Ankle joint stiffness estimation method and system based on musculoskeletal anatomical statistical model

Abstract

The present invention provides a method and system for estimating ankle joint stiffness based on a skeletal muscle anatomical statistical model, belonging to the field of human mechanical impedance measurement. The method comprises: inputting preprocessed surface electromyographic signals into a muscle activation dynamics model to calculate muscle activation signals; inputting human limb characteristic parameters into a musculoskeletal statistical model to obtain muscle structure micro-parameters; calculating muscle force and muscle-tendon unit force at a specific muscle activation level using a muscle force solution model based on the muscle activation signals and muscle structure micro-parameters; calculating the stiffness of the muscle-tendon unit using a muscle-tendon unit stiffness model based on the muscle force and muscle structure micro-parameters; and calculating the joint stiffness using a muscle-tendon to joint mapping model based on the muscle-tendon unit force, muscle structure micro-parameters, and muscle-tendon unit stiffness. Compared with traditional stiffness identification and stiffness estimation methods, the present invention is more convenient, faster, and has higher accuracy.

Description

Technical Field

[0001] The present invention belongs to the field of mechanical impedance measurement, and more specifically, relates to a method and system for estimating ankle joint stiffness based on a skeletal muscle anatomical statistical model. Background Art

[0002] Existing research on human stiffness generally uses two methods to obtain ankle joint stiffness: stiffness identification and stiffness estimation. While ankle joint stiffness identification methods can characterize human ankle joint stiffness in multiple postures, they typically require the development of dedicated identification devices, which in turn necessitates a certain hardware foundation for stiffness acquisition. Furthermore, due to the limitations of the identification devices, the subject's posture is typically relatively limited, making it difficult to extend to other motion scenarios. Therefore, stiffness identification methods have significant limitations in practical engineering applications.

[0003] With the recent advancement of computer technology, data-driven models with predictive capabilities, built using black-box models and neural networks, have become increasingly common in engineering applications. While data-driven models offer advantages such as simple construction and excellent estimation accuracy, their use in joint stiffness estimation is limited due to the difficulty in measuring joint stiffness, the limited use of models in specific scenarios, and the unclear effects of parameters. An analysis of published stiffness estimation research reveals that the current mainstream approach is to predict human joint stiffness by constructing musculoskeletal models based on stiffness generation mechanisms. However, while these musculoskeletal models align with neural transmission principles and human anatomy, and offer advantages in biological interpretability and generalizability, they suffer from low estimation accuracy and increased modeling difficulty due to the large number of submodels and the amplification of intermediate variables that can lead to estimation errors. Summary of the Invention

[0004] In response to the defects of the prior art, the purpose of the present invention is to provide an ankle joint stiffness estimation method and system based on a musculoskeletal anatomical statistical model, aiming to solve the problem that the existing model is difficult to model and the increase of intermediate variables will cause the estimation error to be amplified and the estimation accuracy of the model is low.

[0005] To achieve the above objectives, the present invention provides, on the one hand, a method for estimating ankle joint stiffness based on a musculoskeletal anatomical statistical model, comprising the following steps:

[0006] S1: Measure human limb characteristic parameters and surface electromyography signals; the human limb characteristic parameters include ankle joint angle, knee joint angle, height, and weight; the surface electromyography signals include the surface electromyography signals of the tibialis anterior, medial gastrocnemius, lateral gastrocnemius, and soleus muscles that drive the corresponding joints;

[0007] S2: After high-pass filtering to remove motion artifacts, the surface electromyographic signals from each muscle are subjected to full-wave rectification and normalization to complete the electromyographic signal preprocessing process; the preprocessed surface electromyographic signals are input into the muscle activation dynamics model to calculate the activation signals of each muscle;

[0008] S3: Inputting the characteristic parameters of human limbs into the musculoskeletal statistical model, and obtaining the microscopic parameters of muscle structure through the musculoskeletal statistical model;

[0009] S4: Based on the muscle activation signals and the microscopic parameters of each muscle structure, the muscle force and muscle-tendon unit force of each muscle at a specific muscle activation level are calculated using the muscle force solution model;

[0010] S5: Based on the microscopic parameters of muscle force and muscle structure, the stiffness of each muscle-tendon unit is calculated using the muscle-tendon unit stiffness model;

[0011] S6: Based on the muscle force, microscopic parameters of muscle structure and the stiffness of muscle-tendon unit, the joint stiffness of the ankle joint driven by the tibialis anterior, medial gastrocnemius, lateral gastrocnemius and soleus muscles is calculated through the muscle-tendon-joint mapping model;

[0012] Among them, the musculoskeletal anatomical statistical model includes: muscle force solution model, musculoskeletal statistical model, muscle activation dynamics model, muscle tendon unit stiffness model and muscle tendon-joint mapping model.

