A bone pin path optimization method and an orthopedic consumable management system

Abstract

The application belongs to the technical field of path optimization, and discloses a bone nail path optimization method and an orthopedic consumable management system; the method comprises the following steps: collecting patient bone image data and performing data cleaning to obtain an orthopedic anatomy medical image; combining historical case data to perform image segmentation on the orthopedic anatomy medical image to obtain a tissue boundary medical image; constructing a three-dimensional bone model and constructing a bone nail path in the model, performing risk avoidance output on the bone nail path to obtain a risk-avoiding bone nail path; screening a load-applicable bone nail path from the risk-avoiding bone nail path; performing disturbance screening on the load-applicable bone nail path to generate a structure-stable bone nail path; performing end-closedness analysis on the structure-stable bone nail path to determine a final optimized bone nail path, and taking the final optimized bone nail path as a clinical application bone nail path scheme; the method significantly improves the mechanical reliability and biological fusion adaptability of the bone nail path scheme, and also strengthens the process avoidance mechanism for risk areas and the postoperative stability guarantee capability.

Description

Technical Field

[0001] This invention relates to the field of path optimization technology, and more specifically, to a method for optimizing bone screw paths and an orthopedic consumables management system. Background Technology

[0002] Bone screw placement surgery plays a crucial role in orthopedic diagnosis and treatment. The surgical path planning directly determines the fixation strength, implantation accuracy, and postoperative stability of the bone screw. Currently, clinical bone screw placement path design mainly relies on the surgeon's experience and two-dimensional image guidance. In actual operation, there are often problems with inappropriate path selection for complex anatomical structures or pathological bone tissues, which seriously affect postoperative recovery and surgical outcomes. The accuracy of path planning directly affects the success or failure of bone screw implantation. Therefore, how to achieve higher precision path planning and optimization is the current research focus of bone screw placement surgery. In current visualization path optimization processes, grayscale images acquired by sampling devices may show similar grayscale values ​​for different tissue structures, leading to the misidentification of hollow structures as solid bone structures. This can result in the path entering abnormal areas and causing damage. For example, the presacral venous plexus is often identified as a solid structure during structural recognition because its grayscale value is similar to that of cancellous bone. Consequently, the path planning may misjudge the area as suitable for bone nail placement, leading to damage such as bone nail penetration of blood vessels. Furthermore, for structurally complex areas, current technologies struggle to accurately distinguish boundary boundaries, often classifying areas with similar grayscale values ​​as the same tissue, thus mistakenly using boundary transition zones as path channels and resulting in unreasonable bone nail placement points. In terms of path planning, current technologies often lack multi-dimensional assessment of the actual placement effect, relying solely on angle to determine path feasibility. Simultaneously, current technologies tend to overlook the path traversing tissues. There is the issue of matching the physiological load capacity of the structure with the individual tissue repair capacity of the patient. For example, in adolescent or elderly patients, the development of different tissue structures varies significantly. If the conventional bone screw placement path crosses an underdeveloped area or a high-load, difficult-to-repair area, problems such as screw instability or secondary fracture displacement may occur postoperatively. Moreover, traditional methods lack consideration for the risk of bone screw resonance caused by postoperative patient movement or certain factors that easily generate vibration. For example, in vibration-sensitive populations such as elderly patients, the bone screw placement path structure is very prone to resonance amplification during long walks, ultimately leading to screw loosening or failure of osseointegration. It should be noted that existing techniques often do not consider the sealing problem at the end of the path. When the end of the path is close to key tissue structures such as the posterior bone wall or the angle of deviation from the anatomical plane of the end is too large, the tail of the screw is prone to slippage, jamming, or perforation of the synovium of the adjacent joint, thereby causing serious postoperative complications.

[0003] In view of this, the present invention proposes a bone screw path optimization method and an orthopedic consumables management system to solve the above problems. Summary of the Invention

[0004] To overcome the aforementioned deficiencies of the prior art and to achieve the above objectives, the present invention provides the following technical solution: a bone screw path optimization method, comprising: S1. Collect patient skeletal imaging data and historical case data; perform data cleaning on the patient skeletal imaging data to obtain orthopedic anatomical medical images; S2. Combine historical case data to perform image segmentation on orthopedic anatomical medical images to obtain tissue boundary medical images; S3. Construct a three-dimensional skeletal model based on medical images of tissue boundaries; S4. Construct the nail path in the 3D skeleton model, and output the risk-avoidance nail path by performing risk avoidance on the nail path; S5. Perform path load assessment on the risk avoidance nail paths, and select load-suitable nail paths based on the assessment results; perform disturbance screening on the load-suitable nail paths to generate structurally stable nail paths. S6. Perform end-closure analysis on the structurally stable bone screw path, determine the final optimized bone screw path based on the analysis results, and use the final optimized bone screw path as the bone screw path scheme for clinical application.

[0005] Furthermore, the method for performing image segmentation includes: Perform a region growing operation on any orthopedic anatomical medical image to extract the image region whose gray value is within a preset gray value range to obtain a bone tissue mask; measure the aspect ratio of the bounding box and the mean angle between the contour point normals of any closed region in the bone tissue mask, and determine the closed region whose aspect ratio of the bounding box and the mean angle between the contour point normals are both less than the preset corresponding threshold as a cavity suspected region. The bone tissue mask images of suspected cavity regions are sorted by time and stacked to form a mask image group; the centroid projection points of suspected cavity regions in each layer of the mask image group are obtained, and the centroid offset of adjacent image slices is calculated. At the same time, the number of connected image slices in the mask image group is counted. If the centroid offset is less than a preset offset threshold and the number of connected image slices is greater than a preset number threshold, the corresponding suspected cavity region is determined to be a cavity region. The cavity region is removed from the bone tissue mask image to obtain the adjusted mask image; the adjusted mask image is then delineated to output a tissue boundary medical image.

[0006] Furthermore, the method for boundary delineation includes: After positioning and adjusting the mask image, connect the closed regions between them, calculate the difference in gray values ​​between adjacent pixels within the connected regions, and select the pixels whose gray value difference is greater than a preset transition threshold as boundary candidate points; construct a dynamic sliding window to traverse the connected regions, and calculate the texture contrast within the window centered on the boundary candidate points. Extract the tissue variation probability and bone density fluctuation of the region within a window centered on a candidate boundary point from historical case data, and construct a boundary determination index by stitching together texture contrast; calculate the confidence score of the boundary determination index, and determine the candidate boundary point as the boundary determination point if the confidence score is higher than the preset confidence level threshold; define the boundary determination point whose Euclidean distance from any boundary determination point is greater than the preset radius and does not belong to the same window as the adjacent node; connect the first boundary determination point with its adjacent nodes and set the boundary direction, and connect all boundary determination points with the same boundary direction to form a boundary path; fit a continuous boundary curve based on the boundary path, and project the continuous boundary curve onto the adjusted mask to obtain a medical image of tissue boundary.

[0007] Furthermore, the method for constructing the three-dimensional skeletal model includes: A three-dimensional skeletal model is constructed based on an image set consisting of continuous medical images of tissue boundaries and the spatial coordinates of each tissue region therein, while all tissue structures in the three-dimensional skeletal model are labeled.

