OpenSim-based integrated whole-body musculoskeletal model, construction method and application

Abstract

The invention provides an integrated whole-body musculoskeletal model based on OpenSim, a construction method and application. The method comprises the following steps: integrating a knee joint ligament and a contact structure into a musculoskeletal model containing complete muscles of a trunk and lower limbs to obtain a musculoskeletal model containing the knee joint ligament and the contact structure; and then model adjustment is carried out to obtain an integrated whole-body musculoskeletal model. The integrated whole-body musculoskeletal model can be used for predicting the load response of the anterior cruciate ligament. When a knee joint ligament structure is added, an algorithm and a program for automatically searching grid nodes at ligament and muscle insertion points are designed, so that the knee joint ligaments and muscles can be automatically adjusted and accurately positioned and inserted into a musculoskeletal model containing complete muscles of a trunk and lower limbs; the muscle and the knee joint ligament can be accurately attached to the surface of the skeleton, and accurate integration of the knee joint ligament and the lower limb muscle is achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of biomechanical modeling, relates to a musculoskeletal model, and specifically relates to an integrated whole-body musculoskeletal model based on OpenSim, a construction method and an application thereof. Background Art

[0002] The anterior cruciate ligament (ACL) is a dense connective tissue composed mainly of elastic fibers within the knee joint, connecting the tibia and femur. It primarily limits anterior displacement of the tibia and maintains knee stability. As one of the most common sports injuries, its annual incidence rate reaches 68.6 per 100,000 people, with over 2 million new cases worldwide each year. ACL injury can easily lead to knee instability, meniscus / cartilage damage, and osteoarthritis, seriously affecting function and quality of life. As the core factor of injury, biomechanical mechanisms are often studied using in vitro and in vivo measurement methods, but this is limited by ethics and costs. Advances in computer modeling technology have made musculoskeletal models an important supplement to mechanism research and are of great significance for injury prevention and rehabilitation.

[0003] Current general-purpose basic musculoskeletal models have the following major drawbacks: Most existing lower limb musculoskeletal models fail to fully capture the soft tissue structure of the knee joint, and relatively few models can accurately depict the complex soft tissue structure of the knee joint. While existing knee joint models include soft tissue structures, they lack a complete representation of the entire lower limb musculature, particularly the core trunk muscles.

[0004] Based on the OpenSim platform, it is expected to build an integrated whole-body musculoskeletal model that includes the lower limbs, trunk muscles and complete ligament tissue of the knee joint, but the actual construction faces the following challenges: First, after directly moving the knee ligaments and muscles in the knee joint model to the lower limb musculoskeletal model, it is difficult for the muscles and ligaments to fit accurately on the bone surface. Second, when building an integrated model, it is difficult to achieve accurate prediction of ligament strain. Third, the knee joint degrees of freedom of the existing model only include internal and external valgus, internal and external rotation angles, and anterior and posterior translation of the knee joint, and the degrees of freedom are insufficient. The above problems make it difficult for the model to simulate the physiological structure of the human body to achieve the functional coordination of muscles, ligaments and bones, which in turn limits the application of the model in the prediction of anterior cruciate ligament load response. Summary of the Invention

[0005] In response to the shortcomings of the existing technology, the purpose of the present invention is to provide an integrated whole-body musculoskeletal model based on OpenSim, a construction method and application, to solve the technical problem in the existing technology that after the knee joint and knee ligaments in the knee joint model are directly moved to the lower limb musculoskeletal model, the muscles and ligaments are difficult to fit accurately to the bone surface.

[0006] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0007] A method for constructing an integrated whole-body musculoskeletal model based on OpenSim is characterized in that the method includes specifically: integrating knee joint ligaments and contact structures into a musculoskeletal model including complete muscles of the trunk and lower limbs to obtain a musculoskeletal model including knee joint ligaments and contact structures; then adjusting the model to obtain an integrated whole-body musculoskeletal model.

[0008] Integrating the knee ligaments and contact structures includes locating the knee ligaments and muscles. Specifically, formula I is used to calculate the shortest distance from the insertion point of the knee ligament and muscle to the adjacent bone surface, and then the minimum Euclidean distance is found from all distances. The point on the bone surface corresponding to this distance is the new insertion point of the knee ligament or muscle.

[0009]

[0010] In formula I: d i represents the Euclidean distance between a knee ligament or muscle and the mesh nodes on the surface of n bones where it is inserted; p i =(x i ,y i ,z i ), represents the coordinates of the mesh node on the i-th bone surface; l=(x l ,y l ,z l ), representing the coordinates of a muscle or ligament insertion point.

[0011] The present invention also includes the following technical features:

[0012] The method specifically includes the following steps:

[0013] Step 1: Obtain a musculoskeletal model containing complete muscles of the trunk and lower limbs:

[0014] Step 1.1: Based on the OpenSim human body model database, obtain a lower limb musculoskeletal model that conforms to human physiological characteristics.

[0015] Step 1.2: Based on the OpenSim human body model database, a trunk musculoskeletal model that conforms to human physiological characteristics is obtained.

[0016] In step 1.3, the trunk muscle groups in the trunk musculoskeletal model obtained in step 1.2 are transferred to the lower limb musculoskeletal model obtained in step 1.1 to obtain a musculoskeletal model containing complete muscles of the trunk and lower limbs.

[0017] Step 2: Based on the OpenSim human body model database, a musculoskeletal model including knee joint ligaments and contact structures that conforms to human physiological characteristics is obtained.

[0018] Step 3: Integrate the knee joint ligaments and contact structures of the musculoskeletal model including the knee joint ligaments and contact structures obtained in step 2 into the knee joint of the musculoskeletal model including the complete muscles of the trunk and lower limbs obtained in step 1:

[0019] In step 3.1, the tibia and fibula in the musculoskeletal model containing the knee ligaments and contact structures are scaled synchronously to ensure that the bone structures and their associated ligament attachment points are accurately aligned with the reference dimensions of the musculoskeletal model containing the complete muscles of the trunk and lower limbs.

