Methods, devices and electronic equipment for simulating shoulder joint range of motion

Abstract

This invention relates to the field of computer technology, providing a method, apparatus, and electronic device for simulating shoulder joint range of motion. The method includes: reconstructing a three-dimensional model of the humerus, a three-dimensional model of the scapula, and a three-dimensional model of the thoracic cage of a target object; expanding the surface of the thoracic cage three-dimensional model outward according to body shape parameters to generate a virtual soft tissue envelope surface; establishing a unified coordinate system between the scapular three-dimensional model and the thoracic cage three-dimensional model; driving the humeral three-dimensional model to perform motion simulation relative to the scapular three-dimensional model under the unified coordinate system; performing collision detection on the humeral three-dimensional model and the virtual soft tissue envelope surface during the motion simulation to obtain collision detection results; and generating the shoulder joint range of motion simulation results of the target object based on the collision detection results. This invention introduces global physical constraints that include the influence of the rigid boundary of the torso and the thickness of the soft tissue on the body surface, avoiding the problem of severely inflated range of motion assessment data caused by independent simulation in a suspended state in related technologies.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a method, apparatus, and electronic device for simulating shoulder joint range of motion. Background Technology

[0002] Shoulder replacement surgery is an effective treatment for shoulder diseases. With technological advancements, clinical evaluation standards have been elevated to a comprehensive assessment of patients' postoperative functional mobility.

[0003] Currently, independent "scapula-humerus" dual-body models are typically constructed based on 3D computed tomography (CT) images for range of motion simulation and collision detection. However, for patients with high body mass index, the virtual arm may pass through the chest cavity during simulation, resulting in significantly inflated range of motion data that fails to reflect the actual physiological limitations after surgery. Summary of the Invention

[0004] This invention provides a method, apparatus, and electronic device for simulating shoulder joint range of motion, in order to overcome the deficiencies in the prior art.

[0005] This invention provides a method for simulating shoulder joint range of motion, comprising the following steps: Based on the medical imaging data of the target object, reconstruct the three-dimensional model of the humerus, the three-dimensional model of the scapula, and the three-dimensional model of the thoracic cage of the target object; Based on the body shape parameters of the target object, the surface of the three-dimensional thoracic model is expanded outward to generate a virtual soft tissue envelope surface for characterizing subcutaneous soft tissue. Establish a unified coordinate system between the 3D model of the scapula and the 3D model of the thorax; Under the unified coordinate system, the motion simulation of the humeral 3D model relative to the scapular 3D model is performed. During the motion simulation, collision detection is performed between the three-dimensional model of the humerus and the virtual soft tissue envelope to obtain the collision detection results; Based on the collision detection results, a simulation result of the shoulder joint range of motion of the target object is generated.

[0006] According to a method for simulating shoulder joint range of motion provided by the present invention, the method includes, during the motion simulation process, performing collision detection between the three-dimensional model of the humerus and the virtual soft tissue envelope to obtain collision detection results, including: The 3D model of the scapula and the prosthesis model disposed on the side of the 3D model of the scapula are set as the first-level stator collision body; The virtual soft tissue envelope is set as the second-level stator collider; During the motion simulation of the humeral 3D model, the internal hard collision between the humeral 3D model and the first-stage stator collision body, and the external soft collision between the humeral 3D model and the second-stage stator collision body are detected in parallel. The triggering states of the internal hard collision and the external soft collision are recorded as the collision detection results.

[0007] According to a method for simulating shoulder joint range of motion provided by the present invention, generating a simulation result of the shoulder joint range of motion of the target object based on the collision detection result includes: If the collision detection result is that the internal hard collision is triggered, it is determined that the current simulated motion is restricted by the bony structure inside the joint, and a first-class feedback information indicating that it belongs to the surgical planning defect is generated as the simulation result of the shoulder joint range of motion. If the collision detection result is that the external soft collision is triggered, it is determined that the current simulated motion is limited by the body surface boundary, and a second category feedback information indicating that it belongs to the physiological structural limit of the target object is generated as the simulation result of the shoulder joint range of motion.

[0008] According to a method for simulating shoulder joint range of motion provided by the present invention, the step of expanding the surface of the three-dimensional thoracic model outward based on the body shape parameters of the target object to generate a virtual soft tissue envelope surface for characterizing subcutaneous soft tissue includes: Obtain the set of surface mesh vertices of the 3D model of the thoracic cavity; Determine the thickness weights of different anatomical regions in the three-dimensional thoracic model; Based on the thickness weight and the body shape parameter, determine the target outward expansion distance corresponding to each surface mesh vertex in the surface mesh vertex set; Each surface mesh vertex is moved outward along its respective surface normal direction by the corresponding target expansion distance to obtain the expanded discontinuous mesh boundary; The virtual soft tissue envelope surface is generated by reconstructing the surface based on the discontinuous mesh boundary.

[0009] According to a method for simulating shoulder joint range of motion provided by the present invention, the step of reconstructing a surface based on the discontinuous mesh boundary to generate the virtual soft tissue envelope surface includes: For the gaps formed by the expansion of adjacent ribs in the discontinuous mesh boundary, voxel-level space filling processing is performed to connect the discrete discontinuous mesh boundaries. The boundaries of the connected meshes are smoothed to generate a closed and continuous outer surface, which serves as the virtual soft tissue envelope.

[0010] According to a method for simulating shoulder joint range of motion provided by the present invention, the step of driving the humeral three-dimensional model to perform motion simulation relative to the scapular three-dimensional model in the unified coordinate system includes: Obtain the spatial target position of the human hand end corresponding to preset daily life function movements; Establish a motion trajectory from the initial set posture to the spatial target position of the human hand end; While keeping the relative positions of the thoracic 3D model and the scapular 3D model locked, the humeral 3D model is driven along the motion trajectory to simulate the execution of the daily life function movements; The humeral 3D model is driven relative to the scapular 3D model to perform a single-plane extreme angle sweep motion simulation along the flexion, extension, abduction, adduction, external rotation, and internal rotation directions, respectively.

[0011] According to a method for simulating shoulder joint range of motion provided by the present invention, the method includes, during the motion simulation process, performing collision detection between the three-dimensional model of the humerus and the virtual soft tissue envelope to obtain collision detection results, including: Construct a first-direction bounding box tree structure that covers the 3D model of the humerus, and a second-direction bounding box tree structure that covers the envelope of the virtual soft tissue. During the motion simulation, the first bounding box in the first bounding box tree structure and the second bounding box in the second bounding box tree structure are obtained in the current test frame. An overlap test is performed on the first directional bounding box and the second directional bounding box; If the first directional bounding box overlaps with the second directional bounding box, collision detection is performed on the overlapping area at the surface triangular mesh level to obtain the collision detection result.

[0012] According to the present invention, a method for simulating shoulder joint range of motion, wherein the reconstruction of a three-dimensional model of the humerus, a three-dimensional model of the scapula, and a three-dimensional model of the thoracic cage based on medical imaging data of the target object includes: The medical image data is subjected to voxel-level semantic segmentation to extract the raw data of the humerus, the raw data of the scapula, and the raw data of the thoracic cavity including spinal segments and ribs. The 3D model of the humerus is generated based on the original humerus data, and the 3D model of the scapula is generated based on the original scapula data; The original thoracic data is processed by filling internal voids and removing external isolated noise to form a three-dimensional mesh of a single connected domain without residual internal noise, and the three-dimensional mesh of the single connected domain is used as the three-dimensional model of the thoracic cage.

