A method and device for virtual simulation of individual mandibular movement based on forward and inverse multi-body dynamics

Abstract

The application provides a virtual simulation method and device for individual mandibular movement based on forward and reverse multi-body dynamics, the method combines personalized forward dynamic parameters and personalized reverse dynamic parameters, considers position changes of hyoid bones in a virtual simulation process, and adopts a forward and reverse bidirectional simulation combined mode to perform virtual simulation on mandibular movement of a patient, so that the virtual simulation can maximally predict postoperative movement forms of the patient.

Description

Technical Field

[0001] This application belongs to the field of computer simulation, and specifically relates to a virtual simulation method and device for individual mandibular movement based on forward and inverse multibody dynamics. Background Technology

[0002] Virtual simulation of mandibular movement plays a crucial role in the diagnosis, functional evaluation, dental material testing, treatment of maxillofacial joint diseases, maxillofacial reconstruction, artificial joints, and biomimetic materials. The temporomandibular joint is a vital joint in the maxillofacial region, and it, along with surrounding muscles, determines mandibular movement. While the force and motion parameters of joints, ligaments, tendons, jawbone, and teeth during mandibular movement can be obtained from in vitro experiments, they still differ from those in the real human body. Due to the complexity of functional anatomy and ethical limitations, it is currently impossible to virtually reproduce complete and accurate mandibular movement based on in vivo experiments. The mandible, masticatory muscles, joints, ligaments, and other tissues that perform mandibular movement in the human body can be considered a sophisticated multibody system, and the use of multibody system dynamics methods to study mandibular movement has gained widespread attention. Existing technologies mainly include the following types of simulation models:

[0003] (1) The individualized virtual display model of mandibular movement is only used as a dynamic display of mandibular movement. It does not contain the relationship between force and movement in essence, and cannot be used as a dynamic research model, biomimetic material platform, etc.

[0004] (2) The three-dimensional finite element analysis of mandibular motion and force is an analysis of the forces on the mandibular joint in a certain motion state. Essentially, it is an analysis of passive forces, without active forces or force-driven motion. Furthermore, this method is slow and takes a long time to calculate, and cannot meet the real-time motion requirements.

[0005] (3) The mandibular movement model driven by muscles uses a mechanical model of skeletal muscles and sets mechanical parameters including viscoelasticity. It drives the movement of the mandible by the relationship between muscle force and length and the connection between muscles and bones. The accuracy of the movement results obtained by this scheme is limited by the setting of the muscle physical model. The physical model cannot truly simulate the muscle mechanics and movement of different individuals.

[0006] (4) The mandibular motion model controlled by inverse motion trajectory is a system dynamics parameter of the control system that achieves the motion trajectory by iteratively solving a set motion trajectory. It is mainly used for the control of chewing robots. Its solution results are often far from the actual muscle activity of the human body and are often used for laboratory material analysis and other purposes. It is difficult to apply it to needs that are directly related to human diseases and rehabilitation. Summary of the Invention

[0007] To address the problems in the prior art, this invention provides a method and device for virtual simulation of individual mandibular movements based on forward and inverse multibody dynamics. The method combines personalized forward dynamic parameters and personalized inverse dynamic parameters, considers the positional changes of the hyoid bone during the virtual simulation process, and uses a combination of forward and inverse bidirectional simulation to virtually simulate the patient's mandibular movements. This virtual simulation can predict the patient's postoperative movement morphology to the greatest extent possible.

[0008] The purpose of this application is to provide the following aspects:

[0009] In a first aspect, this application provides a virtual simulation method for individual mandibular movements based on forward and inverse multibody dynamics, the method comprising:

[0010] Obtain personalized forward dynamic parameters;

[0011] Personalized inverse dynamics parameters are obtained through inverse dynamics methods;

[0012] Dynamic simulation data is generated using a forward dynamics method based on the personalized forward dynamics parameters and the personalized inverse dynamics parameters.

[0013] In one feasible approach, obtaining the personalized positive dynamic parameters includes:

[0014] Collect surface electromyography (EMG) values;

[0015] The surface electromyography values ​​are preprocessed to generate positive kinetic parameters.

[0016] Further, the preprocessing of the surface electromyography values ​​includes:

[0017] The collected surface electromyography data were then corrected.

[0018] The surface electromyography (EMG) after the corrected surface is filtered.

[0019] Obtain the maximum value of the mean of the filtered data within each time window;

[0020] The surface muscle data are normalized based on the maximum value of the mean of the filtered data within each time window.

[0021] In one feasible approach, obtaining the personalized inverse dynamics parameters includes:

[0022] Obtain mandibular kinematic parameters, including hyoid bone kinematic parameters;

[0023] Obtain skeletal muscle geometry parameters;

[0024] Inverse dynamic parameters are generated based on the mandibular kinematic parameters and the skeletal muscle geometric parameters.

[0025] Furthermore, obtaining the mandibular kinematic parameters includes:

[0026] Collect the movement trajectory of the mandible;

[0027] Calculate the centroid trajectory r and rotation matrix R based on the mandibular motion trajectory;

[0028] The motion trajectory of any point on the mandible in the laboratory coordinate system is generated based on the centroid motion trajectory r and the rotation matrix R.

