A navigation method for reduction of high-energy injuries of lower limb bones

Abstract

The present invention discloses a method for repositioning and navigating high-energy injuries of the lower limb skeleton. The method comprises establishing a dynamic biomechanical model of a robot and the musculoskeletal system and performing repositioning and navigation simulation, analyzing the motion and force relationship between the fracture fragments and the robot, planning a skeletal repositioning and navigation path, and recording the path as a basis for optimal path selection and a reference for model optimization design. Simulation is performed in conjunction with medical imaging and anatomical histological data to construct an optimal path planning cost function, determine the parameters for searching the repositioning and navigation path, and determine and fix the target of the optimal path algorithm. The present invention is the first to consider avoiding important soft tissues as a key factor in repositioning and fixing the path. It comprehensively considers the repositioning action mode, the longitudinal traction muscle force, the joint position, and other multi-information fusion decisions to optimize the planning path, thereby solving the problems of insufficient flexibility and the susceptibility to secondary injuries in the previous repositioning and navigation process of parallel robots.

Description

Technical Field

[0001] The present invention relates to the field of reduction and fixation, and in particular to a navigation method for reduction of high-energy injuries of lower limb bones. Background Art

[0002] Reduction and fixation are key steps in fracture treatment, and various computer- and robotic-assisted methods have been developed to avoid problems associated with reduction and fixation. Despite advances in this field, research into the problems that can occur during actual reduction and fixation remains insufficient. These can lead to erroneous procedures, prolonged operative time, and increased radiation exposure. Therefore, the optimal description of the reduction and fixation pathway is considered a key approach for analyzing problems associated with reduction and fixation and for further improving reduction and fixation procedures. Detailed studies of dynamic reduction and fixation procedures are currently lacking in the literature, resulting in a lack of description of the optimal reduction and fixation pathway.

[0003] Graham AE et al. and Joung et al. developed musculoskeletal models to determine muscle forces during fracture reduction and fixation. However, the effects of different reduction motions on force generation were not studied in detail. Westphal et al. proposed a reduction path planning method as part of their robot-assisted development. This method aims to minimize distraction and thus reduce applied forces, but analysis of the optimal selection of reduction motions is lacking and warrants further investigation. Summary of the Invention

[0004] The purpose of the present invention is to provide a method for the reduction and navigation of high-energy injuries of the lower limb skeleton. For the first time, the avoidance of important soft tissues is taken as an important consideration in the reduction path planning and navigation. The planning path is optimized by comprehensively considering the reduction action mode, the longitudinal traction muscle force, the joint position and other multi-information fusion decisions, so as to solve the problems of insufficient flexibility and easy secondary injury in the previous parallel robot reduction and navigation process.

[0005] The present invention provides a navigation method for lower limb bone high-energy injury reduction, the method comprising the following steps:

[0006] Step 1: Based on the reduction force prediction research of the musculoskeletal model of the fracture site, a dynamic biomechanical model of the robot's musculoskeletal system was established using OpenSim;

[0007] Step 2: Perform repositioning navigation simulation on the established dynamic biomechanical model of the robot's musculoskeletal system to analyze the relationship between the motion and force between the fracture fragments and the robot, and predict the impact of muscle resistance on the robot system;

[0008] Step 3: Based on the simulation results of step 2, perform bone reduction navigation path planning and record it as the basis for selecting the optimal path and the reference for model optimization design;

[0009] Step 4: Based on the above simulation results, the optimal bone reduction navigation path is described by comprehensively considering multiple information, including: the sequential changes in movement patterns, the relationship between muscle force and joint angles synthesized by longitudinal traction, and the avoidance of important tissues; simulation is performed in combination with medical imaging and anatomical histological data, and the optimal weight of multiple information is determined according to different fracture types;

[0010] Step 5: According to the simulation results, a planning strategy is predetermined, an optimal path planning cost function is constructed, the parameters and spatial position conversion matrix of the bone reduction navigation path are determined, and the optimal bone reduction navigation path is obtained.