[0020] In step 3.2, the tibia and fibula in the musculoskeletal model including the knee joint ligaments and contact structures are used to replace the corresponding anatomical structures in the musculoskeletal model including the complete muscles of the trunk and lower limbs.

[0021] In step 3.3, find the same set of bone surface landmarks on the tibia and fibula in the musculoskeletal model including the complete muscles of the trunk and lower limbs and the musculoskeletal model including the knee joint ligaments and contact structures.

[0022] In step 3.4, an affine transformation is used to project the bone marker points in the musculoskeletal model containing the knee ligaments and contact structure to the corresponding bone marker points in the musculoskeletal model containing the complete muscles of the trunk and lower limbs, and then the knee ligament and muscle path points in the musculoskeletal model containing the knee ligaments and contact structure are mapped to the basic musculoskeletal model.

[0023] Step 3.5: Integrate the knee ligaments and muscles.

[0024] Step 3.5.1: Locate the knee ligaments and muscles.

[0025] Step 3.5.2, locate the contact structure of the knee ligament: extract the femoral cartilage and tibial plateau from the musculoskeletal model containing the knee ligaments and contact structure, scale them, and translate them to the precise contact position with the bone structure; export the geometric features and positioning information of the contact structure to OpenSim, adjust them in OpenSim, and integrate them into the musculoskeletal model containing the complete muscles of the trunk and lower limbs.

[0026] Step 4: Adjust the musculoskeletal model including the knee ligaments and trunk muscles obtained in step 3 to obtain an integrated whole-body musculoskeletal model:

[0027] Step 4.1, adjust the slack length of the ligaments: Generate the model's ligament strain at the randomly generated knee angle, which is the predicted value. Use the reference strain of the musculoskeletal model that includes the knee ligaments and contact structure as the true value, and use the loss function to calculate the error between the predicted value and the true value until the error approaches the minimum value. The loss function is shown in Equation II below:

[0028]

[0029] In formula II: minStrain represents the minimum error between the predicted value and the true value, Represents the error between the predicted value and the true value; represents the ligament strain of the model at the jth angle; represents the reference strain of the musculoskeletal model including the knee ligaments and contact structures at the jth angle; m represents the total number of randomly generated knee angles; Strain represents the ligament strain.

[0030] In step 4.2, a forward dynamics test is performed, and based on the forward dynamics results, the ligament stiffness parameters are adjusted.

[0031] In step 4.3, a forward dynamics test is performed, and the tibial plateau angle is adjusted based on the forward dynamics results.

[0032] Step 4.4, improve the knee joint degrees of freedom: Delete the original functional relationship between knee varus and valgus, internal rotation and external rotation, and flexion and extension degrees of freedom, and define them as completely independent. Modify the varus and valgus, internal rotation and external rotation as two new degrees of freedom; define the original anterior and posterior translation degrees of freedom of the knee joint as absolute freedom, and introduce anterior displacement and posterior displacement as a new degree of freedom.

[0033] The present invention also protects an OpenSim-based integrated whole-body musculoskeletal model obtained by the construction method described above, wherein the model includes complete knee joint ligaments and lower limb and trunk muscle groups.

[0034] The integrated whole-body musculoskeletal model includes four degrees of freedom of the knee joint, namely: knee flexion and extension, knee adduction and abduction, knee internal rotation and external rotation, and knee anterior and posterior displacement.

[0035] The knee joint ligaments include: the anterior cruciate ligament anteromedial bundle, the anterior cruciate ligament posterolateral bundle, the posterior cruciate ligament anterolateral bundle, the posterior cruciate ligament posteromedial bundle, the superficial bundle of the medial collateral ligament, the deep bundle of the medial collateral ligament, the central bundle of the superficial medial collateral ligament, the anterior bundle of the deep medial collateral ligament, the posterior bundle of the superficial medial collateral ligament, the lateral collateral ligament, the popliteal ligament, the oblique popliteal bundle of the posterior capsule, the arcuate popliteal bundle of the posterior capsule, the lateral bundle of the posterior capsule, the medial bundle of the posterior capsule, the lateral patellar tendon ligament, the medial patellar tendon ligament and the central patellar tendon ligament.

[0036] The present invention also protects the use of the above-mentioned integrated whole-body musculoskeletal model based on OpenSim for the prediction of anterior cruciate ligament load response.

[0037] Compared with the prior art, the present invention has the following technical effects:

[0038] (I) When adding the knee ligament structure, the present invention designs an algorithm and program for automatically searching for grid nodes at the ligament and muscle insertion points, so that the knee ligaments and muscles can automatically adjust and accurately locate and insert into the musculoskeletal model that includes the complete muscles of the trunk and lower limbs, ensuring that the muscles and knee ligaments can accurately fit the bone surface, thereby achieving precise integration of the knee ligaments and lower limb muscles.

[0039] (II) When adjusting the musculoskeletal model including the knee ligaments and trunk muscles, the present invention designs an algorithm and procedure for ligament relaxation length calibration, thereby minimizing the difference between the predicted value and the true value of ligament strain.

[0040] (III) The present invention improves the absolute degrees of freedom of the knee joint by redefining the original degrees of freedom of the knee joint and introducing new degrees of freedom.

[0041] (IV) The present invention successfully constructed for the first time an integrated whole-body musculoskeletal model with complete lower limbs, trunk muscle groups and complete ligament tissue of the knee joint. This model has the ability to quantify key physiological parameters and biomechanical characteristics of trunk muscle tissue, including trunk muscle activation patterns and force generation capacity. It can well simulate the physiological structure of the human body and achieve functional coordination of muscles, ligaments and bones, and can simultaneously predict the ligament strain and load in the knee joint, as well as the lower limb muscle force within the full physiological range of activity of the hip and knee joints.