[0029] Furthermore, obtaining the skeletal muscle geometry parameters includes:

[0030] Obtain geometric parameters of the orofacial muscles;

[0031] Obtain the geometric parameters of the skeleton.

[0032] Optionally, obtaining the skeletal geometric parameters includes:

[0033] Acquire CBCT data;

[0034] Reconstructing a maxillofacial skeletal anatomical model based on CBCT data;

[0035] Skeletal geometric parameters are generated based on the maxillofacial skeletal anatomical model.

[0036] Optionally, obtaining the geometric parameters of the orofacial muscles includes:

[0037] Establish an anatomical model of the masticatory muscles;

[0038] Establish a skeletal muscle anatomical model;

[0039] An anatomical model of the orofacial muscles is generated based on the masticatory muscle anatomical model and the skeletal muscle anatomical model.

[0040] Skeletal and muscular anatomical data were obtained based on the anatomical model of the orofacial muscles.

[0041] Furthermore, the establishment of the masticatory muscle anatomical model includes:

[0042] Obtain an anatomical model of the maxillofacial skeleton;

[0043] An anatomical model of the masticatory muscles was established based on the maxillofacial skeletal anatomical model.

[0044] Optionally, generating the orofacial muscle group anatomical model based on the masticatory muscle anatomical model and the skeletal muscle anatomical model includes:

[0045] Determine the muscle attachment points in the masticatory muscle anatomical model;

[0046] Determine the axis of the muscle section in the skeletal muscle anatomical model;

[0047] Connect the muscle attachment point to the axis of the muscle cross section.

[0048] In one feasible approach, generating dynamic simulation data based on the personalized forward dynamic parameters and the personalized inverse dynamic parameters includes:

[0049] Establish a personalized virtual simulation model of the musculoskeletal system;

[0050] Obtain the personalized forward dynamic parameters and the personalized inverse dynamic parameters;

[0051] Based on the personalized forward dynamic parameters and the personalized inverse dynamic parameters, dynamic simulation data is generated using the personalized musculoskeletal system virtual simulation model.

[0052] Secondly, this application also provides a virtual simulation device for individual mandibular movement based on forward and inverse multibody dynamics, the device comprising:

[0053] Personalized forward dynamic parameters are used to obtain personalized forward dynamic parameters.

[0054] A personalized inverse dynamics parameter acquisition module is used to acquire personalized inverse dynamics parameters through inverse dynamics methods;

[0055] The dynamics simulation data generation module is used to generate dynamics simulation data using a forward dynamics method based on the personalized forward dynamics parameters and the personalized inverse dynamics parameters.

[0056] In one possible implementation, the personalized positive dynamics parameter acquisition module is specifically used for:

[0057] Collect surface electromyography (EMG) values;

[0058] The surface electromyography values ​​are preprocessed to generate positive kinetic parameters.

[0059] Further, the preprocessing of the surface electromyography values ​​includes:

[0060] The collected surface electromyography data were then corrected.

[0061] The surface electromyography (EMG) after the corrected surface is filtered.

[0062] Obtain the maximum value of the mean of the filtered data within each time window;

[0063] The surface muscle data are normalized based on the maximum value of the mean of the filtered data within each time window.

[0064] In one possible implementation, the personalized inverse dynamics parameter acquisition module is specifically used for:

[0065] Obtain mandibular kinematic parameters, including hyoid bone kinematic parameters;

[0066] Obtain skeletal muscle geometry parameters;

[0067] Inverse dynamic parameters are generated based on the mandibular kinematic parameters and the skeletal muscle geometric parameters.

[0068] Furthermore, obtaining the mandibular kinematic parameters includes:

[0069] Collect the movement trajectory of the mandible;

[0070] Calculate the centroid trajectory r and rotation matrix R based on the mandibular motion trajectory;

[0071] The motion trajectory of any point on the mandible in the laboratory coordinate system is generated based on the centroid motion trajectory r and the rotation matrix R.

[0072] Furthermore, obtaining the skeletal muscle geometry parameters includes:

[0073] Obtain geometric parameters of the orofacial muscles;

[0074] Obtain the geometric parameters of the skeleton.

[0075] Optionally, obtaining the skeletal geometric parameters includes:

[0076] Acquire CBCT data;

[0077] Reconstructing a maxillofacial skeletal anatomical model based on CBCT data;

[0078] Skeletal geometric parameters are generated based on the maxillofacial skeletal anatomical model.

[0079] Optionally, obtaining the geometric parameters of the orofacial muscles includes:

[0080] Establish an anatomical model of the masticatory muscles;

[0081] Establish a skeletal muscle anatomical model;

[0082] An anatomical model of the orofacial muscles is generated based on the masticatory muscle anatomical model and the skeletal muscle anatomical model.

[0083] Skeletal and muscular anatomical data were obtained based on the anatomical model of the orofacial muscles.

[0084] Furthermore, the establishment of the masticatory muscle anatomical model includes:

[0085] Obtain an anatomical model of the maxillofacial skeleton;

[0086] An anatomical model of the masticatory muscles was established based on the maxillofacial skeletal anatomical model.