[0013] Further preferably, the musculoskeletal statistical model is:

[0014] L shank =HT(0.247+Π F (sex)×0.01)

[0015] L thigh =HT(0.232+Π F (sex)×0.017)

[0016]

[0017]

[0018]

[0019]

[0020] (α o ) sol =28.3π / 180,(α o ) mg =9.9π / 180,(α o )lg =12π / 180,(α o ) ta =9.6π / 180

[0021]

[0022]

[0023]

[0024]

[0025]

[0026]

[0027]

[0028]

[0029]

[0030]

[0031]

[0032] in,

[0033]

[0034]

[0035]

[0036]

[0037] in,

[0038]

[0039]

[0040]

[0041] Wherein, HT is the height of the subject; F (sex) is the subject's gender parameter. When the subject is male, Π F (sex) is 0, when the subject is female, Π F (sex) is 1; L thigh is the length of the subject's thigh; L shank is the length of the subject's calf; Represents the optimal muscle fiber length of the soleus, medial gastrocnemius, lateral gastrocnemius and tibialis anterior muscles respectively (α o ) sol 、(α o ) mg 、(α o ) lg and (α o ) ta are the optimal pennation angles of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively; α SOL , α MG , α LG and α TA represent the pennation angles of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively; The resting lengths of the muscle-tendon unit of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively SOL L MT 、 MG L MT 、 LG L MT 、 TA L MT The muscle-tendon unit lengths L of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are MT , is a function related to the initial ankle angle θ0 and the angle change Δθ; R SOL 、R MG 、R LG 、R TA The muscle moment arms R are the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively. The moment arm slopes of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are K T is tendon stiffness; is the normalized tendon stiffness; f l t 、f l s 、f l p is the proportional coefficient, W is the width of the muscle; a is the muscle activation signal; x1, x2, x3, y1 are mathematical parameters; muscle microscopic parameters are calculated based on human limb parameters, including the resting length of the muscle-tendon unit Optimal muscle fiber length The optimal pennation angle α0 of the muscle, the pennation angle α of the muscle, the muscle moment arm R and the moment arm slope Among them, human limb parameters include: height HT, weight M, ankle joint angle

[0042] Further preferably, the muscle activation dynamics model is:

[0043] u j (t) = f a (e j (t))

[0044]

[0045]

[0046] Among them, f a is the neural activation model; u j (t) is the neural activation signal; e j (t) is the surface electromyography signal; a j (t) is the activation signal reflecting the activation intensity of the j-th muscle, is the activation rate of the jth muscle; c1+c2 is the activation rate constant, corresponding to u j (t) = 1; c2 is the inactivation rate constant, corresponding to u j (t) = 0; t act is the muscle activation time constant; t deact is the inactivation time constant; ζ is the muscle activation-inactivation ratio coefficient; according to the surface electromyography signal e j (t) Calculate muscle activation signal a j (t);

[0047] The muscle force solution model is:

[0048]

[0049] SOL V=2.57HT*M+120

[0050] MG V=1.71HT*M+46.2

[0051] LG V=1.08HT*M+15.7

[0052] TA V=0.796HT*M+36.7

[0053]

[0054]

[0055]

[0056]

[0057]

[0058]

[0059]

[0060]

[0061] Among them, F M is the muscle force, f l s is the proportional coefficient, W is the width of the muscle, α0 is the optimal feathering angle of the muscle, α is the feathering angle of the muscle, For the optimal muscle fiber length, For optimal muscle strength; The optimal muscle strength of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are SOL PCSA, MG PCSA, LG PCSA, TA PCSA is the muscle cross-sectional area PCSA of soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles; SOL V. MG V. LG V. TA V is the muscle volume of soleus, medial gastrocnemius, lateral gastrocnemius and tibialis anterior respectively; HT is the height of the subject, M is the weight of the subject; according to the optimal muscle fiber length The muscle force F is calculated based on the optimal muscle pennation angle α0, muscle pennation angle α and muscle activation signal a. M .