[0042] (V) This invention provides a more comprehensive and effective research tool for studying the biomechanical mechanisms of ACL injury, promoting the advancement of related research. Furthermore, the constructed model can enhance the comprehensive assessment of ACL ligament injury risk factors, which has important clinical value for ACL injury prevention and post-injury rehabilitation. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 Schematic diagram of the Hill muscle-tendon model.

[0044] Figure 2 This is a schematic diagram of the structure of the lower limb musculoskeletal model; Figure 2 (A) is the front view, Figure 2 (B) is the back view.

[0045] Figure 3 This is a schematic diagram of the structure of the trunk musculoskeletal model; Figure 3 (A) is the front view, Figure 3 (B) is a side view. Figure 3 (C) is the back view.

[0046] Figure 4 shows Schematic diagram of the structure of the musculoskeletal model including the complete muscles of the trunk and lower limbs; Figure 4A) is the front view, Figure 4 B) is the back view.

[0047] Figure 5 Schematic diagram of the musculoskeletal model including knee ligaments and contact structures; Figure 5 A) is the front view, Figure 5 B) is the back view.

[0048] Figure 6 This is a schematic diagram of the structure of a full-body musculoskeletal model including knee ligaments and trunk muscles; Figure 6 (A) is the front view, Figure 6 (B) is the back view.

[0049] Figure 7 This is the multi-degree-of-freedom passive motion diagram of the tibiofemoral joint during passive knee flexion; Figure 7 (a) shows the change of knee flexion angle during passive knee flexion. Figure 7 (b) shows the changes in knee adduction / abduction angles during passive knee flexion. Figure 7 (c) is the change of internal rotation / external rotation angle of the knee joint during passive flexion of the knee joint. Figure 7 (d) is the change of the forward / backward displacement of the knee joint during passive flexion of the knee joint. Figure 7 (e) is the change of the upper / lower displacement of the knee joint during passive flexion of the knee joint. Figure 7 (f) Changes in medial / lateral displacement of the knee joint during passive knee flexion; Figure 7 In (a) to (f), the red curve represents the model data of the present invention, and the blue curve represents the real human in vitro measurement data.

[0050] Figure 8 is the anterior-posterior displacement diagram of the model when a 100N anterior-posterior shear force is applied; Figure 8 (a) is the forward displacement diagram of the model when a 100N forward shear force is applied. Figure 8 (b) is the backward displacement diagram of the model when a backward shear force of 100N is applied; Figure 8 In (a) and (b), the red diamond blocks represent the model data of the present invention, and the other legends represent the real human in vitro measurement data.

[0051] Figure 9 The varus and valgus angles of the model when a varus and valgus torque of 10 Nm is applied; Figure 9 (a) is the inversion angle diagram of the model when a 10Nm inversion torque is applied. Figure 9 (b) is the valgus angle diagram of the model when a valgus torque of 10 Nm is applied; Figure 9 In (a) and (b), the red diamond blocks represent the model data of the present invention, and the other legends represent the real human in vitro measurement data.

[0052] Figure 10 The internal and external rotation angles of the model when a 5 Nm rotational torque is applied; Figure 10 (a) is the internal rotation angle diagram of the model when a 5Nm rotation torque is applied. Figure 10 (b) is the external rotation angle diagram of the model when a 5 Nm rotational torque is applied; Figure 10 In (a) and (b), the red diamond blocks represent the model data of the present invention, and the other legends represent the real human in vitro measurement data.

[0053] Figure 11 It is a high-load exercise paradigm; among them, a shows static calibration, b shows walking posture, c shows side-cutting posture, d shows single-foot landing posture, e shows single-foot landing posture after vertical take-off, and f shows single-foot landing posture after lateral take-off.

[0054] Figure 12 The curves of knee joint angle (flexion) changing with time under different movement paradigms; Figure 12 (a) is the curve of knee joint angle (flexion) changing with time under episiotomy paradigm. Figure 12 (b) is the curve of knee joint angle (flexion) changing with time under the single-leg landing paradigm. Figure 12 (c) is the curve of knee joint angle (flexion) changing with time under the single-leg landing lateral jump paradigm. Figure 12 (d) is the curve of knee joint angle (flexion) changing with time under the single-leg landing vertical jump paradigm; Figure 12 (a)-(d) show the changes in knee flexion angle under different movement paradigms.

[0055] Figure 13 The curves of ligament length changes over time under different exercise paradigms are shown; Figure 13 (a) is the curve of ligament length changing with time under episiotomy paradigm. Figure 13 (b) is the curve of ligament length changing with time under the single-leg landing paradigm. Figure 13 (c) is the curve of ligament length changing with time under the condition of single-leg landing lateral jump paradigm. Figure 13 (d) is the curve of ligament length changing with time under the single-leg landing vertical jump paradigm; Figure 13 (a)-(d) show the changes in aACL (anteromedial anterior cruciate ligament) ligament length under different exercise paradigms.

[0056] Figure 14 The curves of ligament strain changing with time under different exercise paradigms; Figure 14 (a) is the curve of ligament strain changing with time under episiotomy paradigm conditions. Figure 14(b) is the curve of ligament strain changing with time under the single-leg landing paradigm. Figure 14 (c) is the curve of ligament strain changing with time under the single-leg landing lateral jump paradigm. Figure 14 (d) is the curve of ligament strain changing with time under the single-leg landing vertical jump paradigm; Figure 14 (a)-(d) show the changes in aACL (anteromedial anterior cruciate ligament) ligament strain under different exercise paradigms.