[0087] Optionally, generating the orofacial muscle group anatomical model based on the masticatory muscle anatomical model and the skeletal muscle anatomical model includes:

[0088] Determine the muscle attachment points in the masticatory muscle anatomical model;

[0089] Determine the axis of the muscle section in the skeletal muscle anatomical model;

[0090] Connect the muscle attachment point to the axis of the muscle cross section.

[0091] In one feasible approach, generating dynamic simulation data based on the personalized forward dynamic parameters and the personalized inverse dynamic parameters includes:

[0092] Establish a personalized virtual simulation model of the musculoskeletal system;

[0093] Obtain the personalized forward dynamic parameters and the personalized inverse dynamic parameters;

[0094] Based on the personalized forward dynamic parameters and the personalized inverse dynamic parameters, dynamic simulation data is generated using the personalized musculoskeletal system virtual simulation model.

[0095] Thirdly, this application also provides a program for virtual simulation of individual mandibular movements based on forward and inverse multibody dynamics, wherein the program is used to implement the steps of the virtual simulation method for individual mandibular movements based on forward and inverse multibody dynamics described in the first aspect.

[0096] Fourthly, a computer-readable storage medium storing computer instructions that, when executed by a processor, implement the steps of the virtual simulation method for individual mandibular movement based on forward and inverse multibody dynamics described in the first aspect.

[0097] Fifthly, a detection device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to cause the at least one processor to perform the virtual simulation method for individual mandibular movement based on forward and inverse multibody dynamics described in the first aspect above.

[0098] Compared with existing technologies, the method provided in this application is based on personalized inverse dynamic parameters, combined with personalized forward dynamic parameters such as the electromyographic activity of the masticatory muscles and the mandibular movement trajectory directly collected from the individual. By using a combination of forward and inverse methods, the musculoskeletal dynamic relationship is corrected and optimized to obtain a more realistic individualized simulation. Attached Figure Description

[0099] Figure 1 The flowchart illustrates a virtual simulation method for individual mandibular movement based on forward and inverse multibody dynamics provided in this application. Detailed Implementation

[0100] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of methods consistent with some aspects of the invention as detailed in the appended claims.

[0101] The following detailed embodiments illustrate the virtual simulation method and device for individual mandibular movement based on forward and inverse multibody dynamics provided in this application.

[0102] First, a brief introduction to the use cases of this solution will be given.

[0103] Traditional positive individual mandibular motion virtual simulation methods only provide personalized skull models, while the motion trajectories are not personalized. That is, uniform kinematic parameters are embedded in different skull models. However, such simulations have significant distortion and can only serve as illustrative examples. For rigorous medical design, such as designing surgical and postoperative rehabilitation plans based on mandibular motion simulation, the simulation effect is obviously insufficient.

[0104] Musculoskeletal system (MSS) models established based on multibody dynamics methods provide numerical simulation tools for simulating temporomandibular joint (TMJ) movements. In recent years, with advancements in computer technology, software platforms for modeling musculoskeletal systems have been continuously developed. Currently, software such as OpenSim, Anybody, and Lifemod are used for motion simulation of large human joints such as the knee and spine. The mandible, masticatory muscles, joints, ligaments, and other tissues that perform mandibular movements can be considered a sophisticated multibody system, and the use of multibody dynamics methods to study mandibular movements has gained widespread attention among scholars.

[0105] Figure 1This application provides a flowchart illustrating a virtual simulation method for individual mandibular movement based on forward and inverse multibody dynamics. Figure 1 As shown, the method includes the following steps S100 to S300:

[0106] Step S100: Obtain personalized forward dynamic parameters.

[0107] In this example, obtaining personalized positive dynamic parameters includes steps S101 to S102:

[0108] Step S101: Collect surface electromyography values.

[0109] In this example, surface electromyography (EMG) values ​​can be acquired using any feasible method available in the prior art. Due to the large number of muscles in the facial and jaw region, multiple acquisition methods can be combined to collect as many EMG values ​​as possible. It is understood that if the acquisition method can simultaneously acquire other information, it can also be acquired simultaneously in this step without interruption, thereby reducing the complexity of the actual operation and improving processing efficiency.

[0110] For example, surface electromyography (EMG) values ​​and the mandibular movement trajectory required in step S211 of this example can be acquired simultaneously. Specifically, an electronic facebow (e.g., WinJaw, Zebris Medical GmbH, Isny, Germany) can be used to acquire the patient's mandibular movement. This electronic facebow acquires six-degree-of-freedom rigid body motion information based on the principle of ultrasonic sensing. The acquisition process first requires fixing the maxillary plate to the patient's maxillary dentition using silicone rubber (Okabet, AK-Y-1, Okadiwei, China). This maxillary plate has three marker points. Then, the mandibular fork is fixed to the labial surface of the subject's mandibular dentition using self-curing resin (Luxatemp Automix Plus, DMG, Hamburg, Germany), while ensuring that it does not affect the subject's mandibular movement. The skin surface of the most prominent muscle bundles of the masseter and temporalis muscles is wiped with 75% alcohol swabs, and the matching Zebris bipolar electrode pads are attached and connected to the EMG measurement module.