[0008] Furthermore, the methods for risk avoidance include: The bone screw insertion area is determined and marked in the 3D bone model. The insertion start point and insertion end point in the bone screw insertion area are selected to form a set of path start and end point pairs. The shortest 3D Euclidean distance of each pair of path start and end points is calculated to obtain the initial bone screw path set. The path axis of any initial bone screw path at the insertion start point is extracted. At the same time, the bone surface normal of the area where the initial bone screw path is located is detected. The angle between the path axis and the bone surface normal is calculated. Initial bone screw paths with an angle greater than or equal to a preset feasible angle threshold are selected as preliminary feasible bone screw paths. A fixed window is used to capture a continuous local area of ​​the cortex along the path axis of the preliminary feasible bone nail path. The minimum penetration thickness of this continuous local area of ​​the cortex is measured, and paths with a minimum penetration thickness less than a preset insertion depth threshold are eliminated. At the same time, if there is a cavity defined area in the remaining preliminary feasible bone nail paths, the corresponding path is deleted to obtain the secondary judgment bone nail path. The secondary judgment of the nail path is divided into segments based on a determined step length to obtain equal-length path segments; connectivity analysis is performed on continuous equal-length path segments to filter out visually connected nail paths; stress distribution is judged on the visually connected nail paths to output the risk avoidance nail path.

[0009] Furthermore, the method for performing connectivity analysis includes: Determine the location of the clinical surgical incision, calculate the spatial straight-line angle from the location of the clinical surgical incision to the center point of each equal-length path segment, and if the spatial straight-line angle is higher than the preset upper limit, the corresponding equal-length path segment is determined to be an obstructed path segment; determine whether the remaining equal-length path segments are in an inoperable tissue structure, if so, determine the corresponding equal-length path segment to be an obstructed path segment; if the number of consecutive obstructed path segments exceeds the preset obstruction number threshold, delete the corresponding secondary judgment bone nail path, and set the remaining paths as visible connected bone nail paths; The methods for determining stress distribution include: Extract equidistant regions on both sides of the visible connected nail path along the path axis to obtain the stress analysis region; detect the axial stress distribution and shear stress distribution in the stress analysis region, and simultaneously obtain the stress change sequence within the region; extract the peak value of axial stress and peak value of shear stress in the stress analysis region, and calculate the standard deviation of the stress change sequence. If any value exceeds the corresponding preset threshold, the corresponding visible connected nail path is considered a dangerous path and is removed, and the remaining visible connected paths, i.e., the avoidance nail paths, are output.

[0010] Furthermore, the method for performing path load assessment includes: Identify the tissue structures traversed by the path of the safety nail and construct a penetrating tissue sequence based on the arrangement order of the tissue structures; label the penetrating tissue sequence with tissue type based on the biological characteristics of the tissue structure; and divide the penetrating tissue sequence into tissue segments based on the tissue type label to obtain tissue classification subsegments. Obtain the regeneration and repair rate and load index of each tissue type corresponding to each tissue classification segment; estimate the repair cycle of each tissue classification segment based on the regeneration and repair rate to obtain the physiological repair time; identify tissue classification segments whose physiological repair time is greater than the standard repair time and whose load index is greater than the preset index threshold, and determine them as high load segments. Based on historical case data, bone age parameters of patients are obtained, and the developmental level of corresponding tissue classification segments is assessed based on these parameters. The biological metabolic response rate is calculated by combining the developmental level and the bone age parameters of the corresponding tissue classification segment. If the biological metabolic response rate is lower than the preset corresponding bone age metabolic threshold, the corresponding tissue classification segment is identified as a metabolic overload segment. High-load segments, metabolic overload segments, and continuous segments composed of adjacent high-load or metabolic overload segments are collectively referred to as unstable segments. If the length of an unstable segment in the risk avoidance nail path is higher than a preset length ratio, the risk avoidance nail path is identified as an unstable path. All risk avoidance nail paths after eliminating unstable paths are integrated to obtain the load-suitable nail path.

[0011] Furthermore, the method for performing perturbation screening includes: A small-amplitude, fixed-frequency vibration signal is added to the load-bearing bone screw path, and the vibration response signal of the local area of ​​the path is received. Based on the vibration response signal, the change in response fluctuation of the corresponding local area of ​​the path is calculated, and a vibration response curve is constructed using the change in response fluctuation. The maximum response frequency in each period of the vibration response curve is extracted, and the response change rate of that period is calculated. A vibration response function is constructed based on the maximum response frequency and the response change rate. The sum of the function values ​​of each period of the local area of ​​the path in each load-bearing bone screw path is calculated using the vibration response function. The local areas of the path whose sum of function values ​​is greater than a preset function and threshold are marked as abnormal response segments. If the frequency range of the abnormal response segment overlaps with the known range of vibration frequencies that are easily triggered after surgery in patients, the abnormal response segment is determined to be a resonance segment. Load-bearing bone screw paths with fewer resonance segments than the limit are selected as structurally stable bone screw paths.

[0012] Furthermore, the method for performing the terminal closure analysis includes: Obtain the spatial coordinates of the endpoint of the structurally stable bone screw path, and identify whether there is a critical bone tissue structure in the area where the endpoint spatial coordinates are located along the path axis. If it does not exist, the corresponding path is eliminated; otherwise, if it exists, a simulated vertical plane is set at the front end of the critical bone tissue structure. Obtain the coordinates of the intersection point of the extension line of the endpoint spatial coordinates along the path axis on the simulated vertical plane, and measure the angle of the extension line at the intersection point with the endpoint of the simulated vertical plane. If the angle of the endpoint is less than the preset closing angle threshold, the endpoint is determined to be a near-wall endpoint. Measure the normal direction of the continuous region through which the extension line passes, and calculate the change in the angle between the extension line and the normal direction of the continuous region. If the change in the angle is higher than the preset normal drift threshold, the corresponding endpoint is determined to be a deflection endpoint. If the structurally stable bone nail path also meets the condition that there is a key bone tissue structure along the path axis in the region where the endpoint spatial coordinates are located, and the endpoint is either a near-wall endpoint or a deflection endpoint, the corresponding path is determined to be an insufficiently closed path and is eliminated. The remaining structurally stable bone nail paths are determined as the final optimized bone nail paths.

[0013] An orthopedic consumables management system for implementing a bone screw path optimization method includes: The data acquisition module is used to collect patient bone imaging data and historical case data; and to clean the patient bone imaging data to obtain orthopedic anatomical medical images. The image segmentation module is used to segment orthopedic anatomical medical images by combining historical case data to obtain boundary-defined medical images. The model building module is used to construct three-dimensional skeleton models based on medical images of tissue boundaries. The risk avoidance module is used to construct the nail path in the 3D skeletal model and output the risk-avoiding nail path by performing risk avoidance on the nail path. The path selection module is used to perform path load assessment on the risk avoidance nail paths and select load-suitable nail paths based on the assessment results; it also performs disturbance screening on the load-suitable nail paths to generate structurally stable nail paths. The final path generation module is used to perform end closure analysis on the structurally stable bone screw path, determine the final optimized bone screw path based on the analysis results, and use the final optimized bone screw path as the bone screw path scheme for clinical application; the modules are connected to each other via wired and / or wireless means.