[0057] Figure 15 The curves of ligament stress changes with time under different exercise paradigms; Figure 15 (a) is the curve of ligament stress changing with time under episiotomy paradigm conditions. Figure 15 (b) is the curve of ligament stress changing with time under the single-leg landing paradigm. Figure 15 (c) is the curve of ligament stress variation with time under the condition of single-leg landing lateral jump paradigm. Figure 15 (d) is the curve of ligament stress variation with time under the single-leg landing vertical jump paradigm; Figure 15 (a)-(d) show the changes in aACL (anteromedial anterior cruciate ligament) ligament stress under different exercise paradigms.

[0058] Figure 16 The relationship between knee flexion and extension angle and the superior-inferior (a) and lateral-medial (b) displacement of the knee joint is shown.

[0059] The specific contents of the present invention are further explained in detail below with reference to the embodiments. DETAILED DESCRIPTION

[0060] It should be noted that all software in the present invention, unless otherwise specified, are software known in the art. For example: the modeling software uses OpenSim, nmsBuilder and Geomagic software known in the prior art. OpenSim is an open software developed based on C++ and JAVA languages ​​and used for muscle model development, simulation and analysis of neuromuscular system functions. nms Builder is a free software tool for creating personalized skeletal muscle models, especially suitable for the OpenSim environment. Its main function is to use skeletal data to establish skeletal entities, bone punctuation and muscle punctuation, thereby automatically generating corresponding reference systems and muscle models. Geomagic is a commonly used 3D software suite, widely used in manufacturing, medical and other fields

[0061] In accordance with the above technical solution, specific embodiments of the present invention are given below. It should be noted that the present invention is not limited to the following specific embodiments, and all equivalent changes made on the basis of the technical solution of this application fall within the protection scope of the present invention.

[0062] Example 1:

[0063] This embodiment provides an integrated whole-body musculoskeletal model based on OpenSim, such as Figure 1 As shown in the figure, the model includes complete knee ligaments and lower limb and trunk muscles. Specifically, it includes 148 Hill-type muscle-tendon units (the tendons are rigid tendons), 36 ligament bundles (18 on each side), 2 femoral cartilages (one on each side), and 2 tibial plateaus (one on each side).

[0064] As a specific solution of this embodiment, the ligament bundles are: anteromedial bundle of the anterior cruciate ligament, posterolateral bundle of the anterior cruciate ligament, anterolateral bundle of the posterior cruciate ligament, posteromedial bundle of the posterior cruciate ligament, superficial bundle of the medial collateral ligament, deep bundle of the medial collateral ligament, central bundle of the superficial medial collateral ligament, anterior bundle of the deep medial collateral ligament, posterior bundle of the superficial medial collateral ligament, lateral collateral ligament, popliteal ligament, oblique popliteal bundle of the posterior capsule, arcuate popliteal bundle of the posterior capsule, lateral bundle of the posterior capsule, medial bundle of the posterior capsule, lateral patellar tendon ligament, medial patellar tendon ligament, and central patellar tendon ligament; precise alignment of the tibial plateau and the cartilage model ensures minimal bone contact during maximum knee extension.

[0065] As a specific solution of this embodiment, the integrated whole-body musculoskeletal model includes multiple degrees of freedom of the lower limbs: three rotational degrees of freedom of the hip joint, specifically flexion and extension, adduction and abduction, and internal rotation and external rotation; four degrees of freedom of the knee joint, specifically flexion and extension, adduction and abduction, internal rotation and external rotation, and anterior displacement and posterior displacement; one degree of freedom of the ankle joint, specifically plantar flexion and dorsiflexion; and one degree of freedom of the subtalar joint, specifically adduction and abduction.

[0066] As a specific solution of this embodiment, Figure 2 As shown in FIG, the Hill muscle-tendon unit in the integrated whole-body musculoskeletal model includes a contraction unit and two nonlinear elastic units. The contraction unit is first connected in parallel with a nonlinear elastic unit and then in series with another nonlinear elastic unit.

[0067] Example 2:

[0068] This embodiment provides a method for constructing an integrated whole-body musculoskeletal model based on OpenSim, according to Example 1. The method includes: first, based on human anatomical and physiological parameters, obtaining a musculoskeletal model comprising complete muscles of the trunk and lower limbs; then, based on physiological characteristic parameters of the human knee joint ligaments and contact structures, adding the knee joint ligaments and contact structures to the knee joint of the musculoskeletal model comprising complete muscles of the trunk and lower limbs, thereby obtaining an integrated whole-body musculoskeletal model based on OpenSim. The method specifically comprises the following steps:

[0069] Step 1: Obtain a musculoskeletal model containing complete muscles of the trunk and lower limbs:

[0070] Step 1.1, based on the OpenSim human body model database, obtain a lower limb musculoskeletal model that conforms to human physiological characteristics; Figure 3 As shown in the figure, the lower limb musculoskeletal model consists of a skeletal system and a muscular system, wherein the skeletal system includes: skull, trunk, humerus, ulna, radius, hand bones, pelvis, femur, patella, tibia, fibula, talus, calcaneus (including tarsal bones and metatarsal bones) and toe bones; the muscular system includes 80 muscle-tendon units of the lower limb (40 muscle-tendon units on each side), specifically: adductor longus / brevis / major muscles, long head / short head of biceps femoris, extensor / flexor digitorum longus, extensor hallucis longus / brevis muscles, gluteus maximus / medius / minimus muscles, gracilis muscle, iliacus muscle, peroneus longus and brevis muscle, piriformis muscle, psoas major muscle, rectus femoris muscle, sartorius muscle, semimembranosus muscle, semitendinosus muscle, soleus muscle, tensor fasciae latae muscle, tibialis anterior / posterior muscle, vastus intermedius muscle, vastus lateralis muscle, vastus medialis muscle. The lower limb musculoskeletal model contains 20 degrees of freedom for the lower limb (6 degrees of freedom for the pelvis and 7 degrees of freedom for each leg), and the trunk and upper limb contain 17 degrees of freedom (3 degrees of freedom for the lumbar joints and 7 degrees of freedom for each arm).