[0111] Motion acquisition was performed using the WinJaw+ software (v1.4.8; Zebris Medical GmbH) compatible with the electronic facebow, and the initial positions of the maxillary splint and the mandibular fork were recorded. The patient performed mandibular border movements such as protrusion, maximum opening, left lateral movement, and right lateral movement, as well as mandibular functional movements such as chewing, according to the instructions of the experimenter. Each movement started from the intercuspal position and returned to the intercuspal position at the end of the movement. Each movement was repeated 3 times, each lasting about 5 seconds. During this process, the Zebris surface electromyograph synchronously recorded the electromyographic changes of the bilateral masseter muscles and temporalis muscles. The movement trajectories of the landmark points on the maxillary splint were saved as XML files, and the surface electromyogram of the chewing muscles was saved as CSV files.

[0112] In addition, the patient also needed to collect the static response values under the maximum occlusion condition, which were also saved as XML files, and the surface electromyogram of the chewing muscles was also saved as CSV files.

[0113] Step S102, preprocess the surface electromyogram values to generate forward kinetic parameters.

[0114] In this example, any method for preprocessing surface electromyogram values in the prior art can be used to preprocess the surface electromyogram values to normalize the directly collected surface electromyogram values, so that they can be applied to the subsequent steps. Hereinafter, an implementation manner is taken as an example to illustrate. In this example, this step may specifically include steps S121 to S124: [[ID=##]] [[ID=##]]

[0115] Step S121, perform rectification processing on the collected surface electromyogram data.

[0116] In the Matlab environment, rectify the surface electromyogram of the temporalis muscle and masseter muscle and other facial and jaw muscles to facilitate subsequent data processing, and map the electrical signal to the muscle activation degree.

[0117] Step S122, filter the surface electromyogram after rectification processing.

[0118] In this example, the result obtained in step S121 is filtered by a fourth-order Butterworth low-pass filter to achieve good noise reduction and obtain the unnormalized analyzed electromyogram. <000027>It can be understood that other filtering schemes in the prior art can also be used for filtering processing.

[0120] Step S12, obtain the maximum value of the mean of the filtered data within each time sliding window.

[0121] In this example, a time window of appropriate length is selected to obtain the average value of the filtered surface electromyography within the time window of that length, that is, the average electromyography within the window. The length of the time window can be specifically set according to the actual situation, for example, it can be set according to the signal length.

[0122] For the same muscle, the maximum value selected from the average electromyographic values ​​within all sliding windows is the maximum voluntary contraction (MVC). It is understood that each muscle corresponds to only one MVC value.

[0123] Step S124: Normalize the surface electromyography data based on the maximum value of the mean of the filtered data within each time window.

[0124] In this example, the MVC value obtained in step S123 is used to normalize the filtered surface electromyography (EMG) obtained in step S122, resulting in an analytical surface EMG time series u between 0 and 1. i (t), 1≤i≤N mus N mus This indicates the number of muscle bundles in the masseter muscle group, thus mapping the degree of muscle activation.

[0125]

[0126] EMG stands for Electromyography.

[0127] In this step, the analytical reference electromyography used in the normalization process can be obtained by performing the same processing as steps S121 to S123 on the static surface electromyography signal of the same muscle measured under maximum occlusion.

[0128] Step S200: Obtain personalized inverse dynamics parameters through the inverse dynamics method.

[0129] In this example, the personalized inverse dynamics parameters are a series of parameters obtained through the inverse dynamics method. These parameters can be further applied to the subsequent forward dynamics model to obtain dynamic simulation data with personalized characteristics, thereby improving the similarity between the simulation results and the real situation.

[0130] Specifically, this step may include the following steps S201 to S203:

[0131] Step S201: Obtain mandibular kinematic parameters, including hyoid bone kinematic parameters.

[0132] In this example, obtaining the mandibular kinematic parameters includes steps S211 to S213:

[0133] Step S211: Collect the mandibular movement trajectory.

[0134] In this example, the mandibular movement trajectory to be acquired in this step can be acquired synchronously in step S101. For details, please refer to step S101, which will not be repeated here.

[0135] It is understandable that the trajectory of the mandible's movement is the position of its volume coordinates that changes over time.

[0136] Step S212: Calculate the centroid motion trajectory r and rotation matrix R based on the mandibular motion trajectory.

[0137] The CT DICOM file is imported into the digitization software, where the movement trajectories of the maxilla and mandible are segmented. Based on the segmented content, three-dimensional models of the maxilla (STL file) and mandible (STL file) are reconstructed respectively.

[0138] Anatomical landmarks such as cusps and marginal ridges on the dentition are selected in the 3D model of the maxilla. Based on the registration of these landmarks, global registration is achieved between the 3D model of the maxilla (STL file) and the intraoral scan STL file. The intraoral scan STL file can be obtained by scanning with an intraoral scanner (Trios, intraoral scanner, 3shape, Denmark), and the intraoral scan STL file records the relative positional relationship between the maxillary furcation and the maxillary dentition.

[0139] Similarly, the intraoral scan STL file is registered with the standard maxillary plate STL file, which is pre-stored in the system, thereby achieving registration between the three-dimensional model of the maxilla and the standard maxillary plate. The solution provided in this application enables the registration of two models that originally had no common parts.