[0062] Further preferably, the muscle-tendon unit stiffness model is:

[0063]

[0064] in,

[0065]

[0066]

[0067]

[0068]

[0069]

[0070]

[0071]

[0072] (α o ) sol =28.3π / 180,(α o ) mg =9.9π / 180,(α o ) lg =12π / 180,(α o ) ta =9.6π / 180

[0073]

[0074] Among them, f p is the proportionality coefficient, K p is the passive stiffness of the muscle-tendon unit, K MTU is the muscle-tendon unit stiffness, K M is the muscle stiffness, K T is the tendon stiffness, and Tendon stiffness of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively; and The resting lengths of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are is the optimal muscle fiber length; (α o ) sol 、(α o ) mg 、(α o ) lg and (α o ) ta The optimal pennation angles of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are shown in Figure 2. x4 is a mathematical parameter. The resting length of the muscle-tendon unit is given by Optimal muscle fiber length and muscle force F M Calculate the muscle-tendon unit stiffness K MTU .

[0075] Further preferably, the muscle-tendon-joint mapping model is:

[0076]

[0077]

[0078]

[0079]

[0080] Among them, when multiple muscles act on the same joint and drive the same joint movement, the subscript j added to the mathematical variable represents the jth muscle; is the muscle stiffness of the j-th muscle, is the stiffness of the j-th muscle-tendon unit, is the tendon stiffness of the jth muscle, ΔL j is the deformation of the jth muscle, θ is the ankle joint angle; K j is the stiffness of the jth muscle; K is the joint stiffness; R j is the moment arm of the jth muscle; is the slope of the moment arm of the jth muscle; is the muscle force of the jth muscle; according to the muscle force F M , muscle-tendon unit stiffness K MTU , muscle moment arm R and moment arm slope Calculate the muscle stiffness K of each muscle j , and then calculate the joint stiffness K.

[0081] In another aspect, the present invention provides an ankle joint stiffness estimation system based on a musculoskeletal anatomical statistical model, comprising:

[0082] A parameter measurement module is used to measure human limb characteristic parameters and surface electromyography signals; wherein the human limb characteristic parameters include: ankle joint angle, knee joint angle, height and weight; the surface electromyography signals include: surface electromyography signals of the tibialis anterior muscle, medial gastrocnemius muscle, lateral gastrocnemius muscle and soleus muscle that drive the corresponding joints;

[0083] The parameter preprocessing module is used to perform high-pass filtering on the surface electromyographic signals from each muscle to remove motion artifacts, and then perform full-wave rectification and normalization to complete the preprocessing of the electromyographic signals;

[0084] A muscle activation signal calculation module is used to input the pre-processed surface electromyography signal into the muscle activation dynamics model to calculate each muscle activation signal;

[0085] A micro-parameter acquisition module is used to input the characteristic parameters of human limbs into the musculoskeletal statistical model and obtain the micro-parameters of muscle structure through the musculoskeletal statistical model;

[0086] The muscle force calculation module is used to calculate the muscle force and muscle-tendon unit force of each muscle at a specific muscle activation level through a muscle force solution model based on the muscle activation signals and the microscopic parameters of each muscle structure;

[0087] The muscle-tendon unit stiffness calculation module is used to calculate the stiffness of each muscle-tendon unit based on the microscopic parameters of each muscle force and muscle structure through the muscle-tendon unit stiffness model;

[0088] The joint stiffness evaluation module is used to calculate the joint stiffness of the ankle joint driven by the tibialis anterior, medial gastrocnemius, lateral gastrocnemius and soleus muscles based on the muscle-tendon unit force, microscopic parameters of the muscle structure and the stiffness of the muscle-tendon unit through the muscle-tendon-joint mapping model.