[0013] The present invention also provides a shoulder joint range of motion simulation device, comprising the following modules: The reconstruction module is used to reconstruct a three-dimensional model of the humerus, a three-dimensional model of the scapula, and a three-dimensional model of the thoracic cage based on the medical imaging data of the target object. An extension module is used to extend the surface of the three-dimensional thoracic model outward based on the body shape parameters of the target object, and generate a virtual soft tissue envelope surface for characterizing subcutaneous soft tissue. A module is established to create a unified coordinate system between the 3D model of the scapula and the 3D model of the thoracic cavity; The simulation module is used to drive the motion simulation of the humeral three-dimensional model relative to the scapular three-dimensional model in the unified coordinate system. The detection module is used to perform collision detection between the 3D model of the humerus and the virtual soft tissue envelope during motion simulation, and obtain the collision detection results. The generation module is used to generate a simulation result of the shoulder joint range of motion of the target object based on the collision detection result.

[0014] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the shoulder joint range of motion simulation method as described above.

[0015] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the shoulder joint range of motion simulation method as described above.

[0016] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the shoulder joint range of motion simulation method as described above.

[0017] The shoulder joint range of motion simulation method, device, and electronic device provided by this invention reconstructs a three-dimensional model of the humerus, scapula, and thoracic cage based on medical imaging data of the target object. It then generates a virtual soft tissue envelope surface by expanding the surface of the thoracic cage model outwards according to body shape parameters. This allows for realistic external physical collision detection using this virtual soft tissue envelope surface when driving the humerus three-dimensional model in a unified coordinate system. Because this invention establishes the basic bony contour of the human torso by reconstructing the thoracic cage model, and accurately quantifies the thickness of subcutaneous fat and muscle by expanding the virtual soft tissue envelope surface outwards using the target object's body shape parameters, and uses the entire structure as the external physical interference boundary for collision detection during the humerus three-dimensional model movement simulation in a unified coordinate system to obtain the final shoulder joint range of motion simulation result, this complete process introduces global physical constraints including the rigid boundaries of the torso and the influence of the thickness of the soft tissue on the body surface. This objectively restores the real physiologically constrained environment of the human body and avoids the problem of severely inflated range of motion assessment data caused by independent, suspended simulations in related technologies. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating the shoulder joint range of motion simulation method provided by the present invention.

[0020] Figure 2 This is a flowchart illustrating another shoulder joint range of motion simulation method provided by the present invention.

[0021] Figure 3 This is a schematic diagram of the shoulder joint range of motion simulation device provided by the present invention.

[0022] Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0024] Current preoperative planning systems typically construct independent dual-bone motion models based on CT images, focusing only on collisions between bony structures during range of motion simulations. However, this method ignores the obstructing effects of rigid boundaries of the human torso and the thickness of subcutaneous soft tissue. During simulated movement, the virtual arm passes unimpeded through the chest cavity, resulting in significantly inflated range of motion data and a substantial discrepancy between the virtual plan and the patient's actual postoperative physiological activity limitations.

[0025] To address this, the present invention provides a method for simulating shoulder joint range of motion. This method aims to reconstruct a three-dimensional model of the humerus, scapula, and thorax based on medical imaging data of the target object. It combines body shape parameters to extend the thorax surface outwards, generating a virtual soft tissue envelope. Furthermore, it drives the three-dimensional humeral model to perform motion simulation in a unified coordinate system to complete soft tissue collision detection. This achieves accurate restoration of the real physical boundaries and soft tissue constraints of the human torso, thereby objectively and accurately generating shoulder joint range of motion simulation results that conform to the true physiological limits of the target object.

[0026] In this invention, all actions involving the acquisition of signal information or data are carried out in accordance with the relevant data protection laws and policies of the country where the device is located, and with the authorization granted by the owner of the device.

[0027] Figure 1 This is a flowchart illustrating the shoulder joint range of motion simulation method provided by the present invention, as shown below. Figure 1 As shown, the method includes steps 110, 120 and 130.

[0028] Step 110: Based on the medical imaging data of the target object, reconstruct the three-dimensional model of the humerus, the three-dimensional model of the scapula, and the three-dimensional model of the thoracic cage of the target object.

[0029] Here, the target object can be understood as the entity that requires shoulder joint range of motion simulation. This could be a patient with a shoulder joint disorder or a healthy individual requiring sports performance analysis or ergonomic assessment. Medical imaging data is used to characterize the internal anatomical structure of the target object. For example, for surgical planning scenarios requiring extremely high bone resolution, medical imaging data is typically computed tomography (CT) images; for research scenarios requiring observation of both soft tissue and bone, medical imaging data is typically magnetic resonance imaging (MRI) data.

[0030] As an optional embodiment, medical image data of the target object can be acquired through a wide-area imaging input protocol. For example, computed tomography (CT) scan data of the target object can be read through a medical digital imaging and communication interface. Unlike conventional shoulder CT protocols, this embodiment employs an extended field-of-view protocol to obtain the complete physical boundaries of the torso. The scanning range of the acquired medical image data can longitudinally cover from the level of the first thoracic vertebra to the level of the first lumbar vertebra, and laterally completely cover the outer edges of both rib cages.

[0031] Because medical imaging data often contains complex anatomical structures, multi-label anatomical segmentation can be performed based on medical imaging data to extract the desired target structures. Based on the segmentation results, the segmented data is processed into a 3D mesh to reconstruct the 3D models of the humerus, scapula, and thoracic cage of the target object. The humerus 3D model can include the proximal humerus and humeral shaft; the scapula 3D model includes the scapula as a reference bone; the thoracic cage 3D model can be understood as a rigidly connected whole, including vertebral segments, the manubrium of the sternum, the body of the sternum, and the bilateral ribs.

[0032] Step 120: Based on the body shape parameters of the target object, perform an outward expansion operation on the surface of the 3D model of the thoracic cavity to generate a virtual soft tissue envelope surface for characterizing the subcutaneous soft tissue.

[0033] Specifically, body shape parameters refer to quantitative indicators that objectively reflect the fatness and soft tissue distribution characteristics of a target object. Body shape parameters can include body mass index (BMI), body fat percentage, or chest circumference. Since conventional computed tomography (CT) images only show the skeletal outline, directly using a 3D chest model for collision detection would ignore the physical obstruction of movement by subcutaneous fat and muscle. Similarly, differences in body shape parameters directly determine the thickness of soft tissue. For example, for a target object with a high BMI, the thickness of the chest wall soft tissue may reach 30-50 mm. Based on this, this embodiment expands the surface of the 3D chest model outward according to the target object's body shape parameters, generating a shell with thickness, i.e., a virtual soft tissue envelope, to simulate the contour of a real human torso.

[0034] As an optional embodiment, a region-weighted normal expansion algorithm can be used to expand the 3D thoracic model outward based on the body shape parameters of the target object. Specifically, the vertices of the surface mesh of the 3D thoracic model can be expanded outward along the normal vector direction. To simulate the differences in soft tissue thickness in different parts of the real human body, a non-uniform expansion algorithm can be used. The specific logic of this algorithm can be expressed as the expanded vertex coordinates equal to the initial vertex coordinates plus the product of the expansion coefficient and the initial vertex normal vector. The expansion coefficient can be determined by the body shape parameters of the target object and the anatomical region where the vertex is located.

[0035] Step 130: Establish a unified coordinate system between the 3D model of the scapula and the 3D model of the thoracic cavity.

[0036] Specifically, a unified coordinate system can be understood as a global reference system used to uniformly describe the spatial positional relationships of multiple three-dimensional models. Considering that the target object is usually in a static state during medical image data acquisition, in order to avoid non-physiological relative slippage between models during subsequent simulation, this embodiment establishes a unified coordinate system to incorporate the thoracic 3D model and the scapular 3D model into the same spatial reference system.

[0037] As an optional implementation, the local coordinate system of the scapula 3D model can be defined as the world coordinate system. Since the target object is stationary during acquisition, the position of the thoracic 3D model relative to the scapula 3D model is considered initially rigidly locked. The initial transformation matrix between the two is recorded, thereby establishing a unified coordinate system between the scapula 3D model and the thoracic 3D model, ensuring that the thoracic 3D model always maintains the correct relative physical position with the scapula 3D model during subsequent simulations.