[0140] Furthermore, after CT scans were used to register the maxillary plate with the standard maxillary plate, a jawbone model with trajectory markers was obtained.

[0141] In this example, all registrations can be performed using any existing registration and alignment method, such as multiple marker point alignment, three-dimensional shape alignment, or surface alignment and registration.

[0142] The following example illustrates the basic conversion principle of registration:

[0143] Suppose there exist two related point sets d j and m j Where j ranges from 1 to N, and N represents the number of points in the static 3D model, the relationship between these two point sets is expressed by the following equation (2):

[0144] D j =Rm j +T+V j Equation (2)

[0145] Where R is a standard 3×3 rotation matrix, T is a three-dimensional translation vector, and V... j Given the interference vector, find the optimal transformation [R^,T^] mapping set m. j In d j Typically, it is necessary to minimize a least squares error criterion, which can be expressed by the following equation (3):

[0146]

[0147] The optimal solution to this least squares equation holds if there are no incorrect correspondences among the points in the point set.

[0148] Furthermore, assuming the mandible is in a closed position, the local coordinate systems of the maxilla and mandible coincide. Import the XML trajectory file obtained in step S101, and using the singular value decomposition method, solve for the mandibular centroid motion trajectory r and rigid body rotation matrix R' during mandibular movement. Then, convert the rotation matrix into mandibular posture angles α, β, γ defined based on Euler angles.

[0149] Step S213: Generate the motion trajectory of any point on the mandible in the laboratory coordinate system based on the centroid motion trajectory r and the rotation matrix R.

[0150] This step can employ any existing technology. Based on the mandibular centroid motion trajectory r and the rigid body rotation matrix R, the motion trajectory of any point on the mandible in the laboratory coordinate system is obtained through rigid body kinematic transformation.

[0151] Step S202: Obtain skeletal muscle geometry parameters.

[0152] Further, obtaining the skeletal muscle geometric parameters includes steps S221 and S222:

[0153] Step S221: Obtain bone geometry parameters, including steps S2211 to S2213.

[0154] Step S2211: Obtain CBCT data.

[0155] In this example, CBCT imaging (NewTom VG, NewTom, Italy) can be performed on the subject, with the imaging range extending from the hyoid bone to the glenoid fossa.

[0156] The imaging parameters can be set as needed, for example, a field of view of 15cm*15cm and a slice thickness of 0.3cm. During the scanning process, the subject should maintain an intercuspal occlusion.

[0157] Step S2212: Reconstruct the maxillofacial skeletal anatomical model based on CBCT data.

[0158] In this example, this step can use any existing method for reconstructing the maxillofacial skeletal anatomy model from CBCT data, for example, the following methods can be used:

[0159] The CBCT-Dicom images obtained in step S2211 are imported into Mimics 17.0 (Mimics, V17.0.0.436, Materialise, Belgium) software. Threshold segmentation technology is used to segment CT images from 1250 HU to 4095 HU. The cervical spine images below the atlantoaxial joint are erased. The images of the maxilla, mandible, and hyoid bone are separated from the remaining CT images according to anatomical landmarks. 3D reconstruction is performed according to the High Quality option in the software. The reconstructed skeletal anatomical models are then output as STL format files.

[0160] Furthermore, the obtained skeletal anatomy STL model was imported into INSIIDES multibody dynamics software. In this software, three rigid bodies—the skull, mandible, and hyoid bone—were defined. For each rigid body, the mass m and the moment of inertia matrix I can be determined according to… The standard jaw model in the AMMR model library is linearly scaled, and the scaling ratio Scale is obtained by comparing the standard jaw model with the measured center distance of the left and right temporomandibular joints.

[0161] Step S2213: Generate bone geometric parameters based on the maxillofacial skeletal anatomical model.

[0162] In this example, this step can be performed using any of the existing technologies.

[0163] Step S222: Obtain the geometric parameters of the orofacial muscles.

[0164] Optionally, this step may include steps S2221 to S2224:

[0165] Step S2221: Establish an anatomical model of the masticatory muscles.

[0166] Optionally, establishing the masticatory muscle anatomical model includes:

[0167] Step S22211: Obtain the maxillofacial skeletal anatomical model;

[0168] In this example, the maxillofacial skeletal anatomical model established in the aforementioned steps can be used, specifically the maxillofacial skeletal anatomical model established in step S2212.

[0169] Step S22212: Establish a masticatory muscle anatomical model based on the maxillofacial skeletal anatomical model.

[0170] In this example, the origin and insertion points of each muscle are determined in the maxillofacial skeletal anatomical model, and the masticatory muscle is inserted into the maxillofacial skeletal anatomical model.

[0171] Specifically, the inserted masticatory muscles include: the anterior, middle, and posterior bundles of the temporalis muscle, the upper and lower layers of the lateral pterygoid muscle, the deep and superficial layers of the masseter muscle, the medial pterygoid muscle, the anterior bundle of the digastric muscle, the anterior and posterior bundles of the mylohyoid muscle, and the geniohyoid muscle.

[0172] Furthermore, the origin and insertion points of each muscle can be located according to existing anatomical atlases of muscles, for example, precisely located according to the anatomical atlas edited by Gray and Sicher, and marked on the STL model of maxillofacial skeletal geometry by professionals with anatomical background knowledge.