[0014] The technical effects and advantages of the bone screw path optimization method and orthopedic consumable management system of the present invention are as follows: By segmenting collected patient bone images and combining them with historical medical data, a three-dimensional bone model was constructed. Based on this model, the bone screw path was constructed and optimized, ultimately yielding a suitable path and realizing a method for optimizing bone screw insertion paths. Compared to existing methods, the location of the cavity region was determined by calculating the centroid offset and connectivity statistics using stacked mask images, achieving automatic separation of the cavity structure from the closed bone structure. Addressing the difficulty in determining the blurred grayscale tissue boundaries in traditional methods, boundary judgment indicators were constructed to improve the reliability of boundary judgment and ensure clear visibility of pathological boundaries. Paths were screened based on three aspects: entry angle, connectivity, and stress distribution, obtaining suitable paths to ensure optimization at the level of specific operational steps during clinical practice. To accommodate individual differences, bone age parameters were introduced, and... By selecting suitable bone screw placement paths for patients based on tissue type and bone age parameters, the adaptation of path structure to tissue repair speed was achieved, effectively reducing the risk of poor postoperative fusion and indirectly improving the patient's postoperative recovery speed. By adding vibration signals, the path has basic resonance immunity during the planning stage, successfully screening out segments that may have resonance effects, and selecting existing paths based on the number of resonance segments, ensuring the stability of the patient's postoperative recovery. Through analysis of the path's end closure, the reliability of the path endpoint is accurately determined, further reducing the probability of bone screws penetrating tissue structures or having large screw end mobility after surgery, indirectly improving path stability. In summary, this method significantly improves the mechanical reliability and bio-fusion adaptability of bone screw path schemes, and also strengthens the process avoidance mechanism for risk areas and the ability to ensure postoperative stability, especially in complex clinical scenarios where it has greater stability and promotional value. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of a bone screw path optimization method according to the present invention; Figure 2 This is a schematic diagram of an orthopedic consumables management system according to the present invention. Detailed Implementation

[0016] 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.

[0017] Example 1 Please see Figure 1 As shown in this embodiment, a bone screw path optimization method includes: S1. Collect patient skeletal imaging data and historical case data; perform data cleaning on the patient skeletal imaging data to obtain orthopedic anatomical medical images; S2. Combine historical case data to perform image segmentation on orthopedic anatomical medical images to obtain tissue boundary medical images; S3. Construct a three-dimensional skeletal model based on medical images of tissue boundaries; S4. Construct the nail path in the 3D skeleton model, and output the risk-avoidance nail path by performing risk avoidance on the nail path; S5. Perform path load assessment on the risk avoidance nail paths, and select load-suitable nail paths based on the assessment results; perform disturbance screening on the load-suitable nail paths to generate structurally stable nail paths. S6. Perform end-closure analysis on the structurally stable bone screw path, determine the final optimized bone screw path based on the analysis results, and use the final optimized bone screw path as the bone screw path scheme for clinical application.

[0018] In this embodiment, the patient's skeletal image data is grayscale images of various skeletal structures and corresponding spatial coordinate information collected by medical imaging equipment. Historical case data includes clinical indicators such as the patient's bone age, bone density, anatomical variations of certain tissues, and postoperative recovery. The patient's skeletal image data has been subjected to image denoising and artifact removal, and time-axis resampling of multiple frames has been performed to remove motion blur, thereby improving image resolution and clarity and generating high-quality orthopedic anatomical medical images.

[0019] Image segmentation methods include: A region growing operation is performed on any orthopedic anatomical medical image to extract image regions whose gray values ​​fall within a preset gray range, resulting in a bone tissue mask. The region growing operation is a type of image segmentation algorithm in existing technology. It involves selecting a seed point within the selected region and recursively checking whether adjacent pixels fall within the preset gray range. It should be noted that the preset gray range is a range of gray values ​​that can represent bone tissue, based on clinical experience and theoretical considerations from experts in related fields. Extracting the image regions composed of pixels that meet the requirements forms the bone tissue mask, avoiding the errors easily caused by traditional full-image thresholding segmentation.

[0020] The aspect ratio of the bounding box and the mean angle between the contour point normals of any closed region in the bone tissue mask image are measured. Closed regions whose aspect ratio and mean angle between the contour point normals are both less than preset thresholds are identified as suspected cavity regions. The aspect ratio of the bounding box refers to the aspect ratio of the smallest bounding rectangle of the closed region. The mean angle between the contour point normals is the angle between the normal vector of any contour point on the contour of the closed region and the normal vectors of adjacent contour points, and the average value is calculated. Since cavity regions such as nerve canals, blood vessel channels, or intraosseous cavities have significantly different boundary characteristics from bone structures, a small aspect ratio of the bounding box of a closed region indicates that the shape of the region is similar to a lumen structure such as a blood vessel. A small mean angle between the contour point normals indicates that the boundary of the region is relatively smooth. At this time, based on existing medical theoretical knowledge, thresholds for the aspect ratio of the bounding box and the mean angle between the contour point normals are set respectively. Closed regions that meet both threshold restrictions are initially identified as suspected cavity regions.

[0021] The bone tissue masks marking suspected cavity areas were sorted by time and stacked to form a mask group. It should be noted that, due to the natural extensibility of human anatomy, the bone tissue masks marking suspected cavity areas were sorted frame by frame according to the sampling time, and these images belonging to the same patient were stacked to form a three-dimensional image block, that is, a mask group. This mask group shows the geometric trajectory and change trend of the suspected cavity areas across multiple layers.

[0022] Obtain the centroid projection point of the suspected cavity region in each slice of the mask image group, calculate the centroid offset of adjacent slices, and count the number of connected slices in the mask image group. If the centroid offset is less than a preset offset threshold and the number of connected slices is greater than a preset number threshold, then the corresponding suspected cavity region is determined to be a cavity region.

[0023] In this embodiment, the centroid of the suspected cavity region in each slice is calculated based on geometric knowledge. The centroids of these slices are then projected into the same space, and the difference in Euclidean distance between the projection points of the centroids of adjacent slices is calculated to obtain the centroid offset in space. This value is used to reflect whether the suspected cavity region exhibits a linear extension trend in space. The number of connected slices in the statistical mask group refers to the number of consecutive slices in which the corresponding region exists. If the number is large, it indicates that the region is a real structure. It should be noted that the preset offset threshold and preset number threshold are values ​​that can be set by those skilled in the art based on historical clinical experience. When the centroid offset is less than the threshold and the number of connected slices is greater than the threshold, it indicates that the corresponding suspected cavity region can be identified as a real cavity region.

[0024] The cavity region is removed from the bone tissue mask image to obtain an adjusted mask image. Removing the cavity region makes the image more consistent with the real anatomical structure, which facilitates the intuitive display of the cavity region when building the model later. It also avoids the cavity region affecting the subsequent path optimization. The adjusted mask image is then delineated to output a medical image of tissue boundaries.

[0025] Methods for boundary delineation include: After positioning and adjusting the mask image, the connection regions between the boundaries of closed areas are calculated. The difference in gray values ​​between adjacent pixels within the connection regions is calculated, and pixels whose gray value difference is greater than a preset jump threshold are selected as boundary candidate points. In this embodiment, the boundary of each closed region is morphologically expanded, and the intersection region between the boundaries of any two closed regions is determined as the connection region. The preset jump threshold is a limit value set based on knowledge in the field of image processing and historical boundary division experience. If the difference in gray values ​​is greater than the preset jump threshold, it means that the difference is much greater than the fluctuation of background noise, indicating that there may be a gray value transition region of tissue boundary, and thus the boundary candidate points are obtained.