[0038] Step 140: Under a unified coordinate system, drive the 3D model of the humerus to perform motion simulation relative to the 3D model of the scapula.

[0039] Specifically, motion simulation can be understood as the process of displacement or rotation of a 3D model of the humerus relative to a 3D model of the scapula in a 3D digital environment through algorithmic control. In order to evaluate the mobility of the shoulder joint, the 3D model of the humerus is driven to perform motion simulation along a preset motion trajectory or rotation axis in a unified coordinate system, using the 3D model of the scapula as a reference, such as simulating arm abduction, adduction, or flexion.

[0040] Step 150: During the motion simulation, collision detection is performed between the 3D model of the humerus and the virtual soft tissue envelope to obtain the collision detection results.

[0041] Specifically, collision detection refers to the technical means of determining whether the 3D model of the humerus and the virtual soft tissue envelope have spatial intersection or overlap. Considering that without restrictions in motion simulation, the 3D model of the humerus would penetrate the thoracic region without obstruction, leading to severely inflated range of motion data, this embodiment calculates the spatial relationship between the 3D model of the humerus and the virtual soft tissue envelope in real time in each frame driving the motion of the 3D model of the humerus, determines whether they have contact or penetration, and records the state and position information of contact or penetration to obtain the collision detection result.

[0042] Step 160: Based on the collision detection results, generate simulation results of the shoulder joint range of motion of the target object.

[0043] Specifically, the shoulder joint range of motion simulation results can be understood as the final motion limit data or state assessment report calculated after physical space constraints. For example, if the collision detection results show that the 3D model of the humerus touches the virtual soft tissue envelope at a specific angle, then that angle is determined as the motion limit in the current direction. Based on this, this embodiment summarizes the limit data in each motion direction according to the collision detection results, and generates the shoulder joint range of motion simulation results for the target object, providing objective evidence for clinicians or researchers.

[0044] The shoulder joint range of motion simulation method provided in this embodiment reconstructs a three-dimensional model of the humerus, scapula, and thoracic cage based on the medical imaging data of the target object. It then generates a virtual soft tissue envelope surface by expanding the surface of the thoracic cage model outwards according to body shape parameters. This allows for realistic external physical collision detection using this virtual soft tissue envelope surface when driving the humerus three-dimensional model in a unified coordinate system. Because this embodiment establishes the basic bony contour of the human torso by reconstructing the thoracic cage model, and accurately quantifies the thickness of subcutaneous fat and muscle by expanding the virtual soft tissue envelope surface outwards using the target object's body shape parameters, and uses the entire structure as the external physical interference boundary for collision detection during the humerus three-dimensional model movement simulation in a unified coordinate system, the final shoulder joint range of motion simulation result is obtained. This complete process introduces global physical constraints that include the rigid boundaries of the torso and the influence of the thickness of the soft tissue on the body surface, thus objectively restoring the real physiologically restricted environment of the human body and avoiding the problem of severely inflated range of motion assessment data caused by independent, suspended simulations in related technologies.

[0045] Considering that related technologies only perform single-level collision detection on the bones and prostheses within the joint, they cannot distinguish whether joint restriction is due to internal joint impact caused by unreasonable surgical planning or external soft tissue obstruction caused by the target's own body shape. Therefore, this embodiment performs collision detection on the 3D model of the humerus and the virtual soft tissue envelope during motion simulation, obtaining collision detection results, including: The 3D model of the scapula and the prosthetic model configured on the side of the 3D model of the scapula are set as the first-level stator colliders; The virtual soft tissue envelope is set as the second-level stator collider; During the motion simulation of the 3D model of the humerus, the internal hard collision between the 3D model of the humerus and the first-level stator collision body, as well as the external soft collision between the 3D model of the humerus and the second-level stator collision body are detected in parallel. The triggering states of internal hard collisions and external soft collisions are recorded as collision detection results.

[0046] Here, the first-level stator collider can be understood as a physical entity that remains relatively stationary during motion simulation, used to characterize the core bony structures within the joint and its implants. The prosthesis model refers to a digital artificial joint component virtually implanted onto the target bone according to preoperative planning, such as a base or liner on the glenoid side. As an optional embodiment, the 3D model of the scapula and the prosthesis model configured on the side of the 3D model of the scapula can be geometrically fused or spatially grouped and set as the first-level stator collider, primarily used to describe intra-articular constraints.

[0047] After setting the first-level stator collider, considering that the actual motion of the target object is not only limited by the internal structure of the joint, but also by the external constraints of the thoracic skeleton and soft tissues such as subcutaneous fat and muscles, it is necessary to construct an external detection entity independent of the internal structure of the joint in order to accurately reproduce and reflect this macroscopic physical interference environment outside the joint. Based on this, this embodiment sets the virtual soft tissue envelope surface as the second-level stator collider.

[0048] Here, the second-level stator collider can be understood as a physical entity that remains relatively stationary during motion simulation, used to characterize the boundary of the soft tissue outside the human torso. As an optional embodiment, the virtual soft tissue envelope surface generated above can be directly set as the second-level stator collider, mainly used to describe extra-joint constraints.

[0049] After setting up the first-level stator collider and the second-level stator collider, considering that the 3D model of the humerus may come into spatial contact with the internal structure of the joint and the external soft tissue of the joint simultaneously or sequentially in the actual motion trajectory, in order to ensure the high fidelity of the motion simulation and the real-time calculation efficiency, and to avoid interference omissions or logical delays caused by sequential execution of detection, the internal hard collision between the 3D model of the humerus and the first-level stator collider, and the external soft collision between the 3D model of the humerus and the second-level stator collider can be detected in parallel during the motion simulation of driving the 3D model of the humerus.

[0050] Here, internal hard collision refers to the spatial overlap or surface intersection between the moving humeral 3D model side entity and the first-level stator collider. External soft collision refers to the spatial overlap or surface intersection between the moving humeral 3D model side entity and the second-level stator collider. As an optional embodiment, in each frame of the active range test, the collision criterion function under dual constraints can be calculated in real time, and parallel detection can be achieved through the following formula: ; in, Represents the collision criterion function; Indicates the current joint angle; This represents the internal collision detection function; This represents a kinetic assembly containing a three-dimensional model of the humerus; Indicates the first-stage stator collider; Represents a logical OR operation; This represents the external collision detection function; This represents the second-stage stator collider. The above logical OR operation mechanism indicates that any collision occurring is considered a physical interference triggered, and the current direction of motion is terminated. Used to determine the current joint angle Is it feasible?

[0051] After completing the parallel detection, considering that clinicians need to intuitively understand the specific reasons for restricting the target object's mobility in order to dynamically adjust surgical planning parameters, in order to trace and classify the specific interference types that lead to motion termination, the triggering states of internal hard collisions and external soft collisions need to be recorded as collision detection results. Here, the triggering state can be understood as a spatial distance indicator of whether internal hard collisions and external soft collisions actually occurred at a specific moment in the motion simulation.

[0052] As an optional embodiment, the collision criterion function described above can be used for determination. When true, extract the specific distance parameters that cause the motion to terminate. When the calculated distance between the mover assembly containing the 3D model of the humerus and the first-level stator collider is ≤0, record the trigger state of triggering the internal hard collision as the collision detection result; when the calculated distance between the mover assembly containing the 3D model of the humerus and the second-level stator collider is ≤0 and no internal hard collision is triggered, record the trigger state of triggering the external soft collision as the collision detection result.

[0053] Based on the above embodiments, and based on the collision detection results, a simulation result of the shoulder joint range of motion of the target object is generated, including: If the collision detection result is that an internal hard collision is triggered, it is determined that the current simulated motion is restricted by the bony structure inside the joint, and a first-class feedback information indicating that it belongs to the surgical planning defect is generated as the simulation result of shoulder joint range of motion. If the collision detection result indicates that an external soft collision has been triggered, the current simulated motion is determined to be limited by the body surface boundary, and a second category feedback information indicating that it belongs to the physiological structural limit of the target object is generated as the simulation result of shoulder joint range of motion.