[0173] Step S2222: Establish a skeletal muscle anatomical model.

[0174] In this example, any existing method can be used to establish the skeletal muscle anatomical model in this step. For example, the flexible cable finite element modeling proposed by Guo et al. (Multibody Syst Dyn (2020) 49:315–336) can be used. In this scheme, each finite element contains two nodes (1, 2), and the motion of each node is based on a partial Lagrangian-Eulerian description, thus the generalized coordinates of the element are taken as... Where, r a p is the spatial coordinate of the node in the laboratory coordinate system. a It is the material coordinate of the muscle mass, where 'a' is the number of nodes, which can be 1 or 2.

[0175] Step S223: Generate an anatomical model of the jaw muscles based on the masticatory muscle anatomical model and the skeletal muscle anatomical model.

[0176] In this example, this step may include:

[0177] Step S2231: Determine the muscle attachment points in the masticatory muscle anatomical model.

[0178] In this example, this step can be determined together with the aforementioned step S22212.

[0179] Step S2232: Determine the axis of the muscle section in the skeletal muscle anatomical model.

[0180] In this step, the attachment points determined in step S2231 are connected, wherein the path of the cable unit is defined as the axis of the corresponding muscle section.

[0181] Alternatively, muscle internal forces can be modeled as the Hill model, where the maximum active muscle force of each muscle bundle is... Values ​​are derived from the standard model based on Scale. 1.5 Scaling yields the standard model, which can be a pre-stored model, and includes the active muscle force F generated by active muscle contraction. CE And the passive muscle force F generated by the passive elongation of muscle fibers PE .

[0182] In this step, the geometric model of the temporomandibular joint is modeled as a flexible contact between the condyle and the articular surface. The temporomandibular joint contact model is set as a soft matter contact function established by Flores, wherein the normal contact force... And ignore the tangential friction between the joint surfaces.

[0183] Where δ represents the embedding depth between the articular surfaces. χ represents the embedding velocity between the joint surfaces, K represents the normal contact stiffness, n represents the nonlinear exponent of the contact force, and χ is determined by the restitution coefficient of the joint surface contact.

[0184] In this example, the condylar contact geometry can be simplified to an ellipsoid. In particular, the ellipsoidal geometric parameters are obtained by least-squares fitting of the actual condylar geometry; the contact surface of the temporomandibular joint is obtained by fitting the actual articular surface into a spline swept surface.

[0185] In addition, a rigid Hertz contact model was established between the upper and lower central incisors and the first molars to avoid interlocking of the upper and lower teeth during occlusion.

[0186] Step S2233: Connect the muscle attachment point to the axis of the muscle cross section.

[0187] In this example, this step can be done manually or automatically using a computer.

[0188] Step S224: Obtain skeletal and muscular anatomical data based on the orofacial muscle group anatomical model.

[0189] In this example, the skeletal anatomical data includes the locations of multiple feature points on the bones and the origin and termination points of each muscle.

[0190] In this example, this step can be performed using any of the existing technologies to obtain the corresponding skeletal and muscular anatomical data from the orofacial muscle group anatomical model.

[0191] Step S203: Generate inverse dynamic parameters based on the mandibular kinematic parameters and the skeletal muscle geometric parameters.

[0192] In this example, this step can specifically involve obtaining the time series of the motion trajectory and posture angle of the mandibular centroid.

[0193] In this example, the length of the muscle fibers of the digastric muscle and other depressor muscles is determined by the time-varying positions of the hyoid bone and mandible. In particular, this example considers the change in the length of the depressor muscles caused by hyoid bone traction, which cannot be ignored.

[0194] To obtain a personalized hyoid bone movement trajectory, this example performs interpolation on the measured hyoid bone position line relative to the mandible in the closed and maximum open positions, thereby obtaining the hyoid bone position corresponding to the mandibular posture angle.

[0195] Given the aforementioned skeletal movements, the distance between the attachment points of the depressor muscles on the mandible and hyoid bone is calculated in reverse, serving as the target muscle fiber length λ for further forward calculation. i , 1≤i≤N mus , where N mus This indicates the number of muscle fibers in each model.

[0196] Step S300: Generate dynamic simulation data using the forward dynamics method based on the personalized forward dynamics parameters and the personalized inverse dynamics parameters.

[0197] In this example, we consider the sagittal tilt θ of the head and neck during a wide-open mouth movement. skull Specifically, we can assume that the head and neck extend backward by 15° when the mouth is at its maximum opening. Understandably, this angle can also be set to other angles to make the simulation results more realistic.

[0198] In this example, the step of generating dynamic simulation data based on the personalized forward dynamic parameters and the inverse dynamic parameters includes steps 301 to S303:

[0199] Step S301: Establish a personalized virtual simulation model of the musculoskeletal system.

[0200] In this example, the personalized musculoskeletal system virtual simulation model can be any available personalized musculoskeletal system virtual simulation model in the prior art, and it is understood that the relevant parameters can be optimized.