[0071] Step 1.2: Based on the OpenSim human body model database, a trunk musculoskeletal model that conforms to the physiological characteristics of the human body is obtained; Figure 4 The trunk musculoskeletal model consists of a skeletal system and a muscular system. The skeletal system includes the skull, trunk, humerus, ulna, radius, hand bones, and pelvis. The trunk muscle group includes 142 muscle-tendon units (71 muscle-tendon units on each side), specifically the rectus abdominis, external oblique muscles, internal oblique muscles, erector spinae, multifidus muscles, and quadratus lumborum muscles.

[0072] Step 1.3, transfer the trunk muscle groups in the trunk musculoskeletal model obtained in step 1.2 to the lower limb musculoskeletal model obtained in step 1.1 to obtain a musculoskeletal model containing complete muscles of the trunk and lower limbs: first, identify and mark the same bony landmarks on the spine, thorax and pelvis of the lower limb musculoskeletal model and the trunk musculoskeletal model; then use affine transformation to project the landmarks of the trunk musculoskeletal model onto the landmarks of the lower limb musculoskeletal model, and then map the muscle path points of the trunk musculoskeletal model to the lower limb musculoskeletal model; finally, adjust the model to ensure the accurate positioning of the transferred muscle path points in the base model. Figure 5The results showed that the newly constructed musculoskeletal model containing complete muscles of the trunk and lower limbs consists of 148 Hill-type motion units with rigid tendons (29 on each side of the trunk and 45 on each lower limb). The model has the ability to move with multiple joint degrees of freedom of the spine and lower limbs (3 lumbar degrees of freedom, 3 hip degrees of freedom, 2 knee degrees of freedom, 1 ankle degree of freedom and 1 subtalar joint degree of freedom).

[0073] Step 2: Based on the OpenSim human body model database, a musculoskeletal model including knee joint ligaments and contact structures that conforms to human physiological characteristics is obtained; Figure 4 The musculoskeletal model including knee ligaments and contact structures is composed of a skeletal system, a muscular system and a knee ligament system; the skeletal system includes: pelvis, femur, patella, tibia and fibula; the muscular system includes three muscle-tendon units of the lower limbs, specifically: vastus medialis, vastus lateralis and vastus intermedius; the ligament system includes 18 bundles of knee ligaments, specifically: anterior cruciate ligament anteromedial bundle, anterior cruciate ligament posterolateral bundle, posterior cruciate ligament anterolateral bundle, Posteromedial bundle of the posterior cruciate ligament, superficial bundle of the medial collateral ligament, deep bundle of the medial collateral ligament, central bundle of the superficial medial collateral ligament, anterior bundle of the deep medial collateral ligament, posterior bundle of the superficial medial collateral ligament, lateral collateral ligament, popliteal ligament, oblique popliteal bundle of the posterior capsule, arcuate popliteal bundle of the posterior capsule, lateral bundle of the posterior capsule, medial bundle of the posterior capsule, lateral patellar tendon ligament, medial patellar tendon ligament, central patellar tendon ligament; the musculoskeletal model including the knee joint ligaments and contact structures has 6 degrees of freedom for the tibial joint and 1 degree of freedom for the patellofemoral joint.

[0074] Step 3: Add the knee ligaments and contact structures in the musculoskeletal model including the knee ligaments and contact structures obtained in Step 2 to the musculoskeletal model including the complete muscles of the trunk and lower limbs obtained in Step 1 to obtain a preliminary integrated model. After adjustment, a musculoskeletal model including the knee ligaments and trunk muscles is obtained:

[0075] In step 3.1, nmsBuilder was used to synchronously scale the tibia and fibula in the musculoskeletal model containing the knee ligaments and contact structures to ensure that the bone structures and their associated ligament attachment points were precisely aligned with the reference dimensions of the musculoskeletal model containing the complete muscles of the trunk and lower limbs.

[0076] In step 3.2, the tibia and fibula in the musculoskeletal model including the knee joint ligaments and contact structures are used to replace the corresponding anatomical structures in the musculoskeletal model including the complete muscles of the trunk and lower limbs.

[0077] In step 3.3, find the same set of bone surface landmarks on the tibia and fibula in the musculoskeletal model including the complete muscles of the trunk and lower limbs and the musculoskeletal model including the knee joint ligaments and contact structures.

[0078] In step 3.4, affine transformation is used to project the bone landmark points in the musculoskeletal model containing the knee joint ligaments and contact structure to the corresponding bone landmark points in the musculoskeletal model containing the complete muscles of the trunk and lower limbs. The ligament and muscle path points in the musculoskeletal model containing the knee joint ligaments and contact structure are then mapped to the basic musculoskeletal model to obtain a preliminary integrated model.

[0079] Step 3.5, Integrate knee ligaments and contact structures:

[0080] Step 3.5.1, Positioning the Knee Ligaments and Muscles: Mesh the tibia and fibula. Because the bone structures of the musculoskeletal model containing the knee ligaments and contact structures differ from those of the musculoskeletal model containing the complete muscles of the trunk and lower limbs in shape, position, and orientation, the bone structures of the preliminary integrated model are derived from these two models. The muscle and ligament insertion points obtained after mapping may not fit the bone surface, requiring muscle and ligament positioning as follows:

[0081] First, calculate the closest distance between the muscle and ligament insertion points and the adjacent bone surface. The distance calculation formula is shown in the following formula I:

[0082]

[0083] In formula I: d i represents the Euclidean distance between a muscle or ligament and the mesh nodes on which it is inserted into n (i.e., all) bone surfaces; p i =(x i ,y i ,z i ), represents the coordinates of the mesh node on the i-th bone surface; l=(x l ,y l ,z l ), representing the coordinates of a muscle or ligament insertion point.