[0054] Here, the obstruction of the bony structures within the joint can be understood as the physical phenomenon where the humeral side directly impacts the scapular bone or the glenoid prosthesis during movement, preventing the continuation of motion. The first category of feedback information refers to the generated data indicators or interface prompts that clearly indicate an unreasonable configuration of the current prosthesis parameters and suggest modifications to the treatment plan.

[0055] As an optional embodiment, the minimum spatial Euclidean distance between the mover assembly containing the 3D model of the humerus and the first-stage stator collider can be calculated in real time. .when When an internal hard collision is triggered, it is determined that the current simulated movement is restricted by the bony structure inside the joint, and a first-category feedback message indicating a surgical planning defect is generated as the simulation result of the shoulder joint range of motion. For example, a text report marked as a surgical planning defect can be output on the interface, prompting the doctor that the current prosthesis position is not good and suggesting adjustments to parameters such as prosthesis offset or tilt angle.

[0056] Furthermore, considering that the target subject may have special physiological characteristics such as a high body mass index or severe kyphosis, their thick soft tissue or deformed ribcage may preemptively block arm movement before the bony structures impact. To avoid misjudging such unchangeable physiological limitations as surgical planning errors, thereby guiding the doctor to make ineffective or even harmful over-adjustments, it is necessary to independently characterize the interference caused by soft tissue. Based on this, this embodiment introduces an independent determination branch for external obstruction.

[0057] Specifically, if the collision detection result is that an external soft collision is triggered, it is determined that the current simulated motion is limited by the body surface boundary and a second category feedback information indicating that it belongs to the physiological structural limit of the target object is generated as the simulation result of shoulder joint range of motion.

[0058] Here, "trunk surface boundary obstruction" can be understood as a phenomenon where the humeral side of the body, before touching the internal bones of the joint, experiences movement restriction due to friction between the arm and chest wall or sinking into the subcutaneous fat layer. The second category of feedback information refers to a data indicator or interface prompt generated by the system to clearly indicate that the current activity limitation is caused by the target object's own body shape and is an objective limitation not related to prosthetic position error.

[0059] As an optional embodiment, the minimum spatial Euclidean distance between the mover assembly containing the 3D model of the humerus and the first-stage stator collider can be calculated in real time. And the minimum spatial Euclidean distance between the mobilizing assembly containing the 3D model of the humerus and the virtual soft tissue envelope. .when and Upon detection of an external soft collision, the system determines that the current simulated motion is limited by the body surface boundary and generates a second-category feedback message indicating that the motion falls within the physiological limits of the target subject. This feedback serves as the simulation result for shoulder joint range of motion. For example, a report marked as anatomical and physiological limits can be output on the interface, indicating to the physician that this limitation is caused by the target subject's body shape. If further improvement of the patient's postoperative quality of life is required, special compensatory planning strategies should be adopted to overcome the impact risk caused by trunk deformities or obesity.

[0060] Considering that the distribution of subcutaneous fat and muscle in the real human body is not uniform in thickness, traditional proportional global expansion methods cannot accurately reproduce the true soft tissue morphology of the human torso, leading to distortion in subsequent physical collision detection. To generate a non-uniform soft tissue boundary that accurately matches the personalized physical characteristics and anatomical distribution patterns of the target object, this embodiment expands the surface of the 3D thoracic model outwards based on the target object's body shape parameters, generating a virtual soft tissue envelope surface to characterize the subcutaneous soft tissue, including: Obtain the set of surface mesh vertices of the 3D model of the thoracic cavity; Determine the thickness weights of different anatomical regions in the 3D model of the thoracic cavity; Based on the thickness weight and size parameters, determine the target outward expansion distance corresponding to each surface mesh vertex in the surface mesh vertex set; Each surface mesh vertex is moved outward along its respective surface normal direction by the corresponding target expansion distance to obtain the expanded discontinuous mesh boundary; Surface reconstruction is performed based on discontinuous mesh boundaries to generate virtual soft tissue envelope surfaces.

[0061] Here, the surface mesh vertex set can be understood as a collection of multiple three-dimensional spatial coordinate points that constitute the outermost topological structure of the thoracic 3D model. As an optional embodiment, a thoracic 3D model file generated by 3D reconstruction of medical images can be read, and the three-dimensional coordinate data of all discrete points constituting the surface of the model can be extracted using a geometric topology analysis algorithm and stored as a surface mesh vertex set.

[0062] After obtaining the vertex set of the surface mesh of the 3D thoracic model, considering the significant differences in the thickness of fat and muscle accumulation in different anatomical regions such as the anterior chest, back, and lateral rib areas, it is necessary to assign differentiated adjustment coefficients to different regions to ensure that the generated outward-expanding surface conforms to the actual anatomical distribution of the human body. Based on this, the thickness weights of different anatomical regions of the 3D thoracic model are determined.

[0063] Here, thickness weight can be understood as a numerical coefficient used to characterize the relative thickness ratio of soft tissues in various specific anatomical regions of the thoracic cage three-dimensional model. As an optional embodiment, the thoracic cage three-dimensional model can be divided into regions using a preset anatomical partition template, such as the pectoralis major region, latissimus dorsi region, and rib surface region, and different values ​​can be assigned as thickness weights to each specific anatomical region based on statistical anthropometric data.

[0064] After determining the thickness weights of different anatomical regions in the 3D thoracic model, considering that the final soft tissue thickness is not only affected by the distribution of anatomical regions but also directly related to the overall body shape of the target object, it is necessary to comprehensively consider both factors to calculate the specific displacement in order to achieve personalized thickness quantification. Based on this, this embodiment determines the target outward expansion distance corresponding to each surface mesh vertex in the surface mesh vertex set according to the thickness weights and body shape parameters.

[0065] Here, body shape parameters can be understood as quantitative indicators such as body mass index (BMI) that characterize the fatness or thinness of the target object; target outward expansion distance refers to the absolute length that each vertex of the surface mesh needs to move outward along its spatial normal. As an optional embodiment, the basic soft tissue thickness can be calculated first based on the body shape parameters, specifically using the following formula: ; in, Indicates the base thickness; This represents the function that takes the maximum value. This represents the body shape parameter of the target object, namely the body mass index.

[0066] After determining the target outward expansion distance for each surface mesh vertex in the surface mesh vertex set, considering that distance values ​​alone cannot form spatial geometric deformation, in order to actually construct the outer contour point cloud of soft tissue in three-dimensional space, it is necessary to perform directional displacement operations on the vertices of the basic skeletal model based on the calculated distance. Therefore, in this embodiment, each surface mesh vertex is moved outward along its respective surface normal direction by the corresponding target outward expansion distance to obtain the expanded discontinuous mesh boundary.

[0067] Here, the surface normal direction refers to the spatial vector direction perpendicular to the local microplane where each surface mesh vertex is located; the extended discontinuous mesh boundary refers to a set of discrete point clouds or mesh structures that may have stretching defects formed in space after all vertices have completed differentiated displacements. As an optional embodiment, the spatial displacement operation of each surface mesh vertex can be achieved using the following formula: ; in, In the extended discontinuous mesh boundary, the first... The three-dimensional spatial coordinate vectors of the vertices; Represents the vertices of the surface mesh set. The initial three-dimensional spatial coordinate vectors of the vertices of the surface mesh; Indicates the first The target outward distance corresponding to each vertex of the surface mesh is determined by the vertex coordinates and the corresponding anatomical region. and body shape parameters Joint decision; Indicates the first The unit vector of the surface normal direction corresponding to each vertex of the surface mesh. By adding the displacement vector to the three-dimensional coordinates of the original vertex using this formula, all vertices can be pushed outward in space, forming an expanded discontinuous mesh boundary.