[0011] The present invention has the following beneficial effects: This reduction and navigation method proposes the establishment of a dynamic biomechanical model of the musculoskeletal system, describing complex biological systems to analyze the relationship between the motion and force between the bone fragments and the robot, predicting the impact of muscle force on the robot system, using CTA and MRI images to achieve three-dimensional geometric reconstruction of biological tissues before surgery, identifying the location of blood vessels and nerves, and for the first time taking the avoidance of important soft tissues as a key consideration in reduction path planning and navigation. The planning path is optimized by comprehensively considering the reduction action mode, longitudinal traction muscle force, joint position, and other multi-information fusion decisions, which is more in line with clinical requirements. Based on the required mechanical performance and spatial range requirements, the surgery is combined with the postoperative fixation method of the first ring installation of the hexapod external fixator to solve the problems of insufficient flexibility and easy secondary injury in the reduction and fixation process of previous parallel robots. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 is a schematic diagram of the integration of the fracture-specific bone model and the original bone model;

[0013] Figure 2 This is a schematic diagram of Hill's muscle-tendon mechanics model;

[0014] Figure 3 is a schematic diagram of the bone reduction force model;

[0015] Figure 4 It is a schematic diagram of collision avoidance path planning;

[0016] Figure 5 It is a schematic diagram of fracture reduction path transformation and inverse transformation;

[0017] Figure 6 is a general operation diagram of path planning in the frontal plane; Figure 6 (a) is the initial node position C at the distal end of the fracture init and the desired target node position C target Schematic diagram of; Figure 6(b) in the figure is a route map starting from the starting node s (reset position) and planning to the target position z (initial position); Figure 6 (c) in the figure is a simplified diagram of the path planning. DETAILED DESCRIPTION

[0018] The present invention will be further described in detail below with reference to the accompanying drawings.

[0019] like Figures 1-6 As shown, the present invention provides a method for reduction and fixation of high-energy injuries of lower limb bones, the method comprising:

[0020] Step 1: Based on the reduction force prediction study of the musculoskeletal model of the fracture site, a dynamic biomechanical model of the lower limb skeletal-muscular system and a robotic system were established using OpenSim;

[0021] Step 2: Perform repositioning navigation simulation on the established lower limb skeletal-muscular system model to analyze the motion and force relationship between the fracture fragments and the robot, and predict the impact of muscle resistance on the robot system;

[0022] Step 3: Based on the simulation results of step 2, perform bone reduction navigation path planning and record it as the basis for optimal path selection and model optimization design reference;

[0023] Step 4: Based on the aforementioned simulation results, a method for describing the optimal bone reduction navigation path was proposed that comprehensively considers multiple information, including the sequential changes in movement patterns (rotation, stretching, alignment, etc.), the muscle force synthesized by longitudinal traction, the relationship between joint angle and force, and the avoidance of critical tissues (such as bone fragments, blood vessels, and nerves). Simulations were conducted in conjunction with medical imaging and anatomical histological data, and the optimal weighting of multiple information was determined based on different fracture types.

[0024] Step 5: Based on the simulation results, a planning strategy is predetermined, an optimal path planning cost function is constructed, the parameters for searching the bone reduction navigation path and the spatial position conversion matrix are determined, the reduction navigation process of the optimal reduction path is realized, and the goal of the optimal complex navigation path algorithm is obtained;

[0025] Step 6: Conduct experimental analysis to verify the optimal reset navigation path algorithm.

[0026] The present invention is further configured as follows: in step one, the establishment of a dynamic biomechanical model includes the establishment of a lower limb skeletal-muscle structure model, the expression and calculation of muscle-tendon-mechanical relationships, the kinematic expansion of the fracture joint, the calculation of muscle force during reduction, and the verification of the model.

[0027] The present invention is further configured as follows: the establishment of the lower limb skeletal-muscle structure model comprises the following steps:

[0028] i) Using OpenSim to build a virtual 3D bone model from CTA scans, segment the distal and proximal ends (including bone fragments for complex fractures) from the CTA images, generate the 3D model, and align it;

[0029] ii) Since any position and orientation of the 3D bone model relative to the coordinate system must be registered with the original bone of the original bone model, an optimal registration transformation matrix is ​​selected through registration error analysis and comparison;

[0030] iii) At the registration position C Fem Define a fracture-specific bone model to replace the undamaged femur in the original bone model. The integration process of the fracture-specific bone model and the original bone model is as follows:

[0031] (a) Calculation of the initial position offset and target position of the bone fragments, including the fracture coordinate system visualized in the fracture-specific bone model;

[0032] (b) Integration of the registered original bone model and the complete fracture-specific bone model with respect to the registered position, including the transferred bone model in the fracture-specific bone model and all associated muscle pathways.