[0084] The minimum Euclidean distance among all these distances is then found. The point on the bone surface corresponding to this distance is the new insertion point of the muscle or ligament. After the program is set up according to the above algorithm, the origin and insertion points of the transferred ligament and its own muscle are automatically adjusted to ensure the accurate positioning of the ligament and muscle path points in the basic musculoskeletal model. In this embodiment, step 3.5 ensures that the muscle and ligament insertion points are aligned with the bone surface. This step simply fine-tunes the insertion point position based on the previous affine transformation step to ensure that the structure conforms to actual physiological conditions.

[0085] Step 3.5.2, locate the contact structure of the knee ligaments: Extract the femoral cartilage and tibial plateau from the musculoskeletal model containing the knee ligaments and contact structures, scale them separately in Geomagic software, and translate them to the precise contact position with the bone structure; export the geometric features and positioning information of the contact structure to OpenSim, make appropriate adjustments in OpenSim, and integrate it into the musculoskeletal model containing the complete muscles of the trunk and lower limbs to obtain a musculoskeletal model that includes the knee ligaments and trunk muscles.

[0086] Step 4: Adjust the musculoskeletal model including the knee ligaments and trunk muscles obtained in step 3 to obtain an integrated whole-body musculoskeletal model:

[0087] Step 4.1, adjust the slack length of the ligaments: Due to changes in ligament geometry, the slack length of the ligaments needs to be further adjusted to improve the accuracy of ligament strain prediction. Therefore, based on the ligament length data under multiple knee angles, a ligament slack length calibration algorithm is designed to minimize the difference between the ligament strain of the integrated whole-body musculoskeletal model and the predicted value of the musculoskeletal model containing the knee ligaments and contact structures under the same posture. The ligament slack length calibration algorithm is specifically as follows: at randomly generated knee angles, generate the ligament strain of the preliminary integrated model, which is the predicted value; use the reference strain of the musculoskeletal model containing the knee ligaments and contact structures as the true value, and use the loss function to calculate the error between the predicted value and the true value until the error approaches the minimum value; based on the above algorithm, set up the program and calibrate the ligament slack length to minimize the difference between the ligament strain and the predicted value of the local knee model under the same posture; the loss function is shown in the following formula III:

[0088]

[0089] In formula II: min Strain represents the minimum error between the predicted value and the true value, Represents the error between the predicted value and the true value; represents the ligament strain of the preliminary integrated model at the jth angle (determined by the angle parameter θ); represents the ligament strain of the musculoskeletal model including the knee ligaments and contact structures at the jth angle (determined by the angle parameter θ); m represents the total number of randomly generated knee angles; Strain represents the ligament strain, which is defined as (ligament length - ligament relaxation length) / ligament relaxation length, and is calculated as shown in Equation III below:

[0090]

[0091] In formula III, L represents the instantaneous length of the ligament bundle, in mm; L0 represents the relaxed length of the ligament bundle, in mm. L0 is personalized by scaling based on the anthropometric information of different subjects.

[0092] Step 4.2: Perform forward dynamics testing under conditions of passive knee flexion and extension, 100 N anteroposterior shear force, 10 Nm varus and valgus torque, and 5 Nm internal and external rotation torque. Based on the forward dynamics results, adjust the ligament stiffness parameters so that the joint angles and displacements generated by the forward dynamics are close to the acquired musculoskeletal model including the knee ligaments and contact structures and fall within a reasonable physiological range.

[0093] Step 4.3: Perform forward dynamics testing under conditions of passive knee flexion and extension, 100 N anteroposterior shear force, 10 Nm internal and external valgus torque, and 5 Nm internal and external rotation torque. Based on the forward dynamics results, adjust the tibial plateau angle so that the joint angle and displacement changes generated by the forward dynamics are close to the acquired musculoskeletal model including the knee ligaments and contact structures and fall within a reasonable physiological range.

[0094] In step 4.4, the absolute degrees of freedom of the knee joint in the musculoskeletal model are increased to accurately capture the biomechanical changes of the ACL during specific movement patterns such as high-intensity exercise. The definitions of knee valgus, internal and external rotation angles, and anterior-posterior translation are modified to introduce knee valgus / valgus and internal / external rotation as new degrees of freedom. Changes in knee valgus / valgus and internal / external rotation angles no longer depend on the knee flexion angle and can be adjusted according to the subject's knee anatomy. In the newly created model, anterior-posterior displacement of the knee joint is introduced as a new degree of freedom. In addition, based on the forward dynamics results, the relationship between the knee flexion and extension angles and the upper and lower displacements and the internal and external displacements of the knee joint are optimized.

[0095] After the model adjustment is completed, an integrated whole-body musculoskeletal model based on OpenSim is obtained. The specific structure is shown in Example 1.