[0068] After obtaining the extended discontinuous mesh boundary, considering that continuous, smooth, and closed 3D surfaces are necessary to ensure the accuracy and stability of intersection calculations in physical collision detection, it is necessary to smooth the discrete boundary and perform topology reconstruction to correct the mesh self-intersection or tearing problems caused by non-uniform displacement operations. Based on this, surface reconstruction is performed on the discontinuous mesh boundary to generate a virtual soft tissue envelope surface.

[0069] Here, the virtual soft tissue envelope refers to a closed and continuous three-dimensional geometric surface formed after smooth reconstruction, used in subsequent simulations as the external collision detection entity boundary representing subcutaneous soft tissue. As an optional embodiment, the Poisson surface reconstruction algorithm can be used to implicitly fit the spatially scattered point cloud of discontinuous mesh boundaries to extract a watertight continuous mesh model. Then, the Laplacian smoothing algorithm is applied to remove local noise and sharp abrupt changes on the surface, ultimately generating a smooth virtual soft tissue envelope.

[0070] Discrete mesh fragments, with their physically discontinuous geometry, cannot accurately represent the continuous surface features of the human torso, which is completely enveloped by muscles and skin. To eliminate structural discontinuities caused by gaps between bones and thus construct a complete physical barrier for the torso, this embodiment introduces voxelization spatial processing techniques to bridge and close discrete boundaries.

[0071] Specifically, in this embodiment, surface reconstruction is performed based on discontinuous mesh boundaries to generate a virtual soft tissue envelope surface, including: For the gaps formed by the expansion of adjacent ribs in the boundary of a discontinuous mesh, voxel-level space filling is performed to connect the discrete discontinuous mesh boundaries. The boundaries of the connected meshes are smoothed to generate a closed and continuous outer surface, which serves as the virtual soft tissue envelope.

[0072] Here, the discontinuous mesh boundary refers to the three-dimensional geometric structure with damage and faults formed after the vertices of each surface mesh in the preceding steps have completed the spatial expansion displacement; the gap space formed after the expansion of adjacent ribs refers to the geometrically dataless region formed after the expansion mapping of the spatial region that originally existed between the upper and lower ribs of the thoracic 3D model; the space filling processing at the voxel level can be understood as the calculation process of assigning entity attributes to the originally empty voxels through a specific mathematical morphology algorithm after the 3D space meshing.

[0073] As an alternative embodiment, the 3D bounding box containing discontinuous mesh boundaries can be divided into a uniform 3D voxel array. Then, using the closing operation in 3D mathematical morphology, a dilation operation is first performed on the voxel array to cause voxels on adjacent extended rib surfaces to overlap and merge in the gap space, followed by an erosion operation of the same scale to restore the original external contour volume. Through this process, empty voxels in the gap space formed after the expansion of adjacent ribs can be effectively identified and filled, thereby connecting the discrete discontinuous mesh boundaries to form a voxel set with unified internal properties.

[0074] After connecting discrete, discontinuous mesh boundaries, it's important to consider that while voxel-level space filling macroscopically bridges structural faults, it microscopically creates noticeable artifacts like stepped or jagged surfaces. Such rough surfaces not only contradict the smooth physiological characteristics of natural transitions in human subcutaneous soft tissue but also easily trigger abrupt changes in collision normals during subsequent physical interference calculations, leading to jitter or misjudgments in collision detection results. To provide a physical boundary that conforms to the real body surface morphology and meets the requirements of high-precision collision detection, geometric smoothing of the aforementioned rough transition surfaces is necessary. Therefore, this embodiment smooths the connected mesh boundaries, generating a closed and continuous outer surface as the virtual soft tissue envelope.

[0075] Here, the connected mesh boundary refers to the preliminary geometric shell with complete topological connections but with stepped artifacts on the surface, extracted after space filling at the voxel level; smoothing refers to the computational process of filtering out high-frequency geometric noise on the surface and reducing local curvature abrupt changes through algorithms without changing the macroscopic volume and core contour of the model; closed and continuous outer surface refers to an ideal three-dimensional watertight surface without any holes or edges and with natural transition of normal vectors.

[0076] As an alternative implementation, the moving cube algorithm can be used to extract isosurfaces from the filled voxel set, thereby converting the voxel data back into polygonal mesh data, i.e., the connected mesh boundary. Subsequently, the Laplacian smoothing algorithm is applied to iteratively adjust the position of each vertex in this connected mesh boundary, moving it towards the geometric center of adjacent vertices. Through multiple smoothing iterations, the jagged features caused by voxelization can be effectively eliminated, ultimately generating a closed and continuous outer surface, which is then output as a virtual soft tissue envelope to the collision detection engine.

[0077] Considering that the core indicator most important to patients after shoulder replacement surgery is not the extreme range of motion in any single direction, but rather the ability to smoothly perform complex daily activities such as combing hair, washing the face, and dressing, traditional preoperative planning often focuses only on extreme sweeps of the anatomical plane, neglecting the simulation of daily functional movements. This results in situations where, although surgery removes restrictions in a single direction, pain or inability to complete movements due to skeletal or soft tissue obstruction still exists in complex daily activities. To ensure that preoperative planning directly serves the patient's final quality of life, this embodiment simulates motion between a 3D model of the humerus and a 3D model of the scapula in a unified coordinate system, including: Obtain the spatial target position of the human hand end corresponding to preset daily life function movements; Establish the motion trajectory from the initial set posture to the spatial target position of the human hand end; While keeping the relative positions of the 3D model of the thoracic cavity and the 3D model of the scapula locked, the 3D model of the humerus is driven along the motion trajectory to simulate the performance of daily life functions. The 3D model of the humerus, relative to the 3D model of the scapula, is used to simulate the extreme angle sweep motion of a single plane along the flexion, extension, abduction, adduction, external rotation, and internal rotation directions.

[0078] Here, the preset daily living function movements refer to the standardized set of movements used in clinical medicine to assess a patient's ability to take care of themselves, such as touching the back of the head, touching the opposite shoulder, or touching the back of the back with the back of the hand; the spatial target position of the human hand can be understood as the specific three-dimensional coordinate point or spatial area that the target object's hand needs to reach in a unified coordinate system when performing the above movements.

[0079] As an optional embodiment, a standard template library containing typical action end position data can be built in. Based on the currently selected preset daily life function action, such as combing hair, the spatial target position of the human hand end corresponding to the action can be directly extracted from the template library.

[0080] After obtaining the spatial target position of the human hand end corresponding to the preset daily life function movement, considering that the three-dimensional movement of the humerus is not instantaneous but a continuous posture change process, in order to avoid severe clipping or collision detection omission when directly jumping to the target position, this embodiment establishes the motion trajectory from the initial set posture to the spatial target position of the human hand end.

[0081] Here, the initial setting posture refers to the spatial pose of the 3D model of the humerus with the arm hanging naturally or in a standard anatomical zero position; the motion trajectory refers to a set of positions and postures of the 3D model of the humerus that change continuously over time in three-dimensional space from the starting point to the ending point in a unified coordinate system. As an optional embodiment, an inverse kinematics algorithm can be used, with the spatial target position of the human hand end as the end constraint and the initial setting posture as the starting state, to generate a continuous and smooth spatial curve through a cubic spline interpolation function as the motion trajectory driving the 3D model of the humerus.

[0082] After establishing the motion trajectory from the initial set posture to the target position of the human hand's end-effector, it was considered that when simulating shoulder joint range of motion, allowing relative sliding between the rib cage and scapula would make it difficult to accurately determine whether the prosthesis placement or the soft tissues of the trunk constituted the decisive limiting factor. Therefore, while keeping the relative positions of the 3D rib cage model and the 3D scapula model locked, the 3D humerus model was driven along the motion trajectory to simulate the performance of daily living functional movements.

