exceeds a preset threshold. If it does, it is marked as a suspected pseudo displacement screw; otherwise, it is marked as a real displacement screw, and the mean of the displacement vector set is used as the real displacement vector. The stability assessment module is used to update the equivalent coordinates of the screws during the follow-up period based on the actual displacement vector, calculate the change in the magnitude of the screw vector based on the coordinates of the screw-bar connection point and the equivalent coordinates of the screw, and determine the stability level of the screw-bar system based on the change in the magnitude of the screw vector.

Images