[0089] Further preferably, the musculoskeletal statistical model is:

[0090] L shank =HT(0.247+Π F (sex)×0.01)

[0091] L thigh =HT(0.232+Π F (sex)×0.017)

[0092]

[0093]

[0094]

[0095]

[0096] (α o ) sol =28.3π / 180,(α o ) mg =9.9π / 180,(α o ) lg =12π / 180,(α o ) ta =9.6π / 180

[0097]

[0098]

[0099]

[0100]

[0101]

[0102]

[0103]

[0104]

[0105]

[0106]

[0107]

[0108] in,

[0109]

[0110]

[0111]

[0112]

[0113] in,

[0114]

[0115]

[0116]

[0117] Wherein, HT is the height of the subject; F (sex) is the subject's gender parameter. When the subject is male, Π F (sex) is 0, when the subject is female, Π F (sex) is 1; L thigh is the length of the subject's thigh; L shank is the length of the subject's calf; Represents the optimal muscle fiber length of the soleus, medial gastrocnemius, lateral gastrocnemius and tibialis anterior muscles respectively (α o ) sol 、(α o ) mg 、(α o ) lg and (α o ) ta are the optimal pennation angles of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively; α SOL , α MG , α LG and α TA represent the pennation angles of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively; The resting lengths of the muscle-tendon unit of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively SOL L MT 、 MG L MT 、 LG L MT 、 TA L MT The muscle-tendon unit lengths L of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are MT , is a function related to the initial ankle angle θ0 and the angle change Δθ; R SOL 、R MG 、R LG 、R TA The muscle moment arms R are the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively. The moment arm slopes of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are K T is tendon stiffness; is the normalized tendon stiffness; f l t 、f l s 、f l p is the proportional coefficient, W is the width of the muscle; a is the muscle activation signal; x1, x2, x3, y1 are mathematical parameters; muscle microscopic parameters are calculated based on human limb parameters, including the resting length of the muscle-tendon unit Optimal muscle fiber length The optimal pennation angle α0 of the muscle, the pennation angle α of the muscle, the muscle moment arm R and the moment arm slope Among them, human limb parameters include: height HT, weight M, ankle joint angle

[0118] Further preferably, the muscle activation dynamics model is:

[0119] u j (t) = f a (e j (t))

[0120]

[0121]

[0122] Among them, f a is the neural activation model; u j (t) is the neural activation signal; e j (t) is the surface electromyography signal; a j (t) is the activation signal reflecting the activation intensity of the j-th muscle, is the activation rate of the jth muscle; c1+c2 is the activation rate constant, corresponding to u j (t) = 1; c2 is the inactivation rate constant, corresponding to u j (t) = 0; t act is the muscle activation time constant; t deact is the inactivation time constant; ζ is the muscle activation-inactivation ratio coefficient; according to the surface electromyography signal e j (t) Calculate muscle activation signal a j (t);

[0123] The muscle force solution module is:

[0124]

[0125] SOL V=2.57HT*M+120

[0126] MG V=1.71HT*M+46.2

[0127] LG V=1.08HT*M+15.7

[0128] TA V=0.796HT*M+36.7

[0129]

[0130]

[0131]

[0132]

[0133]

[0134]

[0135]

[0136]

[0137] Among them, F M is the muscle force, f l s is the proportional coefficient, W is the width of the muscle, α0 is the optimal feathering angle of the muscle, α is the feathering angle of the muscle, For the optimal muscle fiber length, For optimal muscle strength; The optimal muscle strength of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are SOL PCSA, MG PCSA, LG PCSA, TA PCSA is the muscle cross-sectional area PCSA of soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles; SOL V. MG V. LG V. TA V is the muscle volume of soleus, medial gastrocnemius, lateral gastrocnemius and tibialis anterior respectively; HT is the height of the subject, M is the weight of the subject; according to the optimal muscle fiber length The muscle force F is calculated based on the optimal muscle pennation angle α0, muscle pennation angle α and muscle activation signal a. M .

[0138] Further preferably, the muscle-tendon unit stiffness model is:

[0139]

[0140] in,

[0141]

[0142]

[0143]

[0144]

[0145]

[0146]

[0147]

[0148] (α o ) sol =28.3π / 180,(α o ) mg =9.9π / 180,(α o ) lg =12π / 180,(α o ) ta =9.6π / 180

[0149]

[0150] Among them, f p is the proportionality coefficient, Kp is the passive stiffness of the muscle-tendon unit, K MTU is the muscle-tendon unit stiffness, K M is the muscle stiffness, K T is the tendon stiffness, and Tendon stiffness of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively; and The resting lengths of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are is the optimal muscle fiber length; (α o ) sol 、(α o ) mg 、(α o ) lg and (α o ) ta The optimal pennation angles of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are shown in Figure 2. x4 is a mathematical parameter. The resting length of the muscle-tendon unit is given by Optimal muscle fiber length and muscle force F M Calculate the muscle-tendon unit stiffness K MTU .