[0083] Here, relative position locking can be understood as rigidly constraining all degrees of freedom of the thoracic and scapular 3D models in a unified coordinate system, preventing any relative displacement or rotation during motion simulation. Driving the humeral 3D model along the motion trajectory means that the system changes the posture matrix of the humeral 3D model, causing it to move continuously in space frame by frame along a pre-set path. As an optional embodiment, the thoracic and scapular 3D models are integrated as a whole into the same static rigid body node. Subsequently, the humeral 3D model is attached to the system as the sole motion degree-of-freedom node. The 3D transformation matrix of the humeral 3D model is continuously updated along the established motion trajectory using a time-series interpolation function, thereby simulating the performance of daily life functional movements.

[0084] After completing the simulation of compound trajectories for daily living functions, it was considered that while this could assess a patient's ability to perform specific movements, in clinical surgical planning, surgeons also needed to understand the absolute limits of joint movement in various basic anatomical planes to determine the redundancy of movement after prosthesis placement. To provide more comprehensive, multi-dimensional assessment data, it was necessary to go beyond compound movements and conduct standardized limit range of motion tests. Based on this, a 3D model of the humerus was used relative to a 3D model of the scapula to simulate limit angle sweep movements in a single plane along flexion, extension, abduction, adduction, external rotation, and internal rotation.

[0085] Here, the simulation of the ultimate angle sweep motion in a single plane can be understood as the process of driving the three-dimensional model of the humerus to rotate continuously along a specific axis of rotation in anatomically standard planes such as the coronal, sagittal, or horizontal planes until physical interference is triggered and the motion stops. The directions of flexion, extension, abduction, adduction, external rotation, and internal rotation refer to the directions of movement of the shoulder joint in the six basic degrees of freedom defined by anatomical standards.

[0086] As an optional embodiment, after completing the above-mentioned life movement simulation, the humeral 3D model is reset to the initial set posture. Then, by sequentially setting specific rotation axis vectors, the humeral 3D model is driven to continuously rotate relative to the locked scapular 3D model. For example, it is driven to perform a single-plane limit angle sweep motion simulation along the abduction direction in the coronal plane with a preset step length increment angle, until it stops due to obstruction by the prosthesis or chest soft tissue, and the maximum angle of movement in that direction is recorded. The sweeping in the other five directions is completed in sequence to obtain the complete six-degree-of-freedom limit range of motion. Among them, the range of motion in the flexion direction can be a sagittal plane forward scan, the range of motion in the extension direction can be a sagittal plane backward scan, the range of motion in the abduction direction can be a coronal plane outward scan, the range of motion in the adduction direction can be a coronal plane inward scan, the range of motion in the external rotation direction can be external rotation with the upper arm close to the chest (0° abduction), and the range of motion in the internal rotation direction can be internal rotation with the upper arm close to the chest (0° abduction).

[0087] Considering that in complex 3D medical models, both the humerus 3D model and the virtual soft tissue envelope contain hundreds of thousands or even millions of triangular facets, directly performing pairwise intersection calculations on all triangular facets of these two high-precision models in each frame of the motion simulation would result in an enormous computational burden. Therefore, this embodiment performs collision detection on the humerus 3D model and the virtual soft tissue envelope during the motion simulation, obtaining the collision detection results, including: Construct a first-direction bounding box tree structure that covers the 3D model of the humerus, and a second-direction bounding box tree structure that covers the envelope of the virtual soft tissue. During the motion simulation, the first bounding box in the first bounding box tree structure and the second bounding box in the second bounding box tree structure are obtained in the current test frame. Perform an overlap test on the first-direction bounding box and the second-direction bounding box; If the first bounding box and the second bounding box overlap, collision detection is performed on the overlapping area at the surface triangular mesh level to obtain the collision detection result.

[0088] Here, the first-direction bounding box tree structure can be understood as a tree-like hierarchical geometric structure that divides the space of the 3D model of the humerus into different levels, tightly enclosing the internal mesh and allowing arbitrary rotation of the direction; the second-direction bounding box tree structure refers to a bounding box system with the same hierarchical division characteristics established for the virtual soft tissue envelope surface.

[0089] As an optional embodiment, the vertex distribution covariance matrix of the 3D model of the humerus and the virtual soft tissue envelope can be calculated using the principal component analysis algorithm, and the feature vectors can be extracted as the principal axis directions of the bounding box. Then, the mesh model is recursively subdivided from top to bottom to construct a first-direction bounding box tree structure that covers the 3D model of the humerus and a second-direction bounding box tree structure that covers the virtual soft tissue envelope.

[0090] After constructing the first-direction bounding box tree structure covering the 3D model of the humerus and the second-direction bounding box tree structure covering the virtual soft tissue envelope, considering that motion simulation is a dynamic process that changes over time, it is necessary to extract collision detection nodes in the current pose in real time in order to accurately capture potential interference regions within each discrete time step. Based on this, during the motion simulation, the first-direction bounding box in the first-direction bounding box tree structure and the second-direction bounding box in the second-direction bounding box tree structure are obtained in the current test frame.

[0091] Here, the motion simulation process refers to the sequence of operations that drive the 3D model of the humerus to continuously change its pose along a predetermined trajectory; the current test frame refers to a specific instantaneous state in the time series where collision resolution is in progress; the first bounding box and the second bounding box correspond to the spatial rectangular boundary entities at a specific level in the tree structure.

[0092] As an optional embodiment, when updating to the current test frame, the bound tree structure coordinates of the humerus 3D model are synchronously updated according to the latest transformation matrix of the humerus 3D model, and the first bounding box in the first bounding box tree structure and the second bounding box in the second bounding box tree structure under the current test frame are traversed from the root node to obtain the first bounding box in the first bounding box tree structure and the second bounding box in the second bounding box tree structure as the basic input parameters for subsequent spatial intersection tests.

[0093] After obtaining the first-direction bounding boxes in the first-direction bounding box tree structure and the second-direction bounding boxes in the second-direction bounding box tree structure under the current test frame, considering that directly calculating the intersection of faces is still too time-consuming, in order to quickly filter out absolutely safe spatial regions that are impossible to collide, it is necessary to first perform boundary intersection judgment at a coarse level. Based on this, an overlap test is performed on the first-direction bounding boxes and the second-direction bounding boxes. Here, the overlap test refers to the rapid calculation process of using mathematical theorems to determine whether there is a volume intersection between two spatially regular geometric objects in three-dimensional space.

[0094] After performing overlap tests on the first and second bounding boxes, considering that the bounding boxes only roughly wrap around the underlying mesh, overlap does not necessarily mean that the internal anatomical model has physically penetrated. To obtain high-precision anatomical interference results, it is necessary to further refine the assessment at the lowest surface level within the overlapping area. Therefore, if the first and second bounding boxes overlap, collision detection is performed on the overlapping area at the surface triangular mesh level to obtain the collision detection results.

[0095] Here, the overlapping area refers to the common local space where the bounding boxes of two directions intersect in space; collision detection at the surface triangular mesh level refers to the underlying geometric intersection calculation of line-plane or plane-plane intersection for the most basic triangular facets that constitute the three-dimensional model; the collision detection result refers to the state indicator that ultimately determines whether physical penetration has occurred between real medical models.

[0096] As an optional embodiment, if the first bounding box and the second bounding box overlap, the system continues to traverse the tree structure downwards until the leaf node, extracts the real triangular facets contained in the intersecting leaf node, and then performs collision detection at the surface triangular mesh level on the overlapping area. The collision detection result is obtained by calculating the intersection state between the triangles.