[0033] The muscle actuator is simulated in OpenSim using an action line model. The geometry of a muscle consists of a path, which is composed of a series of points connected by polygons. A muscle consists of at least two points, the start point and the end point. The points of the muscle path are attached to the body (skeleton) and are referenced relative to its coordinate system. The total length of the muscle model is calculated from the following formula (1):

[0034]

[0035] Where, l MT is the total length of the muscle model;

[0036] P i are points on the polygon representing the muscle geometry;

[0037] i is a continuous natural number starting from 1 to n-1;

[0038] The present invention is further configured as follows: the expression and calculation of the muscle-tendon-mechanical relationship includes the following steps: a muscle spans one or more joints and generates force when activated, thereby generating a torque around the joint; for muscle-tendon simulation, OpenSim provides multiple models, the mechanical properties of which are described by the Hill model, the contraction element CE represents the contraction dynamics (force-length and force-speed characteristics), the parallel elastic element PE represents the passive properties of the connective tissue structure, and the tendons are arranged continuously and modeled as elastic nonlinear springs.

[0039] The mechanical calculation of a single muscle is expressed by formula (2). By listing the explicit expressions of other parameters, the first-order differential equation can be established and numerically solved. MT The solution is used to calculate the muscle force during the subsequent reduction process.

[0040]

[0041] Where, F T is the force in the tendon direction in the muscle-tendon Hill model;

[0042] F M is the force in the muscle-tendon Hill model;

[0043] F MT Generate force for the muscles and tendons in the Hill model;

[0044] F CE is the contraction force of the muscle in the Hill model;

[0045] F PE is the elastic dynamics of the muscle in the Hill model;

[0046] α is the angle between the muscle and tendon in the Hill model;

[0047] The present invention is further configured as follows: the kinematic expansion of the fracture joint comprises the following steps: the kinematic chain of the integrated model of the fracture-specific bone model and the original bone model connects the femur to the pelvis through a spherical joint, and the lower leg is connected to the femur through a knee joint; the fracture divides the bone into a proximal Prox and a distal Dist, and the distal Dist and the proximal Prox perform relative motion with six degrees of freedom (clinically, the six degrees of freedom of a fracture are also called six reduction parameters, namely, ventral / dorsal (sagittal direction) displacement, lateral / medial (transverse) displacement, distal / proximal (longitudinal) displacement, internal / external rotation deviation, anterior curve / posterior curve, and varus / valgus dislocation); in the original bone model, the integrated model of the fracture-specific bone model and the original bone model is realized by inserting two main bodies, the distal Dist bone and the proximal Prox bone, and the muscle attachment of the fracture is defined as a tendon actuator simulating its line of action;

[0048] Establish a geometric path, including at least two points (origin and insertion point), hide the lines representing muscles in the original bone model, import the surgical robot model, and establish a dynamic model of the connection between the robot and the distal end of the fracture Dist;

[0049] Furthermore, the dynamic behavior of the entire musculoskeletal system is described by its equation of motion, which consists of the following:

[0050] M(q)q=T MT (q)+C(q,q)+G(q)+E(q,q) (3)

[0051] Where q is the generalized coordinate of the system;

[0052] M(q) is the mass matrix of the system;

[0053] C is the centrifugal force and Corioli force and torque vector;

[0054] G is the gravity and torque vector;

[0055] E is the force and torque vector of the external environment on the human body;

[0056] By solving the equations of motion, the motion can be determined by muscle tension (forward dynamics) or the torque that causes the motion (inverse dynamics). For forward dynamics simulations, the equations of motion are given by the acceleration q and the inverse dynamics T MT The simulation is solved accordingly.

[0057] The present invention is further configured as follows: the calculation of the muscle force during the reset comprises the following steps: the reset model provides a reset path as a relative movement between the distal end and the proximal end, and OpenSim calculates the synthetic muscle force based on the change in the length of the muscle actuator.