[0201] Preferably, the control equations of the virtual simulation model of the personalized musculoskeletal system can be written as differential-algebraic equations as shown in equations (4) and (5), and numerically integrated based on the BDF method:

[0202]

[0203] C(q,t)=0 Equation (5)

[0204] Step S302: Obtain the personalized forward dynamic parameters and the personalized reverse dynamic parameters.

[0205] In this example, the simulated muscle fiber length l is set. i The result of the forward solution for the jaw-depressing muscle group is compared with the target muscle fiber length λ obtained from the reverse simulation in step S203. i The gap e λ (t)=l i -λ i To control the amount, establish and determine the activation level of the depressor jaw muscle. The control equations for the model can be represented by the proportional-derivative controller shown in equation (6):

[0206]

[0207] in, This indicates the proportional coefficient of the control controller. The differential coefficients, denoted by 'a', represent the muscle activation 'a' of the depressor jaw muscles when the simulation model reaches its maximum mouth opening. λ (t) is the largest.

[0208] Furthermore, the analytical electromyographic signal of any muscle bundle can be determined by the first-order muscle activation kinetic equation expressed by the following equation (7):

[0209]

[0210] in, The degree of muscle activation of the jaw-raising muscle group can be determined by analyzing surface electromyography results.

[0211] τ a The nonlinear function representing the normalized analytical electromyography and muscle activation can be specifically represented by the following equation (8):

[0212]

[0213] in, The value range is 0 to 1.

[0214] Step S303: Generate dynamic simulation data using the personalized musculoskeletal system virtual simulation model based on the personalized forward dynamic parameters and the personalized inverse dynamic parameters.

[0215] In this example, the activation level a of the orofacial muscles is... λSubstituting (t) into the flexible muscle finite element model represented by equation (6) to solve for the muscle internal force, thereby obtaining the positive driving mandibular motion data and realizing the positive and negative coupling dynamic simulation of the patient's personalized musculoskeletal system virtual simulation model.

[0216] In this example, after obtaining the dynamic simulation data, this data can be further used to dynamically display the data in a coordinate system. Specifically, the coordinate system is transformed to the original CT coordinate system, the lower incisor points and bilateral condylar points on the mandibular model are selected, and a list of trajectory coordinates of the three points over time in the original CT coordinate system is output. The three-dimensional objects of the virtual model of the maxilla, mandible, and hyoid bone are then output in STL or OBJ format.

[0217] In software that can dynamically display three-dimensional objects, the above-mentioned virtual model three-dimensional object is input through a data interface to determine the lower incisor point and bilateral condylar points on the mandible. The trajectory coordinates of the three points on the mandibular model are then input through the data interface.

[0218] For each time point, the three points determined on the mandibular model are aligned with the trajectory coordinates using the minimum distance method to obtain the position of the mandible. The position of the mandible changes over time at all time points to obtain the display of the mandibular movement trajectory. The virtual model can output the degree of activation of multiple sets of masticatory muscles attached to the mandible as the mandible moves. The activation degree can be reflected by the color intensity, and the activation value can also be viewed in real time through the timeline window.

[0219] The solution provided in this application is based on musculoskeletal dynamic physical parameters, combined with the collectable mandibular movement trajectory and surface masticatory muscle electromyography, to inversely solve the physical parameters of deep muscles such as the length of the depressor muscle group. The simulation is performed using a forward and inverse coupled dynamics method, and the musculoskeletal dynamic relationship is corrected and optimized by a muscle activation controller that combines forward and inverse methods. This results in a multi-body dynamics simulation effect that is closer to the actual mandibular movement trajectory and dynamic characteristics of related masticatory muscles, and even individualized simulation, as well as virtual simulation of the hyoid bone movement position.

[0220] Specifically, the solution provided in this application creates a personalized musculoskeletal system biodynamic model for patients based on methods such as mandibular motion capture, surface electromyography (EMG) acquisition, and CBCT measurement, solving the problem that existing virtual simulation models cannot incorporate real test data from subjects. Based on a flexible multibody dynamics method, it realizes the distribution of orofacial skeletal muscle mass along muscle fibers. The skeletal muscle model includes active and passive muscle force models, solving the problem of calculating the activation of deep muscle groups such as the depressor jaw muscles and medial pterygoid muscles during orofacial movements.

[0221] The present application has been described in detail above with reference to specific embodiments and exemplary examples; however, these descriptions should not be construed as limiting the present application. Those skilled in the art will understand that various equivalent substitutions, modifications, or improvements can be made to the technical solutions and implementation methods of the present application without departing from the spirit and scope of the present application, and all such modifications and improvements fall within the scope of the present application. The scope of protection of the present application is determined by the appended claims.