[0097] Considering that a high-precision patient-specific skeletal anatomical model is the geometric basis for all subsequent spatial interference detection and prosthesis matching in the preoperative simulation planning of shoulder joint range of motion, traditional image extraction methods based on a single grayscale threshold struggle to accurately separate individual bones with blurred boundaries in the highly anatomically dense shoulder and chest region. This results in the extracted model exhibiting adhesions or missing parts, severely impacting the realism of the simulation. Therefore, this embodiment reconstructs a three-dimensional model of the humerus, scapula, and thoracic cage based on the target object's medical imaging data, including: Voxel-level semantic segmentation was performed on medical image data to extract raw data of the humerus, scapula, and thoracic cavity including spinal segments and ribs. Generating a 3D model of the humerus based on raw humeral data, and generating a 3D model of the scapula based on raw scapula data; The original thoracic data is processed by filling internal voids and removing external isolated noise to form a three-dimensional mesh of simply connected domains without residual internal noise, and the three-dimensional mesh of simply connected domains is used as the three-dimensional model of the thoracic body.

[0098] Here, medical imaging data refers to a sequence of continuous two-dimensional slice images of a target object obtained through equipment such as computed tomography or magnetic resonance imaging; voxel-level semantic segmentation refers to the classification process of assigning specific anatomical labels to each basic volume unit in the three-dimensional image space; raw data of the humerus, raw data of the scapula, and raw data of the thoracic cavity containing spinal segments and ribs refer to the set of discrete pixels with spatial coordinates extracted independently from the global image after semantic label filtering.

[0099] As an optional embodiment, the acquired medical image data can be input into a pre-trained three-dimensional convolutional neural network. Using the probability distribution map output by the network model, masks belonging to different anatomical categories such as the humerus, scapula, and thoracic cage composed of ribs and spine can be extracted one by one to obtain the above three independent raw data.

[0100] After extracting the raw data of the humerus, scapula, and thoracic cavity (including spinal segments and ribs), it was determined that the extracted raw data, mathematically speaking, is merely a discrete set of three-dimensional voxels, lacking continuous topological connectivity representing the object's surface. Therefore, it cannot be directly applied to the subsequent 3D spatial geometry engine involving normal calculations and surface intersection tests. Consequently, it needs to be converted into a standard polygonal mesh surface. Based on this, a 3D model of the humerus is generated from the raw humerus data, and a 3D model of the scapula is generated from the raw scapula data.

[0101] Here, the 3D model of the humerus and the 3D model of the scapula refer to continuous polygonal mesh entities composed of a series of vertices, edges, and surface triangles, used for visualization in 3D virtual space and for participating in geometric collision calculations.

[0102] As an optional embodiment, the moving cube algorithm can be used to extract the isosurfaces from the original data of the humerus and the original data of the scapula, respectively, and convert the discrete voxel field into an initial surface mesh composed of continuous triangular facets, thereby generating three-dimensional models of the humerus and the scapula for characterizing independent bony structures.

[0103] During or after generating a 3D model of the humerus based on the original humerus data and a 3D model of the scapula based on the original scapula data, considering that the original thoracic data contains numerous intercostal spaces and a complex hollow structure of the spinal canal, directly using conventional algorithms to extract the surface would result in the generated model being filled with a large number of meaningless micropores and debris floating outside due to scanning artifacts. These topological defects would greatly disrupt the normal consistency and volume calculation stability when subsequently expanding outward to generate the soft tissue envelope.

[0104] Next, the original thoracic data is processed by filling internal voids and removing external isolated noise to form a three-dimensional mesh of simply connected domains without residual internal noise, and the three-dimensional mesh of simply connected domains is used as the three-dimensional model of the thoracic body.

[0105] Here, internal void filling refers to the algorithmic process of identifying and filling closed or semi-closed cavities inside the model caused by gaps in bone trabeculae or imaging artifacts; external isolated noise removal refers to the process of removing small, independent voxel groups that are detached from the main skeletal structure; and a single-connected 3D mesh refers to a high-quality mesh structure in which there is an effective path connecting any two mesh vertices, and the mesh is completely solid inside without any free fragments or cavities.

[0106] As an optional embodiment, the original thoracic data can first be expanded and eroded using three-dimensional morphological closing operations to fill internal cavities. Then, the volume and area of ​​all independent clumps can be calculated using a connected component labeling algorithm. Clumps with volumes smaller than a preset threshold are filtered out to remove external isolated noise. Finally, the surface of the largest retained clump is extracted to generate a single connected 3D mesh with no residual noise inside, which is then used as the 3D model of the thoracic cavity for subsequent soft tissue expansion calculations.

[0107] Figure 2 This is a flowchart illustrating another shoulder joint range of motion simulation method provided by the present invention, as shown below. Figure 2 As shown, firstly, medical imaging data of the target object, including the entire thoracic cavity and shoulder joint, is input, and data reconstruction is performed. That is, based on the medical imaging data of the target object, the three-dimensional models of the humerus, scapula, and thoracic cavity of the target object are reconstructed, and the spatial alignment of each three-dimensional model is completed in a unified coordinate system.

[0108] Then, soft tissue envelope processing is performed, which involves expanding the surface of the 3D model of the thoracic cavity outward based on the body shape parameters of the target object to generate a closed and continuous virtual soft tissue envelope surface for characterizing the subcutaneous soft tissue.

[0109] Next, a collision model is constructed. The 3D model of the scapula and the prosthesis model configured on the side of the 3D model of the scapula are set as the first-level stator collision body, and the virtual soft tissue envelope is set as the second-level stator collision body. At the same time, the 3D model of the humerus and the prosthesis model configured on the side of the 3D model of the humerus are used as the mover in the motion simulation.

[0110] Subsequently, during the functional simulation, real-time collision detection was performed. Specifically, during the motion simulation of the humeral 3D model relative to the scapular 3D model, internal hard collisions between the humeral 3D model and the first-level stator collision body, and external soft collisions between the humeral 3D model and the second-level stator collision body were detected in parallel. The triggering states of internal hard collisions and external soft collisions were recorded as collision detection results for collision type determination. If an internal hard collision was triggered, it was determined to be a planning defect requiring prosthesis adjustment, and a first-classification feedback message indicating a surgical planning defect was generated. If an external soft collision was triggered, it was determined to be a physiological limit imposed by body type, and a second-classification feedback message indicating a physiological structural limit of the target object was generated.

[0111] Finally, functional simulations are performed based on the aforementioned collision detection mechanism, specifically including basic range of motion testing and daily living activity simulation. In the basic range of motion test, the humeral 3D model is driven relative to the scapular 3D model to perform single-plane extreme angle sweep motion simulations along flexion, extension, abduction, adduction, external rotation, and internal rotation directions. In the daily living activity simulation, the spatial target position of the human hand end corresponding to preset daily living functions is obtained, and a motion trajectory from the initial set posture to the spatial target position of the human hand end is established. While maintaining the relative position lock between the thoracic 3D model and the scapular 3D model, the humeral 3D model is driven along the motion trajectory to simulate the execution of daily living functions. Finally, a personalized surgical plan is output as the overall shoulder joint range of motion simulation result.

[0112] The shoulder joint range of motion simulation device provided by the present invention is described below. The shoulder joint range of motion simulation device described below can be referred to in correspondence with the shoulder joint range of motion simulation method described above.

[0113] Based on the above embodiments, Figure 3 This is a schematic diagram of the shoulder joint range of motion simulation device provided by the present invention, as shown below. Figure 3 As shown, the device includes: Reconstruction module 310 is used to reconstruct the three-dimensional model of the humerus, the three-dimensional model of the scapula, and the three-dimensional model of the thoracic cage of the target object based on the medical imaging data of the target object. The extension module 320 is used to perform outward expansion operations on the surface of the three-dimensional model of the thoracic cavity according to the body shape parameters of the target object, and generate a virtual soft tissue envelope surface for characterizing the subcutaneous soft tissue. Module 330 is established to create a unified coordinate system between the 3D model of the scapula and the 3D model of the thoracic cavity. Simulation module 340 is used to drive the motion simulation of the humeral 3D model relative to the scapular 3D model in a unified coordinate system; The detection module 350 is used to perform collision detection between the 3D model of the humerus and the virtual soft tissue envelope during motion simulation, and obtain the collision detection results. The generation module 360 ​​is used to generate simulation results of the shoulder joint mobility of the target object based on the collision detection results.