[0151] Further preferably, the muscle-tendon-joint mapping model is:

[0152]

[0153]

[0154]

[0155]

[0156] In all formulas, the subscript j refers to the jth muscle; is the muscle stiffness of the j-th muscle, is the stiffness of the j-th muscle-tendon unit, is the tendon stiffness of the jth muscle, ΔL j is the deformation of the jth muscle, θ is the ankle joint angle; K j is the stiffness of the jth muscle; K is the joint stiffness; R j is the moment arm of the jth muscle; is the slope of the moment arm of the jth muscle; is the muscle force of the jth muscle; according to the muscle force F M , muscle-tendon unit stiffness K MTU , muscle moment arm R and moment arm slope Calculate the muscle stiffness K of each muscle j , and then calculate the joint stiffness K.

[0157] In general, the above technical solutions conceived by the present invention have the following advantages compared with the prior art:

[0158] Beneficial effects:

[0159] The present invention provides an ankle joint stiffness estimation method and system based on a skeletal muscle anatomical statistical model. Compared with traditional stiffness identification methods, the system does not require the use of proprietary equipment to conduct stiffness identification experiments, making it more convenient and quicker. Compared with other stiffness estimation methods, the system not only does not require stiffness identification, but also does not require estimation model training, greatly reducing the complexity of stiffness estimation. The estimation accuracy is comparable to the identification accuracy, and can well reflect the mechanical impedance characteristics of the human body. BRIEF DESCRIPTION OF THE DRAWINGS

[0160] Figure 1 Schematic diagram of a stiffness identification research experiment scenario provided by an embodiment of the present invention;

[0161] Figure 2 is a flow chart of a method for estimating ankle joint stiffness based on a musculoskeletal anatomical statistical model provided by an embodiment of the present invention;

[0162] Figure 3 This is a diagram of anatomical statistical model estimation results provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0163] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0164] This embodiment is an application case of model accuracy inspection based on the estimation of human lower limb ankle joint stiffness. In stiffness estimation research, in order to estimate the accuracy of the anatomical statistical model, the estimation result of the anatomical statistical model is usually compared with the identification value of the research object. The present invention uses the subject's limb characteristic parameters and surface electromyography data collected in the human ankle joint stiffness identification as the input of the anatomical statistical model, and uses the ankle joint stiffness data identified in the human ankle joint stiffness identification as the reference standard to evaluate the estimation performance of the anatomical statistical model. The experimental scenario of the subject in the human ankle joint stiffness identification study is as follows: Figure 1As shown, the subject sits on a table with their foot fixed to the tool end of a collaborative robotic arm with six rotational joints (the tool end of the 6-DOF collaborative robotic arm is positioned below and in front of the subject, ensuring a 90° angle between the ankle and calf at rest). The angle of the subject's ankle joint is altered by adjusting the rotation angle of the end joint. Simultaneously, reflective markers are placed on the subject's toes, heels, ankles, and knees. An optical motion capture system collects the three-dimensional positions of these markers to obtain displacement information, which is then processed to determine the ankle joint angle. A 6-dimensional force / torque sensor attached to the tool end of the robotic arm collects torque information during ankle joint disturbances. The ankle joint stiffness is determined using the least squares method based on the collected ankle joint angle and torque information.

[0165] like Figure 2 As shown, the present invention provides an anatomical statistical model for ankle joint stiffness, which includes five sub-models, namely a muscle activation dynamics model, a muscle force solution model, a musculoskeletal statistical model, a muscle-tendon unit stiffness model, and a muscle-tendon-joint mapping model;

[0166] a. Muscle Activation Dynamics Model

[0167] The muscle activation dynamics model converts raw EMG signals into muscle activation signals. First, the raw EMG signals are high-pass filtered to remove motion artifacts, followed by full-wave rectification and normalization to complete the EMG signal preprocessing process. The normalized EMG signals are then processed using the neural activation model to obtain neural signals, and finally, the muscle activation signals are obtained based on the muscle activation dynamics model.