[0058] Furthermore, muscle length is influenced by the position of the hip and knee joints and therefore depends on the following parameters:

[0059]

[0060] Among them, F i MT is the force in each individual tendon complex i, F res The analysis will be F i MT The force is decomposed into three components related to the fracture joint, F sag is the force in the sagittal direction, F trans is the lateral force, F long is the longitudinal force. We intend to borrow the concept of torque from physics and calculate the torque around the fracture joint in OpenSim based on the theory of virtual work.

[0061]

[0062] Substitute the above F MT 、l MT , and synthesized by the three force components, the force vector can be shown as follows:

[0063]

[0064] Furthermore, the validation of the model includes the following steps:

[0065] 1. Calculate the amount of muscle deformation caused by manipulating the hip and knee angles;

[0066] 2. Compare the muscle forces detected by clinical surgery and clinical anatomical data with the OpenSim model. If the values ​​are within the deviation range, it is assumed that the model has been created correctly.

[0067] 3. Analyze the causes of errors, adjust simulation system parameters, and establish a sample database to make the simulation model more accurate and reasonable.

[0068] Furthermore, in the method for reduction and fixation of high-energy injuries of the lower limb skeleton, in step 4, the optimization process of the reduction and fixation path is as follows:

[0069] 1) Analyze each factor influencing the reduction process, including the sequential changes in movement patterns (rotation, stretching, alignment), the forces of each muscle in longitudinal traction, and the relationship between joint angle and force. Specifically: ① By comparing two simulated reduction processes with different movement pattern sequences, we analyze the reduction force, plot data curves, and determine the combined order of movement patterns such as rotation, stretching, and alignment. ② The length of the muscle path and the resulting force depend on the joint angle. By simulating changes in joint angle in each individual step, we analyze the impact of various joint angles on the generated force, and comprehensively consider the ideal and uniform hip / knee position as the basis for optimal path planning.

[0070] 2) CTA-MRI-based 3D geometry reconstruction of biological tissues. Specifically: ① Based on the characteristics of CTA images, skeletal and skin models are reconstructed using the Marching-Cubes algorithm, and nerve distribution is detected through preoperative MRI. ② CTA-MRI image fusion and registration technology is used to achieve 3D geometric reconstruction of important tissues such as bones, muscles, blood vessels, and nerves, providing basic data for collision detection of the reduction path.

[0071] 3) Based on the data from process 2), collisions of bones, blood vessels, and nerves are detected and avoided, specifically: ① A point polyhedron test method is used, where each virtual bone fragment is represented by a polyhedron consisting of a triangle and its corresponding vertices. If at least one vertex of one bone model is inside another, a collision / intersection occurs. The inclusion is tested using a ray casting method and the number of test triangles is minimized based on recursive subdivision to accelerate the test of points in the polyhedron; ② The movement of the distal end of the fracture is represented by the movement of the center point, which is the intersection between the axis of the fragment and the offset surface. The distal end of the fracture runs along the reduction path, avoiding critical tissues (blood vessels, nerves, which are modeled and displayed as spheres). A parameter, collision risk, is used to evaluate the risk of the distal fragment approaching critical tissues. The parameter ranges from 0 to 1. A higher value means that the reduction path is more likely to collide with critical tissues, such as Figure 4 shown.

[0072] Furthermore, step five mainly includes the establishment of a reset path expression, the predetermination of a planning strategy, and the implementation of an optimal path planning search algorithm.

[0073] Establishment of reduction path expression: The fracture reduction process is guided by the distal end of the fracture from the initial position Dist init To the target location Dist target The required single motion sequence is composed of the following mathematical expressions: Dist target =Γ Dist Dist init , the reference coordinate system C generated by the above i) CT , the reduction parameters can be calculated to determine the position and direction of the distal end of the fracture. The reduction path expression can be expressed as:

[0074]

[0075] Where T i is the transformation of each individual reset step i, all T i The combination of generates the reset path Γ Dist , the distal end of the fracture moves from the starting position (i=1) to the target position (i=N). Each individual transformation T i Corresponding to the actual reduction action in clinical practice, it includes changes in 6 reduction parameters (usually obtained through CT scan): ventral / dorsal (sagittal direction) displacement d s , lateral / medial (transverse) displacement d t , distal / proximal (longitudinal) displacement d l , internal / external rotation deviation α, anterior curve / posterior curve β, varus / valgus misalignment γ, these reduction parameters are used to describe the distal position of the fracture and provide information on the translation and rotation deviation of the distal fracture from the target position.