[0114] Figure 4 This is a schematic diagram of the structure of the electronic device provided by the present invention, such as... Figure 4 As shown, the electronic device may include a processor 410, a communications interface 420, a memory 430, and a communication bus 440, wherein the processor 410, communications interface 420, and memory 430 communicate with each other via the communication bus 440. The processor 410 can call logical instructions in the memory 430 to execute a shoulder joint range of motion simulation method.

[0115] Furthermore, the logical instructions in the aforementioned memory 430 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0116] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the shoulder joint range of motion simulation method provided by the above methods.

[0117] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to perform the shoulder joint range of motion simulation methods provided by the above methods.

[0118] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0119] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0120] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for simulating shoulder joint range of motion, characterized in that, include: Based on the medical imaging data of the target object, reconstruct the three-dimensional model of the humerus, the three-dimensional model of the scapula, and the three-dimensional model of the thoracic cage of the target object; Based on the body shape parameters of the target object, the surface of the three-dimensional thoracic model is expanded outward to generate a virtual soft tissue envelope surface for characterizing subcutaneous soft tissue. Establish a unified coordinate system between the 3D model of the scapula and the 3D model of the thorax; Under the unified coordinate system, the motion simulation of the humeral 3D model relative to the scapular 3D model is performed. During the motion simulation, collision detection is performed between the three-dimensional model of the humerus and the virtual soft tissue envelope to obtain the collision detection results; Based on the collision detection results, a simulation result of the shoulder joint range of motion of the target object is generated; The step of expanding the surface of the 3D thoracic model outward based on the body shape parameters of the target object to generate a virtual soft tissue envelope surface for characterizing subcutaneous soft tissue includes: Obtain the set of surface mesh vertices of the 3D model of the thoracic cavity; Determine the thickness weights of different anatomical regions in the three-dimensional thoracic model; Based on the thickness weight and the body shape parameter, determine the target outward expansion distance corresponding to each surface mesh vertex in the surface mesh vertex set; Each surface mesh vertex is moved outward along its respective surface normal direction by the corresponding target expansion distance to obtain the expanded discontinuous mesh boundary; The virtual soft tissue envelope surface is generated by reconstructing the surface based on the discontinuous mesh boundary.

2. The shoulder joint range of motion simulation method according to claim 1, characterized in that, During the motion simulation, collision detection is performed between the 3D model of the humerus and the virtual soft tissue envelope to obtain collision detection results, including: The 3D model of the scapula and the prosthesis model disposed on the side of the 3D model of the scapula are set as the first-level stator collision body; The virtual soft tissue envelope is set as the second-level stator collider; During the motion simulation of the humeral 3D model, the internal hard collision between the humeral 3D model and the first-stage stator collision body, and the external soft collision between the humeral 3D model and the second-stage stator collision body are detected in parallel. The triggering states of the internal hard collision and the external soft collision are recorded as the collision detection results.

3. The shoulder joint range of motion simulation method according to claim 2, characterized in that, The step of generating a simulation result of the shoulder joint range of motion of the target object based on the collision detection result includes: If the collision detection result is that the internal hard collision is triggered, it is determined that the current simulated motion is restricted by the bony structure inside the joint, and a first-class feedback information indicating that it belongs to the surgical planning defect is generated as the simulation result of the shoulder joint range of motion. If the collision detection result is that the external soft collision is triggered, it is determined that the current simulated motion is limited by the body surface boundary, and a second category feedback information indicating that it belongs to the physiological structural limit of the target object is generated as the simulation result of the shoulder joint range of motion.

4. The shoulder joint range of motion simulation method according to any one of claims 1 to 3, characterized in that, The step of reconstructing the surface based on the discontinuous mesh boundary to generate the virtual soft tissue envelope includes: For the gaps formed by the expansion of adjacent ribs in the discontinuous mesh boundary, voxel-level space filling processing is performed to connect the discrete discontinuous mesh boundaries. The boundaries of the connected meshes are smoothed to generate a closed and continuous outer surface, which serves as the virtual soft tissue envelope.

5. The shoulder joint range of motion simulation method according to any one of claims 1 to 3, characterized in that, The step of driving the humeral 3D model to perform motion simulation relative to the scapular 3D model in the unified coordinate system includes: Obtain the spatial target position of the human hand end corresponding to preset daily life function movements; Establish a motion trajectory from the initial set posture to the spatial target position of the human hand end; While keeping the relative positions of the thoracic 3D model and the scapular 3D model locked, the humeral 3D model is driven along the motion trajectory to simulate the execution of the daily life function movements; The humeral 3D model is driven relative to the scapular 3D model to perform a single-plane extreme angle sweep motion simulation along the flexion, extension, abduction, adduction, external rotation, and internal rotation directions, respectively.

6. The shoulder joint range of motion simulation method according to any one of claims 1 to 3, characterized in that, During the motion simulation, collision detection is performed between the 3D model of the humerus and the virtual soft tissue envelope to obtain collision detection results, including: Construct a first-direction bounding box tree structure that covers the 3D model of the humerus, and a second-direction bounding box tree structure that covers the envelope of the virtual soft tissue. During the motion simulation, the first bounding box in the first bounding box tree structure and the second bounding box in the second bounding box tree structure are obtained in the current test frame. An overlap test is performed on the first directional bounding box and the second directional bounding box; If the first directional bounding box overlaps with the second directional bounding box, collision detection is performed on the overlapping area at the surface triangular mesh level to obtain the collision detection result.

7. The shoulder joint range of motion simulation method according to any one of claims 1 to 3, characterized in that, The reconstruction of the three-dimensional models of the humerus, scapula, and thoracic cage of the target object based on the medical imaging data of the target object includes: The medical image data is subjected to voxel-level semantic segmentation to extract the raw data of the humerus, the raw data of the scapula, and the raw data of the thoracic cavity including spinal segments and ribs. The 3D model of the humerus is generated based on the original humerus data, and the 3D model of the scapula is generated based on the original scapula data; The original thoracic data is processed by filling internal voids and removing external isolated noise to form a three-dimensional mesh of a single connected domain without residual internal noise, and the three-dimensional mesh of the single connected domain is used as the three-dimensional model of the thoracic cage.

8. A shoulder joint range of motion simulation device, characterized in that, include: The reconstruction module is used to reconstruct a three-dimensional model of the humerus, a three-dimensional model of the scapula, and a three-dimensional model of the thoracic cage based on the medical imaging data of the target object. An extension module is used to extend the surface of the three-dimensional thoracic model outward based on the body shape parameters of the target object, and generate a virtual soft tissue envelope surface for characterizing subcutaneous soft tissue. A module is established to create a unified coordinate system between the 3D model of the scapula and the 3D model of the thoracic cavity; The simulation module is used to drive the motion simulation of the humeral three-dimensional model relative to the scapular three-dimensional model in the unified coordinate system. The detection module is used to perform collision detection between the 3D model of the humerus and the virtual soft tissue envelope during motion simulation, and obtain the collision detection results. The generation module is used to generate a simulation result of the shoulder joint range of motion of the target object based on the collision detection result; The step of expanding the surface of the 3D thoracic model outward based on the body shape parameters of the target object to generate a virtual soft tissue envelope surface for characterizing subcutaneous soft tissue includes: Obtain the set of surface mesh vertices of the 3D model of the thoracic cavity; Determine the thickness weights of different anatomical regions in the three-dimensional thoracic model; Based on the thickness weight and the body shape parameter, determine the target outward expansion distance corresponding to each surface mesh vertex in the surface mesh vertex set; Each surface mesh vertex is moved outward along its respective surface normal direction by the corresponding target expansion distance to obtain the expanded discontinuous mesh boundary; The virtual soft tissue envelope surface is generated by reconstructing the surface based on the discontinuous mesh boundary.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the shoulder joint range of motion simulation method as described in any one of claims 1 to 7.

Images