[0026] A dynamic sliding window is constructed to traverse the connected regions, and the texture contrast within the window centered on the boundary candidate point is calculated. The size of the dynamic sliding window can be adjusted according to specific needs. In this embodiment, the initial window size of the dynamic sliding window is set to... Pixel; Texture contrast is obtained by calculating the ratio of the maximum gray value to the minimum gray value within a window. This value is used to measure the difference in brightness between pixel blocks within the window.

[0027] The probability of tissue variation and the amount of bone density fluctuation are extracted from the window region centered on the candidate boundary point in historical case data. The texture contrast is then concatenated to construct a boundary determination index. The probability of tissue variation and the amount of bone density fluctuation refer to the statistical results obtained from expert diagnoses in historical case data. The probability of tissue variation refers to the variation probability parameter of a certain tissue structure, such as the probability distribution of typical osteoma areas and the abnormal probability of the transition region of osteochondral heterogeneity. The amount of bone density fluctuation refers to the range of bone density change in the region of the constructed dynamic sliding window. The boundary determination index is obtained by vector concatenating the probability of tissue variation and the amount of bone density fluctuation with the previously calculated texture contrast of the window.

[0028] The credibility score of the boundary determination index is calculated, and the boundary candidate points with credibility scores higher than the preset trust level threshold are determined as boundary determination points. In this embodiment, the parameters in the boundary determination index are first normalized, and then processed using an SVM support vector machine model trained on historical data corpus to output a probability score, which is the credibility score. If the credibility score is higher than the preset trust level threshold, it indicates that the boundary candidate point corresponding to the window area has a high boundary correlation, and therefore it is determined as a boundary determination point. The preset trust level threshold is a value set based on the historical experience of determining boundary determination points.

[0029] Boundary points whose Euclidean distance to any given boundary point is greater than a preset radius and which do not belong to the same window are considered as adjacent nodes. A preset radius is set around any given boundary point. This value is based on existing theoretical knowledge. Since candidate points within the same window may be concentrated in a certain area, the path is limited to the nearest neighbor points, which can easily form small loops. Therefore, different boundary points need to satisfy both the conditions of Euclidean distance being greater than the preset radius and not belonging to the same window to avoid dense distribution in local areas.

[0030] Connect the first boundary determination point with its adjacent nodes and set the boundary direction. Connect all boundary determination points with the same boundary direction to form a boundary path. Starting from the first determined boundary determination point, set a unified direction that conforms to the actual organizational boundary, such as a certain angle direction. Then connect all boundary determination points whose angle difference from the unified direction is within an acceptable range to avoid obvious backtracking or sudden changes in direction.

[0031] Based on the boundary path fitting of a continuous boundary curve, the continuous boundary curve is projected onto the adjusted mask to obtain a tissue boundary medical image. In this embodiment, the B-spline fitting algorithm is used to smoothly fuse several boundary paths to generate a smooth and continuous boundary curve. The curve is projected onto the adjusted mask for boundary segmentation and annotation, to identify the accurate boundaries between different tissues, and finally generate a tissue boundary medical image.

[0032] Methods for constructing 3D skeletal models include: A three-dimensional skeletal model is constructed based on an image set consisting of continuous medical images of tissue boundaries and the spatial coordinates of each tissue region within it. Simultaneously, all tissue structures within this three-dimensional skeletal model are labeled. In this embodiment, sequentially acquired medical images of tissue boundaries are used as the basic data. Based on the basic data and the spatial coordinate information of tissue structures in the images, a voxel reconstruction algorithm is used to construct a three-dimensional structure, resulting in a three-dimensional skeletal model. The boundaries of the tissue structures are divided and labeled based on the continuous boundary curves within the model. Furthermore, based on knowledge of human anatomy, the corresponding tissue structures in each region are identified and labeled.

[0033] Risk avoidance methods include: The bone screw insertion area is determined and marked in the 3D bone model. The insertion start point and insertion end point in the bone screw insertion area are selected to form a set of path start and end point pairs. The area where bone screws need to be inserted in the 3D bone model is determined based on the original medical plan and is marked. Subsequent operations are performed in this area to prevent cross-area situations. Multiple alternative insertion start points and insertion end points are arranged in the bone screw insertion area according to medical needs. Each placement of the insertion start point and insertion end point forms a path start and end point pair. All path start and end point pairs are integrated to obtain a set of path start and end point pairs.

[0034] Calculate the shortest three-dimensional Euclidean distance between the first and last points of each path to obtain the initial set of nail paths. The calculation of the shortest three-dimensional Euclidean distance is to extract the shortest spatial straight line between the first and last points as a candidate nail path. Integrate all candidate nail paths to obtain the initial set of nail paths.

[0035] Extract the path axis of any initial bone screw path at the insertion start point, and simultaneously detect the bone surface normal in the area where the initial bone screw path is located. Calculate the angle between the path axis and the bone surface normal. Select initial bone screw paths whose angle is greater than or equal to a preset feasible angle threshold as preliminary feasible bone screw paths. Here, the path axis refers to a direction vector used to represent the opening direction of the path; the bone surface normal refers to the normal vector of the bone tissue surface at the insertion start point of the initial bone screw path in the 3D bone model. Calculate the angle between the opening direction and the bone surface normal, and set a feasible angle threshold based on expert experience and existing medical theories. When the angle value is greater than or equal to the feasible angle threshold, it indicates that the path has sufficient insertion angle space and can be selected as a preliminary feasible bone screw path.

[0036] A fixed window is used to extract a continuous local region of the cortical bone along the path axis of a preliminary feasible bone nail path. The minimum penetration thickness of this continuous local region is measured, and paths with a minimum penetration thickness less than a preset insertion depth threshold are discarded. In this embodiment, the window size of the fixed window is set to... A window is used to slide along the path axis and record the parameter changes of the continuous cortical bone region it passes through. On each local region slice, starting from the path start point, the distance between the ray and the entry and exit points of the cortical bone surface is calculated along the path axis. The minimum distance is the minimum penetration thickness. The preset penetration depth threshold is a value set based on existing medical theory to ensure a reasonable bone screw insertion depth. The minimum penetration thickness is obtained by measurement. If this value is less than the preset penetration depth threshold, it means that the corresponding path is not suitable for bone screw insertion.

[0037] Meanwhile, if a cavity-defined area exists in the other preliminary feasible bone screw paths, the corresponding path is deleted to obtain a secondary judgment bone screw path. In addition to the penetration thickness, it is also necessary to check whether there is a cavity-defined area in the area traversed by the path to avoid the risk of the cavity-defined area being penetrated due to bone screw placement. Therefore, the path with a cavity-defined area is eliminated. The remaining path obtained after judging the penetration thickness and the existence of the cavity-defined area is the secondary judgment bone screw path.

[0038] The secondary judgment nail path is divided into segments based on a determined step length to obtain equal-length path segments. The secondary judgment nail path is further divided into several continuous equal-length path segments according to a fixed-length step length, which are used for subsequent connectivity analysis and stress distribution judgment. Connectivity analysis is performed on the continuous equal-length path segments to filter out visible connected nail paths. Stress distribution judgment is performed on the visible connected nail paths to output the avoidance nail path.