[0076] like Figure 5 The coordinate system shown in the figure is C, which is the current position of the distal end of the fracture according to the transformation relationship. i , T can be calculated N T N- 1...T1, the derivation process is as follows:

[0077]

[0078]

[0079] C i =C target ΔT i ; i=1,2,...,N (10)

[0080]

[0081] By linking the transformations and substituting expression (11) into expression (7), the basic reset path Γ can be obtained. Dist , and obtain a series of basic reset paths for subsequent optimal path search.

[0082] Predetermination of planning strategy: The goal of automatic path planning is to search for the optimal path based on the aforementioned ① basic reset path. This requires considering factors such as the dynamic reset force, action sequence, and collision avoidance mentioned in 1), 2), and 3) above. The specific conditions are: ① Collision-free path: Exclude reset movements that may cause collisions with broken bones, nerves, and blood vessels from path planning to create a collision-free path; ② Minimum force path: Consider changes in muscle strength to plan a minimum force path; ③ Shortest path: The goal of searching for the shortest path is to reach the target position directly and avoid any circuitous paths.

[0083] To simplify the algorithm, based on the simulation results of 1), 2), and 3) dynamic reset force, action sequence, collision avoidance, etc., the path planning strategy is further restricted, including:

[0084] I) Reducing degrees of freedom through translational paths. Each reset motion typically depends on six reset parameters. If more than three parameters are changed simultaneously between two reset steps, it results in a combined translational and rotational movement. Executing such complex maneuvers is not always feasible. Furthermore, considering all six degrees of motion in each reset step increases the complexity of the planning algorithm. To reduce the degrees of freedom, only a translational path is planned, compensating for rotational deviations in the first reset step, followed by a purely translational offset toward the target.

[0085] II) Repositioning to the base position. Due to the different positions of the hip and knee joints, different muscle conditions lead to different tension effects, affecting the reduction results. Therefore, as a boundary condition, the base position for adjusting the reduction movement is determined by the simulation results of the action sequence in 1).

[0086] III) Reverse Path Planning. By planning the reverse path, the efficiency of path search can be improved. Therefore, path planning is proposed to return from the target position to the initial position.

[0087] like Figure 6 , implementation of the optimal path planning search algorithm: Figure 6 a shows the initial node position C at the distal end of the fracture init and the desired target node position C target .like Figure 6As shown in b, reverse path planning is used. The planning algorithm starts from the starting node s (reset position) to plan the route to the target position z (initial position). First, the node u to be checked corresponds to the starting node s. The neighboring nodes v of this node are checked. Each of these nodes v corresponds to a displacement of the distal segment in the spatial direction with a step length of Δt, thus forming a grid of neighboring nodes. The motion is then checked for collisions. The neighboring nodes that collide are in Figure 6 The nodes are marked dark in b and are excluded from further path planning. Considering only the points that are allowed to move, the following cost function is constructed to select and evaluate the neighboring nodes v of node u:

[0088]

[0089] Where sum(v) is the sum of all path costs from the starting node s to the current node v.

[0090] h(v,z): distance as a prediction function used to estimate the path cost from the current node v to the destination z.

[0091] ω·F res (v): Muscle load cost. F res is the force produced by all muscles, and ω is a weight factor that describes how much the force affects the cost function.

[0092] Considering that the step length Δt and weight coefficient ω of muscle force can be changed, the algorithm checks the node closest to the target and selects the node with the lowest cost f(v). The simplified representation of path planning is as follows Figure 6 c.

[0093] The following experimental analysis of the navigation path established by the present invention's method primarily includes intraoperative skeletal muscle force / torque testing, reduction testing, and data analysis and processing. The intraoperative skeletal muscle force / torque testing, performed during traditional orthopedic surgery, uses a six-degree-of-freedom force / torque sensor to measure the forces and torques applied during bone and soft tissue manipulation, recording the maximum forces and torques encountered during fracture reduction.