[0096] In this embodiment, the forward dynamic test is used to evaluate the anteroposterior, varus, and internal and external rotation laxity between 0 and 90 degrees, with each 15-degree increase in knee flexion angle. The specific results are as follows:

[0097] (A) Testing the laxity of the knee joint under passive flexion and extension conditions: The femur is fixed and muscle activation is set to zero. The model is then passively flexed to 90°, with the remaining five degrees of freedom unconstrained. This process is controlled by ligament and contact loads. In OpenSim, this test is completed by defining the knee flexion angle as a time-varying function specified in the model file and performing a forward dynamic simulation. The results are shown in Figure 2. Figure 8As shown, after comparing the results with those disclosed in the prior art, it is confirmed that each degree of freedom under passive flexion and extension conditions is within the previous reasonable range;

[0098] (B) Testing the anteroposterior laxity of the tibia: A 100N anteroposterior shear force was applied to the tibia 10cm below the origin of the tibial coordinate system, and forward dynamics was used to simulate the movement of the tibia and patella to test the anteroposterior laxity of the tibia. Figure 9 As shown, after comparing the results with the results disclosed in the prior art, it is confirmed that the anterior-posterior displacement of the tibiofemoral joint under the conditions of anterior-posterior shear force is within a reasonable range. The test results disclosed in the prior art for comparison are shown in Table 2 below:

[0099] Table 2. Anterior-posterior translation distance of the tibial joint at various knee flexion angles under the condition of 100N anterior-posterior shear force

[0100]

[0101] (C) Testing the internal and external rotation laxity of the tibia: The internal and external rotation laxity was tested by applying an axial rotational torque of 5 Nm to the long axis of the tibia and using dynamic simulation to predict the resulting tibiofemoral rotation. The test results are shown in Figure 2. Figure 10 As shown, after comparing the results with those disclosed in the prior art, it is confirmed that under the condition of axial rotation torque, the internal and external rotation angles of the tibiofemoral joint are all within a reasonable range. The test results disclosed in the prior art for comparison are shown in Table 3 below:

[0102] Table 3. Tibial varus and valgus angles at various knee flexion angles under the condition of 10 Nm varus and valgus torque

[0103]

[0104]

[0105] (D) Testing varus and varus laxity of the tibia: Varus and varus laxity was assessed by applying a 10 Nm varus torque to the tibia and simulating the resulting tibiofemoral rotation using forward dynamics. Figure 11 As shown, after comparing the results with the results disclosed in the prior art, it is confirmed that the valgus and valgus angles of the tibiofemoral joint are within a reasonable range under the conditions of valgus and valgus torque. The test results disclosed in the prior art for comparison are shown in Table 4 below:

[0106] Table 4. Internal and external rotation angles of the tibial joint at various knee flexion angles under the condition of applying 5Nm internal and external rotation torque

[0107]

[0108]

[0109] Example 3:

[0110] This embodiment describes the application of the OpenSim-based integrated whole-body musculoskeletal model of Example 1 to the prediction of ACL ligament load response. The application is as follows:

[0111] First, the model preserves the anatomical and physiological parameters of the original model's ligaments and muscles (origin and insertion paths, tendon slack length, and optimal muscle fiber length). It also preserves the original model's lower limb ligament dynamics and muscle inclusion morphology. These features enable reliable quantification of geometric parameters and biomechanical variables, such as muscle forces, through advanced muscle-tendon force models or ligament loading patterns using advanced computational frameworks.

[0112] As a specific solution of this embodiment, the ligament is regarded as a non-dissipative nonlinear elastic spring, and its force-strain relationship is defined by a quadratic interval and a linear interval exceeding 3% strain, which is expressed by the following formula IV:

[0113]

[0114] Where: f represents the force of the ligament, in N; ε represents the ligament strain, which is a numerical ratio; k represents the reference linear stiffness, which is defined as the slope of the f-ε curve; ε l represents the linear strain limit.

[0115] The ligament strain is defined as (ligament length - ligament relaxation length) / ligament relaxation length, and is expressed as follows:

[0116]

[0117] Where: L represents the instantaneous length of the ligament bundle, in mm; L0 represents the relaxed length of the ligament bundle, in mm. L0 is personalized by scaling based on the anthropometric information of different subjects.

[0118] As a specific solution of this embodiment, the properties of the ligament (including its relaxation length and stiffness) are shown in the following table: Table 5. Ligament properties

[0119] Second, an optical motion capture system was used to collect human kinematic data under various high-load exercise paradigms (e.g. Figure 12 shown).

[0120] Third, demographic data such as height and weight of the subjects were collected and personalized scaling was performed on the newly created model in OpenSim.

[0121] Third, the collected kinematic data were input into the newly created model and inverse kinematics analysis was performed in OpenSim. By using experimental marker data from different task trials, the error between paired experimental markers and model markers was minimized to obtain the curve of joint angle change over time (e.g. Figure 13 shown).

[0122] Fourth, based on the inverse kinematics results, ligament length analysis was performed to calculate the ACL length under various motion paradigms (e.g. Figure 14 shown).

[0123] Fifth, the ACL strain size was calculated based on the absolute length of the ACL under various motion paradigms and the ACL relaxation length (e.g. Figure 15 shown).

[0124] Sixth, based on the strain values ​​under various motion paradigms and combined with the ACL stiffness values, the ACL stress magnitude (e.g. Figure 16 shown).

Claims

1. A method for constructing an integrated whole-body musculoskeletal model based on OpenSim, characterized in that: The method comprises: integrating knee joint ligaments and contact structures into a musculoskeletal model including complete muscles of the trunk and lower limbs to obtain a musculoskeletal model including the knee joint ligaments and contact structures; then adjusting the model to obtain an integrated whole-body musculoskeletal model; Integrating the knee ligaments and contact structures includes locating the knee ligaments and muscles. Specifically, the shortest distance from the insertion point of the knee ligaments and muscles to the adjacent bone surface is calculated using Formula I. Then, the minimum Euclidean distance is found among all distances. The point on the bone surface corresponding to this distance is the new insertion point of the knee ligament or muscle. In formula I: d i represents the Euclidean distance between a knee ligament or muscle and the mesh nodes on the surface of n bones where it is inserted; p i =(x i ,y i ,z i ), represents the coordinates of the mesh node on the i-th bone surface; l=(x l ,y l ,z l ), representing the coordinates of a muscle or ligament insertion point.