[0168] The neural activation signal u of muscle j j (t) and surface electromyography signal e j The relationship between (t) can be expressed as:

[0169] u j (t) = f a (e j (t)) (1)

[0170] Among them, f a is the neural activation model, representing the neural activation signal u j (t) and surface electromyography signal e j (t) relationship, f a It is a direct proportional function with a slope of 1;

[0171] The muscle activation dynamics model has the following expression:

[0172]

[0173] Among them, aj (t) is the activation signal reflecting the activation intensity of the jth muscle, c1+c2 is the activation proportional constant, c2 is the inactivation proportional constant, c1, c2 and the muscle activation time constant t act and the inactivation time constant t deact The following relationship exists:

[0174]

[0175] Where ζ is the muscle activation-inactivation ratio coefficient, taking ζ = 0.5, and the muscle activation time constant t act Take 15ms;

[0176] According to formula (1) to formula (3), using the surface electromyography signal e j (t) can be calculated to get the muscle activation signal a j (t);

[0177] b. Muscle force solution model

[0178] As an important component of the anatomical statistical model, the muscle force solution model is used to calculate the muscle contraction force F according to the degree of muscle activation, the microscopic parameters of the muscle structure and the movement characteristics. M ; According to the muscle activation signal a, it is solved according to the statistical law shown in formula (4):

[0179]

[0180] Among them, f l s is the proportional coefficient, W is the width of the muscle, α is the feathering angle of the muscle, For the optimal muscle fiber length, For optimal muscle strength;

[0181] The proportional coefficient f in formula (4) l s and the muscle width W are calculated using formulas (5) and (6), respectively:

[0182]

[0183]

[0184] in, is the optimal muscle fiber length, α is the muscle pennation angle,

[0185] The optimal muscle fiber length in formula (4) and the muscle pennation angle α are obtained in the musculoskeletal statistical model; optimal muscle force Calculate using formulas (7)(8)(9):

[0186]

[0187]

[0188]

[0189] in, The optimal muscle strength of the four muscles are soleus (SOL), medial gastrocnemius (MG), lateral gastrocnemius (LG) and tibialis anterior (TA) SOL PCSA, MG PCSA, LG PCSA, TA PCSA is the muscle cross-sectional area PCSA of four muscles: soleus (SOL), medial gastrocnemius (MG), lateral gastrocnemius (LG), and tibialis anterior (TA); is the optimal muscle fiber length;

[0190] SOL V. MG V. LG V. TA V is the muscle volume V of the four muscles; HT is the height of the subject, and M is the weight of the subject;

[0191] c. Musculoskeletal statistical model

[0192] The function of the musculoskeletal statistical model is to solve the microscopic parameters of muscle structure based on the characteristic parameters of human limbs. The musculoskeletal statistical model in the present invention establishes the law between human height, weight and muscle microscopic parameters;

[0193] The model between the human body's thigh length, calf length, height, and gender is shown in formula (10):

[0194]

[0195] Wherein, HT is the height of the subject; F sex is the subject's gender parameter. When the subject is male, Π F sex is 0, when the subject is female Π F sex is 1; L thigh is the length of the subject's thigh; L shank is the length of the subject's calf;

[0196] Delving into the internal muscle structure, obtaining various muscle structural parameters is the key to solving muscle force, muscle stiffness, and joint stiffness. Among them, solving muscle force requires the degree of muscle activation, optimal muscle fiber length, optimal muscle force, muscle pennation angle, resting length of muscle-tendon unit, and length of muscle-tendon unit. Obtaining muscle stiffness requires solving for muscle force on the one hand, and optimal muscle fiber length and resting length of muscle-tendon unit on the other. Then, through the spatial relationship between muscle and joint formed by muscle torque arm, muscle-tendon unit stiffness can be converted into joint stiffness.

[0197] Optimal muscle fiber length There is the following relationship between HT and human height:

[0198]

[0199] in, Represents the optimal fiber length of the soleus, medial gastrocnemius, lateral gastrocnemius and tibialis anterior muscles respectively

[0200] The feathering angles α of the four muscles closely related to ankle joint movement in the present invention are:

[0201] (α o ) sol =28.3π / 180,(α o ) mg =9.9π / 180,(α o ) lg =12π / 180,(α o ) ta =9.6π / 180 (12)

[0202] Among them, α SOL , α MG , α LG and α TA represent the pennation angles of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively;

[0203] The resting length of the muscle-tendon unit in the present invention Muscle-tendon unit length L MT The following relationship exists with limb parameters:

[0204]

[0205]

[0206]

[0207] in, The resting lengths of the muscle-tendon unit of the four muscles are the soleus (SOL), medial gastrocnemius (MG), lateral gastrocnemius (LG), and tibialis anterior (TA). SOL L MT 、 MG L MT 、 LG L MT 、 TA L MT The muscle-tendon unit lengths L of the four muscles are MT , is the angle θ with the ankle joint ankle Related functions;