[0094] The reduction test involved several SYNBONE femoral models (SYNBONE AG, Marans, Switzerland) that were fractured in various ways and then repositioned using the designed robot and reduction algorithm. The repositioning algorithm calculated the target position and alignment of the bone fragments. Starting from a randomly selected starting position, the robot used an optical positioning sphere to simulate tissue, such as blood vessels and nerves, and automated path planning, which accounted for muscle strength and collision detection, was performed to measure reduction accuracy.

[0095] Data analysis and processing is based on the test data obtained from various experiments. MATLAB is used to analyze the data, compare the test data with the simulation curve, and verify the normal distribution of the data set. The error between the test and the theory is calculated, the factors that cause the error are analyzed, and further improvements are made based on the results.

[0096] The specific embodiments are merely explanations of the present invention and are not limitations of the present invention. After reading this specification, those skilled in the art may make non-creative modifications to the embodiments as needed. However, as long as they are within the scope of the claims of the present invention, they are protected by patent law.

Claims

1. A navigation method for lower limb bone high-energy injury reduction, characterized in that: The method comprises the following steps: Step 1: Based on the research on the reduction force prediction of the musculoskeletal system at the fracture site, a dynamic biomechanical model of the robot's musculoskeletal system was established using OpenSim; Step 2: Perform repositioning navigation simulation on the established dynamic biomechanical model of the robot's musculoskeletal system to analyze the relationship between the motion and force between the fracture fragments and the robot, and predict the impact of muscle resistance on the robot system; Step 3: Based on the simulation results of step 2, perform bone reduction navigation path planning and record it as the basis for selecting the optimal path and the reference for model optimization design; Step 4: Based on the above simulation results, the skeletal reduction navigation path is described by comprehensively considering multiple information, including the sequential changes in movement patterns, the relationship between the muscle force synthesized by longitudinal traction and joint angles, and the avoidance of important tissues; Combining medical imaging and anatomical histological data for simulation, the weight of multivariate information is determined according to different fracture types; The step 4 is specifically as follows: 4.1) Analyze the factors affecting the reset process one by one, specifically: ① By comparing two simulated reset processes with different execution motion pattern sequences, reset force analysis is performed, data curves are drawn, and the combination sequence of rotation, stretching, and alignment motion patterns is determined; ② The length of the muscle path and the resulting force depend on the joint angle. By simulating changes in the joint angle in each individual step, the influence of various joint angles on the generated force is analyzed. The ideal and uniform hip / knee position is comprehensively considered as the basic position for optimal path planning. 4.2) Reconstructing 3D geometry of biological tissue based on CTA-MRI, specifically: ① Based on the characteristics of CTA images, the bone and skin models are reconstructed in three dimensions using the Marching-Cubes algorithm, and the nerve distribution is detected through preoperative magnetic resonance imaging. ② Using CTA-MRI image fusion and registration technology to achieve three-dimensional geometric reconstruction of bones, muscles, blood vessels, and nerve tissues, providing basic data for collision detection of the reset path; 4.3) Based on the data from 4.2), collisions of bones, blood vessels, and nerves are detected and avoided. Specifically: ① Using a point-polyhedron testing method, each virtual bone segment is represented by a polyhedron consisting of a triangle and its corresponding vertices. If at least one vertex of one skeletal model is located inside another, a collision / intersection occurs. Ray casting is used to test the polyhedron containing triangles and their corresponding vertices, and the number of test triangles is minimized based on recursive subdivision to accelerate the testing of points in the polyhedron. ② The motion of the distal end of the fracture is represented by the motion of the center point, which is the intersection between the axis of the fragment and the offset surface. The distal end of the fracture moves along the reduction path, avoiding critical tissue. The risk of the distal fragment approaching critical tissue is assessed using a parameter, collision risk, which ranges from 0 to 1. Higher values ​​mean that the reduction path is more likely to collide with critical tissue. Step 5: According to the simulation results, a planning strategy is predetermined, an optimal path planning cost function is constructed, the parameters and spatial position conversion matrix of the bone reduction navigation path are determined, and the optimal bone reduction navigation path is obtained.