Claims

1. A method of virtual simulation of individual mandibular movement based on forward and inverse multi-body dynamics, characterized in that, The method comprises: obtaining individualized forward kinetic parameters; obtaining individualized reverse kinetic parameters by a reverse kinetic method; generating kinetic simulation data by a forward kinetic method according to the individualized forward kinetic parameters and the individualized reverse kinetic parameters; wherein the obtaining of the individualized reverse kinetic parameters comprises: obtaining mandibular kinematic parameters, the mandibular kinematic parameters comprising hyoid kinematic parameters; obtaining skeletal muscle geometry parameters; generating reverse kinetic parameters according to the mandibular kinematic parameters and the skeletal muscle geometry parameters; the generating of the reverse kinetic parameters according to the mandibular kinematic parameters and the skeletal muscle geometry parameters comprises: obtaining a time sequence of a motion trajectory of a mandibular barycenter and an attitude angle; determining muscle fiber lengths of a digastric muscle group according to time-varying positions of a hyoid bone and a mandible indicated by the time sequence of the motion trajectory of the mandibular barycenter and the attitude angle; obtaining a hyoid bone position corresponding to a mandibular attitude angle by performing linear interpolation on measured hyoid bone relative mandibular bone position lines in a closed mouth position and a maximum open mouth position; obtaining target muscle fiber lengths by reversely solving distances of the digastric muscle group between mandibular and hyoid bone attachment points; the generating of the kinetic simulation data by the forward kinetic method according to the individualized forward kinetic parameters and the individualized reverse kinetic parameters comprises: establishing an individualized muscle-bone system virtual simulation model; obtaining the individualized forward kinetic parameters and the individualized reverse kinetic parameters; generating kinetic simulation data by using the individualized muscle-bone system virtual simulation model according to the individualized forward kinetic parameters and the individualized reverse kinetic parameters; the obtaining of the individualized forward kinetic parameters and the individualized reverse kinetic parameters comprises: taking a difference between a muscle fiber simulation length and a target muscle fiber length as a control quantity, establishing a control equation for determining a digastric muscle activation degree model, the control equation being a proportional-differential controller, and determining muscle activation degrees of the digastric muscle group by the proportional-differential controller.

2. The method of claim 1, wherein, the obtaining of the individualized forward kinetic parameters comprises: collecting surface electromyography values; preprocessing the surface electromyography values to generate forward kinetic parameters.

3. The method according to claim 1 or 2, characterized in that, the obtaining of the mandibular kinematic parameters comprises: collecting a mandibular motion trajectory; calculating a barycenter motion trajectory r and a rotation matrix R according to the mandibular motion trajectory; generating a motion trajectory of an arbitrary point on a mandibular bone in a laboratory coordinate system according to the barycenter motion trajectory r and the rotation matrix R.

4. The method according to claim 1 or 2, characterized in that, the obtaining of the skeletal muscle geometry parameters comprises: obtaining masticatory muscle group geometry parameters; obtaining skeletal geometry parameters.

5. The method according to claim 1 or 2, characterized in that, further comprising: obtaining a craniofacial bone dissection model; establishing a masticatory muscle dissection model according to the craniofacial bone dissection model.

6. A device for virtual simulation of individual mandibular movement based on forward and inverse multi-body dynamics, characterized in that, the device comprises: an individualized forward kinetic parameter obtaining module, configured to obtain individualized forward kinetic parameters; an individualized reverse kinetic parameter obtaining module, configured to obtain individualized reverse kinetic parameters by a reverse kinetic method; a dynamics simulation data generation module configured to generate dynamics simulation data by using a forward dynamics method according to the personalized forward dynamics parameters and the personalized inverse dynamics parameters; The personalized inverse dynamics parameter acquisition module is configured to perform the following operations: obtain mandibular kinematics parameters, the mandibular kinematics parameters including hyoid kinematics parameters; obtain skeletal muscle geometry parameters; generate inverse dynamics parameters according to the mandibular kinematics parameters and the skeletal muscle geometry parameters; The personalized inverse dynamics parameter acquisition module generates inverse dynamics parameters according to the mandibular kinematics parameters and the skeletal muscle geometry parameters, including: obtain a time sequence of a motion trajectory of a mandibular barycenter and an attitude angle; determine muscle fiber lengths of the mylohyoid muscle group according to time-varying positions of the hyoid bone and the mandible indicated by the time sequence of the motion trajectory of the mandibular barycenter and the attitude angle; obtain hyoid bone positions corresponding to mandibular attitude angles by performing interpolation on measured hyoid bone relative mandibular bone position lines in a closed-mouth position and a maximum open-mouth position; obtain target muscle fiber lengths by inversely solving distances of the mylohyoid muscle group between the mandibular bone and the hyoid bone attachment points; The dynamics simulation data generation module is specifically configured to perform the following operations: establish a personalized musculoskeletal system virtual simulation model; obtain the personalized forward dynamics parameters and the personalized inverse dynamics parameters; generate dynamics simulation data by using the personalized musculoskeletal system virtual simulation model according to the personalized forward dynamics parameters and the personalized inverse dynamics parameters; The dynamics simulation data generation module specifically obtains the personalized forward dynamics parameters and the personalized inverse dynamics parameters, including: determine muscle activation of the mylohyoid muscle group by using a proportional-differential controller.

7. A computer-readable storage medium, characterized in that, The computer instructions are executed by a processor to implement steps of the individual mandibular movement virtual simulation method based on forward and inverse multi-body dynamics according to any one of claims 1 to 5.

8. A detection device, characterized by The detection device includes at least one processor and a memory connected to the at least one processor in communication; wherein the memory stores instructions executable by the processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the individual mandibular movement virtual simulation method based on forward and inverse multi-body dynamics according to any one of claims 1 to 5.

Images