[0039] Methods for performing connectivity analysis include: The clinical surgical incision location is determined, and the spatial angle between the incision location and the center point of each equal-length path segment is calculated. If the spatial angle is higher than a preset upper limit, the corresponding equal-length path segment is considered an obstructing path segment. The incision location for the surgery is determined based on medical needs, and this incision location is connected to the center point of each equal-length path segment. The angle between the connecting line and the horizontal plane along the path axis is calculated, which is the spatial angle. If the spatial angle between the incision location and a certain equal-length path segment is greater than the preset upper limit, it indicates that the path segment is difficult to see clearly from the perspective of the clinical surgeon, and it is therefore considered an obstructing path segment. The preset upper limit is an angle threshold determined based on clinical theoretical knowledge.

[0040] The system determines whether the remaining equal-length path segments are located within unoperable tissue structures. If so, the corresponding equal-length path segment is identified as an occluded path segment. If the number of consecutive occluded path segments exceeds a preset occlusion threshold, the corresponding secondary-judgment bone screw path is deleted, and the remaining paths are set as visible connected bone screw paths. To further filter occluded path segments, the system checks whether the equal-length path segments previously identified as occluded are located in areas marked as unoperable tissue structures in the 3D bone model. Unoperable tissue structures refer to areas unsuitable for bone screw placement, such as articular cartilage surfaces or nerve channels. If a path segment is located in such an unoperable tissue structure, the corresponding equal-length path segment is also identified as an occluded path segment. A preset occlusion threshold is set based on historical experience. If the number of consecutively occluded path segments exceeds this threshold, it indicates that a large section of the path is unsuitable for surgical operation, and therefore the corresponding path is deleted.

[0041] Methods for determining stress distribution include: Extract equidistant regions on both sides of the visible connected bone nail path along the path axis to obtain the stress analysis region. This region is formed by extending the path axis to both sides by a preset distance, creating an approximate cuboid that surrounds the original visible connected bone nail path. This region is used as the stress analysis region. For example, using the path axis as a baseline, extending it to both sides by the same distance yields approximate cuboids with the same length, width, and height as the visible connected bone nail path. The extended distance is the width of the approximate cuboid.

[0042] The axial stress distribution and shear stress distribution in the stress analysis region are detected, and the stress change sequence within this region is obtained. In this embodiment, the finite element method is used to calculate two stress variables in the stress analysis region, including axial stress distribution and shear stress distribution. Axial stress distribution refers to the normal stress value at each discrete point under compression along the path axis, reflecting the stress state of bone tissue resisting the entry of bone screws. Shear stress distribution refers to the shear stress value acting in the cross-sectional direction of the path, reflecting the tendency of bone tissue to rebound or fracture due to transverse shear. The axial stress value and shear stress value at discrete points in each stress analysis region are sorted according to the path axis direction to form a stress change sequence including axial stress change and shear stress change.

[0043] The peak values ​​of axial stress and shear stress in the stress analysis region are extracted, and the standard deviation of the stress change sequence is calculated. If any value exceeds the corresponding preset threshold, the corresponding visible connected bone nail path is considered a dangerous path and is removed. The remaining visible connected paths, i.e., the avoidance bone nail paths, are output. The peak value of axial stress represents the maximum value of the axial stress in the path, and the peak value of shear stress represents the maximum value of the shear stress in the path. The standard deviation of axial stress and shear stress in the stress change sequence are calculated. The preset thresholds include the maximum value limit threshold of axial stress, the maximum value limit threshold of shear stress, and the standard deviation threshold of the corresponding axial stress and shear stress. Based on medical theoretical knowledge, the maximum value limit threshold of axial stress, the maximum value limit threshold of shear stress, and the standard deviation threshold of the corresponding axial stress and shear stress are set respectively. If any parameter exceeds the preset threshold, it means that the corresponding path is unsafe from a mechanical point of view and is therefore deleted.

[0044] Methods for performing path load assessment include: The system identifies the tissue structures traversed by the path of the safety nail and constructs a penetrating tissue sequence based on the arrangement of these tissue structures. Specifically, by performing spatial intersection detection between each safety nail path and the tissue structures in the 3D bone model, the system identifies which tissue structures the safety nail path passes through. Furthermore, it constructs a penetrating tissue sequence that reflects the tissue trajectory of the structures traversed by the safety nail path according to the arrangement of the tissue structures along the path axis.

[0045] Based on the biological characteristics of tissue structures, the penetrating tissue sequence is labeled with tissue type. Based on the tissue type label, the penetrating tissue sequence is divided into tissue segments to obtain tissue classification sub-segments. The biological characteristics refer to the relevant parameter information of each tissue structure in the modern biological field, such as osteogenic activity, load-bearing capacity, and blood supply density. Based on the biological characteristics and combined with existing medical knowledge, different tissue structures in the penetrating tissue sequence are labeled with tissue type, including but not limited to tissue types such as cortical bone, cancellous bone, osteoma tissue, and proliferative new bone. Based on the tissue type label, multiple tissue segments are divided to obtain tissue classification sub-segments.

[0046] The regeneration and repair rate and load index of the tissue type corresponding to each tissue classification segment are obtained by querying existing medical and biological data. The regeneration and repair rate refers to the self-healing speed of the tissue per unit time, and the load index refers to the evaluation value of the tissue's maximum physiological tolerance.

[0047] The repair cycle is estimated for each tissue subsegment based on the regeneration and repair rate to obtain the physiological repair time. The formula for calculating the repair cycle estimation is as follows: ;in, The physiological repair time is expressed in months in this embodiment; This represents the length of any organizational classification sub-segment, expressed in this embodiment as " " as the unit; The regeneration and repair rate is expressed as "in this embodiment". " as the unit.

[0048] Tissue classification segments that have a physiological repair time greater than the standard repair time and a load index greater than a preset index threshold are identified as high-load segments. The standard repair time refers to the normal tissue repair time determined based on theoretical knowledge in the fields of biology and medicine. The preset index threshold refers to the normal load index value of the corresponding tissue type in the corresponding tissue classification segment at the theoretical level.

[0049] The bone age parameters of patients are obtained based on historical case data. The development level of the corresponding tissue classification sub-segment is assessed based on the bone age parameters of patients. Specifically, by using the patient's bone age as the basis for judgment, the existing human bone development database is queried to obtain a set of relevant developmental parameters that match the current patient's bone age. These parameters include, for example, developmental stage, bone structure continuity index (TCI), bone mineral density maturity (BMD), and basal metabolic response level. This set of developmental parameters is the developmental level.

[0050] The biological metabolic response rate is calculated by combining the developmental level and the bone age parameters of the corresponding tissue classification sub-segment. If the biological metabolic response rate is lower than the preset corresponding bone age metabolic threshold, the corresponding tissue classification sub-segment is identified as a metabolic overload segment.

[0051] In this embodiment, a feedforward neural network model pre-trained based on historical patient diagnostic data is used. Relevant parameters of developmental level and the patient's bone age are used as model inputs to calculate the biometabolic response rate of the tissue sub-segment. The biometabolic response rate is a functional indicator reflecting the intensity of the metabolic response of a tissue region to stimuli. Simultaneously, based on existing medical knowledge, a standard metabolic threshold is determined for each tissue sub-segment corresponding to the tissue type under the current patient's bone age, and a corresponding bone age metabolic threshold is set. If the biometabolic response rate of a certain sub-segment is lower than the preset corresponding bone age metabolic threshold, it indicates that the tissue in that sub-segment cannot complete self-repair through normal metabolic mechanisms and is identified as a metabolic overload segment.