2. A navigation method for reduction of high-energy injuries of lower limb bones according to claim 1, characterized in that: The step 1 comprises the following steps: 1.1) Establish a fracture-muscle structure model; 1.2) Express and calculate muscle-tendon-mechanical relationships; 1.3) Perform kinematic extension of the fractured joint; 1.4) Calculate muscle force during reduction; 1.5) Validate the model.

3. A navigation method for reduction of high-energy injuries of lower limb bones according to claim 2, characterized in that: Step 1.1) specifically includes: i) Using OpenSim to build a virtual 3D bone model from CTA scans, segment the distal and proximal ends from the CTA images, generate a 3D model, and align them; ii) Since the 3D bone model is relative to the CT coordinate system C CT Any position and orientation of the bone must be aligned with the original bone, and the optimal alignment transformation matrix is ​​selected through alignment error analysis and comparison; iii) At the registration position C Fem A fracture-specific bone model was defined, replacing the undamaged femur in the original bone model; The muscle actuator is simulated in OpenSim using the action line model; the muscle geometry consists of a path consisting of a series of points (P1,...,P n ) is composed of at least two points, namely the starting point P1 and the end point P n The points of the muscle path are attached to the bones, and the total length of the muscle model is calculated from the following formula (1) by referencing it relative to its coordinate system: Where, l MT is the total length of the muscle model; P i are points on the polygon representing the muscle geometry; i is a continuous natural number starting from 1 to n-1; The integration process of the fracture-specific bone model and the original bone model is as follows: (a) Calculate the initial position offset and target position of the bone fragment; (b) Integration of the registered original bone model and the complete fracture-specific bone model with respect to the registered position.

4. A navigation method for reduction of high-energy injuries of lower limb bones according to claim 2, characterized in that: Step 1.2) specifically includes the following steps: A muscle spans one or more joints and generates a force F when activated. MT , thereby generating torque around the joint; For muscle-tendon simulation, OpenSim provides multiple models, the mechanical properties of which are described by the Hill model, the contraction element CE represents the contraction dynamics, the parallel elastic element PE represents the passive properties of the connective tissue structure, and the tendon is arranged continuously and modeled as an elastic nonlinear spring; The mechanical calculation of a single muscle is expressed by formula (2), and the explicit expressions of other parameters are listed. The first-order differential equation is established and numerically solved. MT ; Where, F T is the force in the tendon direction in the muscle-tendon Hill model; F M is the force in the muscle-tendon Hill model; F MT Generate force for the muscles and tendons in the Hill model; F CE is the contraction force of the muscle in the Hill model; F PE is the elastic dynamics of the muscle in the Hill model; α is the angle between the muscle and tendon lines in the Hill model.

5. The method for navigation reduction of high-energy injuries of lower limb bones according to claim 2, characterized in that: Step 1.3) specifically includes the following steps: Integration of the fracture-specific bone model and the original bone model The kinematic chain of the model connects the femur to the pelvis through a spherical joint, and the lower leg to the femur through the knee joint; the fracture divides the bone into the proximal Prox and the distal Dist, and the distal Dist and the proximal Prox move relative to each other with 6 degrees of freedom. The 6 degrees of freedom are also called 6 reset parameters, namely ventral / dorsal displacement, lateral / medial displacement, distal / proximal displacement, internal / external rotation deviation, anterior curve / posterior curve, and varus / valgus dislocation. The integration model of the fracture-specific bone model and the original bone model is realized by inserting the distal Dist bone and the proximal Prox bone as two main bodies, and the muscle attachment of the fracture is defined to simulate the tendon actuator as its line of action; Establish a navigation geometry path, including at least the origin and insertion points, and hide the lines representing the muscles in the original bone model, import the robot model, and establish a dynamic model of the connection between the robot and the distal Dist bone; The dynamic behavior of the entire musculoskeletal system is described by its equations of motion, which consist of the following: M(q)q=T MT (q)+C(q,q)+G(q)+E(q,q) (3) Where q is the generalized coordinate of the system; M(q) is the mass matrix; C is the centrifugal force and Corioli force and torque vector; G is the gravity and torque vector; E is the force and torque vector of the external environment on the human body; By solving the equations of motion, the motion is determined by the muscle tension or torque that causes the motion.