2. The method for constructing an integrated whole-body musculoskeletal model based on OpenSim according to claim 1, characterized in that: Model adjustment involves adjusting the slack length of the ligaments. Specifically, the ligament strain of the model is generated at randomly generated knee angles, which is the predicted value. The reference strain of the musculoskeletal model that includes the knee ligaments and contact structures is used as the true value, and the error between the predicted value and the true value is calculated using a loss function until the error approaches the minimum value. The loss function is shown in Equation II below: In formula II: min Strain represents the minimum error between the predicted value and the true value, Represents the error between the predicted value and the true value; represents the ligament strain of the model at the jth angle; represents the reference strain of the musculoskeletal model including the knee ligaments and contact structures at the jth angle; m represents the total number of randomly generated knee angles; Strain represents the ligament strain.

3. The method for constructing an integrated whole-body musculoskeletal model based on OpenSim as claimed in claim 1, characterized in that: Model adjustments include improving the knee joint degrees of freedom. Specifically, the original functional relationship between knee varus and valgus, internal rotation and external rotation and flexion and extension degrees of freedom in the model is deleted and defined as completely independent. The modified varus and valgus, internal rotation and external rotation are used as two new degrees of freedom; the original anterior and posterior translation degrees of freedom of the knee in the model are defined as absolute freedom, and anterior displacement and posterior displacement of the knee are introduced as a new degree of freedom.

4. The method for constructing an integrated whole-body musculoskeletal model based on OpenSim according to claim 1, wherein: The method specifically includes the following steps: Step 1: Obtain a musculoskeletal model containing complete muscles of the trunk and lower limbs; Step 2: Obtain a musculoskeletal model that conforms to human physiological characteristics and includes knee joint ligaments and contact structures; Step 3: Integrate the knee joint ligaments and contact structures of the musculoskeletal model including the knee joint ligaments and contact structures obtained in step 2 into the knee joint of the musculoskeletal model including the complete muscles of the trunk and lower limbs obtained in step 1; Step 4: Adjust the musculoskeletal model including the knee ligaments and trunk muscles obtained in step 3 to obtain an integrated whole-body musculoskeletal model.

5. The method for constructing an integrated whole-body musculoskeletal model based on OpenSim according to claim 4, characterized in that: The step three is specifically as follows: Step 3.1: synchronously scale the tibia and fibula in the musculoskeletal model containing the knee ligaments and contact structures to ensure that the bone structures and their associated ligament attachment points are precisely aligned with the reference dimensions of the musculoskeletal model containing the complete muscles of the trunk and lower limbs; Step 3.2, using the tibia and fibula in the musculoskeletal model including the knee joint ligaments and contact structures to replace the corresponding anatomical structures in the musculoskeletal model including the complete muscles of the trunk and lower limbs; Step 3.3, find the same set of bone surface landmarks on the tibia and fibula in the musculoskeletal model including the complete muscles of the trunk and lower limbs and the musculoskeletal model including the knee joint ligaments and contact structures; Step 3.4, using affine transformation to project the bone landmarks in the musculoskeletal model containing the knee ligaments and contact structures onto the corresponding bone landmarks in the musculoskeletal model containing the complete muscles of the trunk and lower limbs, and then mapping the knee ligament and muscle path points in the musculoskeletal model containing the knee ligaments and contact structures onto the basic musculoskeletal model; Step 3.5, integrate knee ligaments and muscles; Step 3.5.1, locate the knee ligaments and muscles; Step 3.5.2, locate the contact structure of the knee ligament: extract the femoral cartilage and tibial plateau from the musculoskeletal model containing the knee ligaments and contact structure, scale them, and translate them to the precise contact position with the bone structure; export the geometric features and positioning information of the contact structure to OpenSim, adjust them in OpenSim, and integrate them into the musculoskeletal model containing the complete muscles of the trunk and lower limbs.

6. The method for constructing an integrated whole-body musculoskeletal model based on OpenSim according to claim 4, wherein: The step 4 is specifically as follows: Step 4.1, adjust the slack length of the ligament; Step 4.2: Perform a forward dynamics test and adjust the ligament stiffness parameters based on the forward dynamics results. Step 4.3: Perform a forward dynamics test and adjust the tibial plateau angle based on the forward dynamics results. Step 4.4: Improve knee joint freedom.

7. An integrated whole-body musculoskeletal model based on OpenSim obtained by the construction method according to any one of claims 1 to 6; characterized in that: The model includes complete knee ligaments and trunk muscles.

8. The integrated whole-body musculoskeletal model based on OpenSim according to claim 7, wherein: The integrated whole-body musculoskeletal model includes four degrees of freedom of the knee joint, namely: knee flexion and extension, knee adduction and abduction, knee internal rotation and external rotation, and knee anterior and posterior displacement.

9. The integrated whole-body musculoskeletal model based on OpenSim according to claim 7, wherein: The knee joint ligaments include: the anterior cruciate ligament anteromedial bundle, the anterior cruciate ligament posterolateral bundle, the posterior cruciate ligament anterolateral bundle, the posterior cruciate ligament posteromedial bundle, the superficial bundle of the medial collateral ligament, the deep bundle of the medial collateral ligament, the central bundle of the superficial medial collateral ligament, the anterior bundle of the deep medial collateral ligament, the posterior bundle of the superficial medial collateral ligament, the lateral collateral ligament, the popliteal ligament, the oblique popliteal bundle of the posterior capsule, the arcuate popliteal bundle of the posterior capsule, the lateral bundle of the posterior capsule, the medial bundle of the posterior capsule, the lateral patellar tendon ligament, the medial patellar tendon ligament and the central patellar tendon ligament.

10. Application of the integrated whole-body musculoskeletal model based on OpenSim as claimed in any one of claims 7 to 9 in the prediction of anterior cruciate ligament load response.

Images