[0052] High-load segments, metabolic overload segments, and continuous segments composed of adjacent high-load or metabolic overload segments are collectively referred to as unstable segments. If the length of an unstable segment in a risk-avoidance nail path exceeds a preset length ratio, the risk-avoidance nail path is determined to be an unstable path. All risk-avoidance nail paths after eliminating unstable paths are integrated to obtain a load-suitable nail path. This path can be composed of adjacent high-load segments, adjacent metabolic overload segments, or adjacent high-load and metabolic overload segments. If the total length of unstable segments exceeds a preset length ratio set based on biological and medical knowledge, the entire path is determined to have a high risk and is therefore considered an unstable path.

[0053] Methods for perturbation screening include: A small-amplitude, fixed-frequency vibration signal is added to the load-bearing bone screw path, and the vibration response signal of the local area of ​​the path is received. In order to simulate the slight disturbance caused by patient movement or impact after surgery, in this embodiment, a vibration signal of a preset frequency is applied to each complete load-bearing bone screw path, and the vibration response signal of the local area of ​​the path divided by a fixed step size is received. This signal includes several physical parameters, such as displacement, acceleration and force distribution at the nodes in the area, to reflect the real physical response of the local area of ​​the path to external stimuli.

[0054] The change in response fluctuation in the local area of ​​the corresponding path is calculated based on the vibration response signal. The vibration response curve is constructed using the change in response fluctuation. This is achieved by extracting the timestamp sequence corresponding to the physical parameters of the response from the vibration response signal and calculating the change in the value of each parameter between adjacent time points. The change in response fluctuation is then plotted with time as the horizontal axis and the amplitude of the change in response fluctuation as the vertical axis. The vibration response curve of each physical parameter as a function of time is then drawn.

[0055] The maximum response frequency of each period in the vibration response curve is extracted, and the response rate of change for that period is calculated. A vibration response function is constructed based on the maximum response frequency and the response rate of change. In this embodiment, the main peak vibration frequency of each vibration period is extracted from the vibration response curve using Fourier transform as the maximum response frequency. The response rate of change is obtained by calculating the average of the slopes of the response fluctuation amplitude changes per unit time in the vibration period. The formula for calculating the vibration response function is as follows: ;in, Indicates the first Vibration response function value for each cycle; This indicates the frequency of the small, fixed-frequency vibration signal added to the path; Indicates the first The maximum response frequency per cycle; Indicates the first The rate of change of response over one cycle; and They represent and The weights are assigned, and these weights can be dynamically adjusted based on specific circumstances. In this embodiment, the initial value is set to... , It should be noted that the above calculation formula is a dimensionless calculation.

[0056] The sum of periodic function values ​​in a local area of ​​the path for each applicable bone screw under load is calculated using the vibration response function. Local areas where the sum of function values ​​is greater than a preset function and threshold are marked as abnormal response segments. The sum of periodic function values ​​is used to reflect the total disturbance response intensity of the corresponding local area. The function and threshold are set based on historical experience. If the sum of function values ​​is greater than the preset function and threshold, it indicates that the corresponding area may have an over-response situation, and is therefore determined to be an abnormal response area.

[0057] If the frequency range of the abnormal response segment overlaps with the known frequency range of easily triggered vibrations after surgery, the abnormal response segment is identified as a resonance segment. The frequency range of easily triggered vibrations after surgery is set based on the hospital's historical patient recovery records. The frequency range of the abnormal response segment is detected. If the two frequency ranges completely overlap, it indicates that the corresponding segment may have a high-risk resonance phenomenon, and therefore the corresponding segment is identified as a resonance segment.

[0058] Load-bearing bone screw paths with fewer than the limit of resonance segments are considered structurally stable. It should be noted that for patients with implanted bone screws, if resonance occurs after surgery, it can easily lead to complications such as bone screw vibration and loosening or bone tissue fatigue and fracture. Therefore, a limit is set based on historical experience. Only when the number of resonance segments is less than this limit can the path be considered to be in a controllable response state; otherwise, it is eliminated.

[0059] Methods for performing end-cap closure analysis include: The spatial coordinates of the endpoint of the structurally stable bone nail path are obtained, and the presence of key bone tissue structures along the path axis in the region where the endpoint spatial coordinates are located is identified. Specifically, the spatial coordinates of the endpoint of any structurally stable bone nail path obtained through multiple screenings are determined, and the presence of key bone tissue structures in the anatomical region in front of the endpoint of the path axis is determined in combination with the three-dimensional bone model. Key bone tissue structures refer to structures such as cortical bone edges, bony prominences, subarticular bone surfaces, and cortical bone transition layers. These tissues have the significance of endpoint anchoring and support in the anatomical structure, and are therefore considered as key bone tissue structures.

[0060] If the path does not exist, the corresponding path is eliminated; otherwise, if it exists, a simulated vertical plane is set at the front end of the key bone tissue structure. The coordinates of the intersection point of the extension line of the endpoint spatial coordinates along the path axis direction on the simulated vertical plane are obtained. The angle of intersection between the extension line at the intersection point and the endpoint of the simulated vertical plane is measured. Specifically, a completely vertical simulated vertical plane is constructed at the front end of the key bone tissue structure, and an extension ray is emitted from the endpoint spatial coordinate position along the path axis direction to intersect the simulated vertical plane. The angle of intersection between the extension line at the intersection point and the simulated vertical plane is obtained.

[0061] If the incision angle of the endpoint is less than the preset closure angle threshold, the endpoint is determined to be a near-wall endpoint. The preset closure angle threshold is set based on clinical experience. If the incision angle of the endpoint is less than the threshold, it means that the path is close to the surface of the bone tissue at the endpoint. In this case, the end of the path may not be able to achieve a good anchoring effect, which may easily lead to the bone screw penetrating and penetrating the key bone tissue structure after surgery.

[0062] The normal direction of the continuous region through which the extension line passes is measured, and the change range of the angle between the extension line and the normal direction of the continuous region is calculated. In this process, geometric analysis is performed on several continuous regions through which the extension line extends forward to identify the normal vector of each region, and the angle between the extension line and the normal vector of that region is calculated when the extension line passes through each region. The difference between the consecutive angles is then calculated to obtain the change range of the angle.

[0063] If the angle change exceeds the preset normal drift threshold, the corresponding endpoint is determined to be a deflection endpoint. The normal drift threshold is set based on medical knowledge and clinical experience. If the angle change exceeds this threshold, it indicates that the angle change is large, and the corresponding area has significant directional jumps or structural unevenness. The corresponding endpoint is then determined to be a deflection endpoint. Such endpoints can lead to easy deviation of the bone screw during insertion and poor anchoring effect, posing a safety risk.