6. A method for navigation reduction of high-energy injuries of lower limb bones according to claim 4, characterized in that: Step 1.4) specifically includes the following steps: OpenSim calculates the resultant muscle force based on the changes in the length of the muscle actuator; The muscle actuator length is affected by the position of the hip and knee joints and depends on the following parameters: Among them, F i MT is the force in each individual tendon complex i, F res For its combined force; analysis will F i MT The force is decomposed into three components related to the fracture joint, F sag is the force in the sagittal direction, F trans is the lateral force, F long It is the longitudinal force. Borrowing the concept of torque in physics and based on the theory of virtual work, the torque around the fracture joint is calculated in OpenSim. Substitute F in step 1.2) MT 、l MT , and synthesized by the three force components, the force vector is as follows:

7. The method for navigation reduction of high-energy injuries of lower limb bones according to claim 2, characterized in that: Step 1.5) specifically includes the following steps: 1) Calculate the amount of muscle deformation caused by manipulating the hip and knee angles; 2) the detected muscle forces and clinical anatomical data were compared with the OpenSim model, and if the values ​​were within the deviation range, it was assumed that the model was created correctly; 3) Analyze the causes of errors, adjust simulation system parameters, and establish a sample database to make the simulation model more accurate and reasonable.

8. The method for navigation reduction of high-energy injuries of lower limb bones according to claim 1, characterized in that: The step five specifically includes: 5.1) Establishment of reset path expression: The fracture reduction process is to guide the distal end of the fracture from the initial position Dist init To the target location Dist target The required single motion sequence is composed of the following mathematical expressions: Dist target =Γ Dist Dist init , based on the generated reference coordinate system, the reduction parameters are calculated to determine the position and direction of the distal end of the fracture; the reduction path expression is expressed as: in is the transformation of each individual reset step i, all The combination of generates the reset path Γ Dist , the distal end of the fracture moves from the starting position i = 1 to the target position i = N; each individual transformation Corresponding to the actual reduction action in clinical practice, it includes changes in 6 reduction parameters. These reduction parameters are used to describe the distal position of the fracture and provide information on the translation and rotation deviation of the distal fracture from the target position. According to the transformation relationship, the current position of the distal end of the fracture is C i , calculate T N T N-1 ...T1, the derivation process is as follows: C i =C target ΔT i ;i=1,2,...,N (10) By linking the transformations, substituting expression (11) into expression (7), we can obtain the basic reset path Γ Dist , obtain a series of basic reset paths for subsequent optimal path search; 5.2) Planning strategy: The goal of automatic path planning is to search for the optimal path based on the basic reset path obtained in 5.

1. The specific process is as follows: ① Collision-free path: Exclude the reduction motion that may cause collision with broken bones, nerves, and blood vessels from path planning to create a collision-free path; ② Minimum force path: Taking into account the changes in muscle strength to plan the minimum force path; ③Shortest path: The goal of searching for the shortest path is to reach the target location directly and avoid any circuitous paths; 5.3) Implementation of the optimal path planning search algorithm: Mark the initial node position and the desired target node position C at the distal end of the fracture target ,Using reverse path planning, the planning algorithm starts from the starting node s to plan a route to the target location z; First, the node u to be checked corresponds to the starting node s; the neighboring nodes v of this node are checked; each of these nodes v corresponds to a displacement of the distal fragment in the spatial direction with a step size of Δt, thus forming a grid of neighboring nodes; then the movement is checked for collisions; the neighboring nodes that collide are marked and excluded from the path planning; only the allowed movements are considered, and the following cost function is constructed to select and evaluate the neighboring nodes v of node u: Where sum(v) represents the sum of all path costs from the starting node s to the current node v; h(v,z): distance as a prediction function, used to estimate the path cost from the current node v to the destination z; ω·F res (v): muscle load cost; F res is the force generated by all muscles, ω is a weight factor used to describe the influence of force on the cost function; Considering the step size Δt of the muscle force and the change in the weight coefficient ω, the algorithm checks the nodes closest to the goal and selects the one with the lowest cost f(v).

Images