[0064] If a structurally stable bone screw path simultaneously contains a critical bone tissue structure along the path axis in the region where the endpoint's spatial coordinates are located, and this endpoint is either a near-wall endpoint or a deflection endpoint, then the corresponding path is identified as an insufficiently closed path and eliminated. The remaining structurally stable bone screw paths are determined as the final optimized bone screw paths. Since the absence of a critical bone tissue structure indicates a lack of end support, in this embodiment, paths without a critical bone tissue structure along the path axis after the endpoint have been preferentially eliminated. For paths containing a critical bone tissue structure along the path axis after the endpoint, if the endpoint of the path is either a near-wall endpoint or a deflection endpoint, it indicates that the endpoint may pose a significant safety hazard and is therefore eliminated. The remaining paths after multiple rounds of screening are determined as the final optimized bone screw paths that can be used as clinical application protocols.

[0065] This embodiment segmentes collected patient bone images and constructs a 3D bone model by combining the patient's historical medical data. Based on this 3D bone model, it constructs and optimizes bone nail paths, ultimately obtaining suitable bone nail paths, thus realizing a bone nail path optimization method. Compared with existing experience, it determines the location of cavity regions by calculating centroid offset and connectivity statistics using stacked mask images, achieving automatic separation of cavity structures from closed bone structures. Addressing the difficulty in determining blurred grayscale tissue boundaries in traditional methods, it improves the reliability of boundary judgment by constructing boundary judgment indicators, ensuring clear visibility of pathological boundaries. Paths are screened based on three aspects: cutting angle, connectivity, and stress distribution, obtaining suitable paths to ensure that the paths can be optimized at the level of specific operational steps during clinical operation. To accommodate individual differences, a bone age parameter is introduced, and based on… By selecting bone screw placement paths suitable for patients based on tissue type and bone age parameters, the method achieves a match between path structure and tissue repair speed, effectively reducing the risk of poor postoperative fusion and indirectly improving postoperative recovery speed. Adding vibration signals endows the path with basic resonance immunity during the planning stage, successfully identifying segments potentially affected by resonance. Furthermore, existing paths are selected based on the number of resonant segments, ensuring postoperative recovery stability. Analysis of the path's end closure accurately determines whether the path endpoint has a reliable support structure, further reducing the probability of postoperative tissue perforation or excessive screw end mobility, indirectly improving path stability. In summary, this method significantly improves the mechanical reliability and bio-fusion compatibility of bone screw placement schemes, strengthens the process avoidance mechanism for risk areas and the ability to ensure postoperative stability, and is particularly valuable for application in complex clinical scenarios.

[0066] Example 2 Please see Figure 2 As shown, parts not described in detail in this embodiment are described in Embodiment 1. An orthopedic consumables management system is provided, including: The data acquisition module is used to collect patient bone imaging data and historical case data; and to clean the patient bone imaging data to obtain orthopedic anatomical medical images. The image segmentation module is used to segment orthopedic anatomical medical images by combining historical case data to obtain boundary-defined medical images. The model building module is used to construct three-dimensional skeleton models based on medical images of tissue boundaries. The risk avoidance module is used to construct the nail path in the 3D skeletal model and output the risk-avoiding nail path by performing risk avoidance on the nail path. The path selection module is used to perform path load assessment on the risk avoidance nail paths and select load-suitable nail paths based on the assessment results; it also performs disturbance screening on the load-suitable nail paths to generate structurally stable nail paths. The final path generation module is used to perform end closure analysis on the structurally stable bone screw path, determine the final optimized bone screw path based on the analysis results, and use the final optimized bone screw path as the bone screw path scheme for clinical application; the modules are connected to each other via wired and / or wireless means.

[0067] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

[0068] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0069] In the description of this invention, it should be understood that the terms "first," "second," etc., are used only for distinguishing descriptions and should not be construed as indicating or implying relative importance.

[0070] In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0071] In the description of this invention, "several" means one or more, and "a large number" means two or more.

[0072] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," 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, the 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.

[0073] All formulas in this manual are dimensionless and calculated numerically. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters and thresholds in the formulas are set by those skilled in the art according to the actual situation.

[0074] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A method for optimizing bone screw paths, characterized in that, include: S1. Collect patient skeletal imaging data and historical case data; perform data cleaning on the patient skeletal imaging data to obtain orthopedic anatomical medical images; S2. Combine historical case data to perform image segmentation on orthopedic anatomical medical images to obtain tissue boundary medical images; S3. Construct a three-dimensional skeletal model based on medical images of tissue boundaries; S4. Construct the nail path in the 3D skeleton model, and output the risk-avoidance nail path by performing risk avoidance on the nail path; S5. Conduct a path load assessment on the risk avoidance nail path and select the nail path with suitable load based on the assessment results; Perturbation screening is performed on the applicable bone screw paths for the load to generate structurally stable bone screw paths; S6. Perform end-closure analysis on the structurally stable bone screw path, determine the final optimized bone screw path based on the analysis results, and use the final optimized bone screw path as the bone screw path scheme for clinical application.

2. The bone screw path optimization method according to claim 1, characterized in that, The methods for performing image segmentation include: Perform a region growing operation on any orthopedic anatomical medical image to extract the image region whose gray value is within a preset gray value range to obtain a bone tissue mask; measure the aspect ratio of the bounding box and the mean angle between the contour point normals of any closed region in the bone tissue mask, and determine the closed region whose aspect ratio of the bounding box and the mean angle between the contour point normals are both less than the preset corresponding threshold as a cavity suspected region. The bone tissue mask images of suspected cavity regions are sorted by time and stacked to form a mask image group; the centroid projection points of suspected cavity regions in each layer of the mask image group are obtained, and the centroid offset of adjacent image slices is calculated. At the same time, the number of connected image slices in the mask image group is counted. If the centroid offset is less than a preset offset threshold and the number of connected image slices is greater than a preset number threshold, the corresponding suspected cavity region is determined to be a cavity region. The cavity region is removed from the bone tissue mask image to obtain the adjusted mask image; the adjusted mask image is then delineated to output a tissue boundary medical image.

3. The bone screw path optimization method according to claim 2, characterized in that, The methods for boundary delineation include: After positioning and adjusting the mask image, connect the closed regions between them, calculate the difference in gray values ​​between adjacent pixels within the connected regions, and select the pixels whose gray value difference is greater than a preset transition threshold as boundary candidate points; construct a dynamic sliding window to traverse the connected regions, and calculate the texture contrast within the window centered on the boundary candidate points. Extract the tissue variation probability and bone density fluctuation of the region within a window centered on a candidate boundary point from historical case data, and construct a boundary determination index by stitching together texture contrast; calculate the confidence score of the boundary determination index, and determine the candidate boundary point as the boundary determination point if the confidence score is higher than the preset confidence level threshold; define the boundary determination point whose Euclidean distance from any boundary determination point is greater than the preset radius and does not belong to the same window as the adjacent node; connect the first boundary determination point with its adjacent nodes and set the boundary direction, and connect all boundary determination points with the same boundary direction to form a boundary path; fit a continuous boundary curve based on the boundary path, and project the continuous boundary curve onto the adjusted mask to obtain a medical image of tissue boundary.

4. The bone screw path optimization method according to claim 3, characterized in that, The methods for constructing a three-dimensional skeletal model include: A three-dimensional skeletal model is constructed based on an image set consisting of continuous medical images of tissue boundaries and the spatial coordinates of each tissue region therein, while all tissue structures in the three-dimensional skeletal model are labeled.

5. The bone screw path optimization method according to claim 4, characterized in that, The methods for risk avoidance include: The bone screw insertion area is determined and marked in the 3D bone model. The insertion start point and insertion end point in the bone screw insertion area are selected to form a set of path start and end point pairs. The shortest 3D Euclidean distance of each pair of path start and end points is calculated to obtain the initial bone screw path set. The path axis of any initial bone screw path at the insertion start point is extracted. At the same time, the bone surface normal of the area where the initial bone screw path is located is detected. The angle between the path axis and the bone surface normal is calculated. Initial bone screw paths with an angle greater than or equal to a preset feasible angle threshold are selected as preliminary feasible bone screw paths. A fixed window is used to capture a continuous local area of ​​the cortex along the path axis of the preliminary feasible bone nail path. The minimum penetration thickness of this continuous local area of ​​the cortex is measured, and paths with a minimum penetration thickness less than a preset insertion depth threshold are eliminated. At the same time, if there is a cavity defined area in the remaining preliminary feasible bone nail paths, the corresponding path is deleted to obtain the secondary judgment bone nail path. The secondary judgment of the nail path is divided into segments based on a determined step length to obtain equal-length path segments; connectivity analysis is performed on continuous equal-length path segments to filter out visually connected nail paths; stress distribution is judged on the visually connected nail paths to output the risk avoidance nail path.

6. The bone screw path optimization method according to claim 5, characterized in that, The methods for performing connectivity analysis include: Determine the location of the clinical surgical incision, calculate the spatial straight-line angle from the location of the clinical surgical incision to the center point of each equal-length path segment, and if the spatial straight-line angle is higher than the preset upper limit, the corresponding equal-length path segment is determined to be an obstructed path segment; determine whether the remaining equal-length path segments are in an inoperable tissue structure, if so, determine the corresponding equal-length path segment to be an obstructed path segment; if the number of consecutive obstructed path segments exceeds the preset obstruction number threshold, delete the corresponding secondary judgment bone nail path, and set the remaining paths as visible connected bone nail paths; The methods for determining stress distribution include: Extract equidistant regions on both sides of the visible connected nail path along the path axis to obtain the stress analysis region; detect the axial stress distribution and shear stress distribution in the stress analysis region, and simultaneously obtain the stress change sequence within the region; extract the peak value of axial stress and peak value of shear stress in the stress analysis region, and calculate the standard deviation of the stress change sequence. If any value exceeds the corresponding preset threshold, the corresponding visible connected nail path is considered a dangerous path and is removed, and the remaining visible connected paths, i.e., the avoidance nail paths, are output.

7. The bone screw path optimization method according to claim 6, characterized in that, The methods for performing path load assessment include: Identify the tissue structures traversed by the path of the safety nail and construct a penetrating tissue sequence based on the arrangement order of the tissue structures; label the penetrating tissue sequence with tissue type based on the biological characteristics of the tissue structure; and divide the penetrating tissue sequence into tissue segments based on the tissue type label to obtain tissue classification subsegments. Obtain the regeneration and repair rate and load index of each tissue type corresponding to each tissue classification segment; estimate the repair cycle of each tissue classification segment based on the regeneration and repair rate to obtain the physiological repair time; identify tissue classification segments whose physiological repair time is greater than the standard repair time and whose load index is greater than the preset index threshold, and determine them as high load segments. Based on historical case data, bone age parameters of patients are obtained, and the developmental level of corresponding tissue classification segments is assessed based on these parameters. The biological metabolic response rate is calculated by combining the developmental level and the bone age parameters of the corresponding tissue classification segment. If the biological metabolic response rate is lower than the preset corresponding bone age metabolic threshold, the corresponding tissue classification segment is identified as a metabolic overload segment. High-load segments, metabolic overload segments, and continuous segments composed of adjacent high-load or metabolic overload segments are collectively referred to as unstable segments. If the length of an unstable segment in the risk avoidance nail path is higher than a preset length ratio, the risk avoidance nail path is identified as an unstable path. All risk avoidance nail paths after eliminating unstable paths are integrated to obtain the load-suitable nail path.

8. The bone screw path optimization method according to claim 7, characterized in that, The methods for performing perturbation screening include: A small-amplitude, fixed-frequency vibration signal is added to the load-bearing bone screw path, and the vibration response signal of the local area of ​​the path is received. Based on the vibration response signal, the change in response fluctuation of the corresponding local area of ​​the path is calculated, and a vibration response curve is constructed using the change in response fluctuation. The maximum response frequency in each period of the vibration response curve is extracted, and the response change rate of that period is calculated. A vibration response function is constructed based on the maximum response frequency and the response change rate. The sum of the function values ​​of each period of the local area of ​​the path in each load-bearing bone screw path is calculated using the vibration response function. The local areas of the path whose sum of function values ​​is greater than a preset function and threshold are marked as abnormal response segments. If the frequency range of the abnormal response segment overlaps with the known range of vibration frequencies that are easily triggered after surgery in patients, the abnormal response segment is determined to be a resonance segment. Load-bearing bone screw paths with fewer resonance segments than the limit are selected as structurally stable bone screw paths.

9. The bone screw path optimization method according to claim 8, characterized in that, The methods for performing end-cap closure analysis include: Obtain the spatial coordinates of the endpoint of the structurally stable bone screw path, and identify whether there is a critical bone tissue structure in the area where the endpoint spatial coordinates are located along the path axis. If it does not exist, the corresponding path is eliminated; otherwise, if it exists, a simulated vertical plane is set at the front end of the critical bone tissue structure. Obtain the coordinates of the intersection point of the extension line of the endpoint spatial coordinates along the path axis on the simulated vertical plane, and measure the angle of the extension line at the intersection point with the endpoint of the simulated vertical plane. If the angle of the endpoint is less than the preset closing angle threshold, the endpoint is determined to be a near-wall endpoint. Measure the normal direction of the continuous region through which the extension line passes, and calculate the change in the angle between the extension line and the normal direction of the continuous region. If the change in the angle is higher than the preset normal drift threshold, the corresponding endpoint is determined to be a deflection endpoint. If the structurally stable bone nail path also meets the condition that there is a key bone tissue structure along the path axis in the region where the endpoint spatial coordinates are located, and the endpoint is either a near-wall endpoint or a deflection endpoint, the corresponding path is determined to be an insufficiently closed path and is eliminated. The remaining structurally stable bone nail paths are determined as the final optimized bone nail paths.

10. An orthopedic consumables management system, used to implement the bone screw path optimization method according to any one of claims 1-9, characterized in that, include: The data acquisition module is used to collect patient bone imaging data and historical case data; and to clean the patient bone imaging data to obtain orthopedic anatomical medical images. The image segmentation module is used to segment orthopedic anatomical medical images by combining historical case data to obtain boundary-defined medical images. The model building module is used to construct three-dimensional skeleton models based on medical images of tissue boundaries. The risk avoidance module is used to construct the nail path in the 3D skeletal model and output the risk-avoiding nail path by performing risk avoidance on the nail path. The path filtering module is used to perform path load assessment on the risk avoidance nail path and filter the nail path with load applicable based on the assessment results. Perturbation screening is performed on the applicable bone screw paths for the load to generate structurally stable bone screw paths; The final path generation module is used to perform end closure analysis on the structurally stable bone screw path, determine the final optimized bone screw path based on the analysis results, and use the final optimized bone screw path as the bone screw path scheme for clinical application; the modules are connected to each other via wired and / or wireless means.

Images