[0208] The mean value R of the muscle moment arm and the slope of the moment arm change within the range of ankle joint motion in the present invention The statistical model expression is:

[0209]

[0210]

[0211]

[0212] in,

[0213]

[0214]

[0215]

[0216] in,

[0217]

[0218]

[0219]

[0220] Wherein, HT is the height of the subject; F (sex) is the subject's gender parameter. When the subject is male, Π F (sex) is 0, when the subject is female, Π F sex is 1; L thigh is the length of the subject's thigh; L shank is the length of the subject's calf; Represents the optimal muscle fiber length of the soleus, medial gastrocnemius, lateral gastrocnemius and tibialis anterior muscles respectively (α o )sol 、(α o ) mg 、(α o ) lg and (α o ) ta are the optimal pennation angles of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively; α SOL , α MG , α LG and α TA represent the pennation angles of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively; The resting lengths of the muscle-tendon unit of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively SOL L MT 、 MG L MT 、 LG L MT 、 TA L MT The muscle-tendon unit lengths L of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are MT , is a function related to the initial ankle angle θ0 and the angle change Δθ; R SOL 、R MG 、R LG 、R TA The muscle moment arms R are the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively. The moment arm slopes of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are K T is tendon stiffness; is the normalized tendon stiffness; f l t 、f l s 、f l p is the proportional coefficient, W is the width of the muscle; a is the muscle activation signal; x1, x2, x3, y1 are mathematical parameters; muscle microscopic parameters are calculated based on human limb parameters, including the resting length of the muscle-tendon unit Optimal muscle fiber length The optimal pennation angle α0 of the muscle, the pennation angle α of the muscle, the muscle moment arm R and the moment arm slope Among them, human limb parameters include: height HT, weight M, ankle joint angle

[0221] d. Muscle-tendon unit stiffness model

[0222] The function of the muscle-tendon unit stiffness model is to calculate the stiffness of the muscle-tendon unit through muscle force and muscle structure parameters. In this invention, the muscle-tendon unit is regarded as a muscle fiber and tendon in series. Therefore, the stiffness K of the muscle-tendon unit is MTU Usually expressed as:

[0223]

[0224] Among them, K M is the muscle stiffness, K T is tendon stiffness;

[0225] Muscle stiffness K M According to the muscle force F M and optimal muscle fiber length To solve, the specific expression is as follows:

[0226]

[0227]

[0228]

[0229] in,

[0230] Tendon stiffness K of the four muscles T Solve the following expression:

[0231]

[0232]

[0233] (α o ) sol =28.3π / 180,(α o ) mg =9.9π / 180,(α o ) lg =12π / 180,(α o ) ta =9.6π / 180

[0234] in, and The resting lengths of the four muscle tendons are and The lengths L of the optimal fibers of the four muscles are O , (α o ) sol 、(α o ) mg 、(αo ) lg 、(α o ) ta The optimal pennation angles of the four muscles are respectively;

[0235] e. Muscle-tendon-joint mapping model

[0236] The function of the muscle-tendon to joint mapping model is to obtain the relationship between the joint stiffness and the muscle-tendon unit stiffness, so as to convert the muscle-tendon unit stiffness obtained by solving formula (17) into the joint stiffness through the spatial relationship. The modeling process of this spatial relationship is as follows:

[0237] Assume that the torque T generated by the jth muscle at the joint j for:

[0238]

[0239] By definition, the value of joint stiffness K is equal to the derivative of joint torque T with respect to joint angle θ, because each muscle-tendon unit at the joint contributes K to its stiffness. j The calculation expression is:

[0240]

[0241] The muscle-tendon-joint mapping relationship obtained by simplifying the above formula is:

[0242]

[0243]

[0244] in, is the muscle stiffness of the j-th muscle, is the stiffness of the j-th muscle-tendon unit, is the tendon stiffness of the jth muscle, ΔL j is the deformation of the jth muscle, θ is the ankle joint angle; K j is the stiffness of the jth muscle; K is the joint stiffness; R j is the moment arm of the jth muscle; is the slope of the moment arm of the jth muscle; is the muscle force of the jth muscle; according to the muscle force F M , muscle-tendon unit stiffness K MTU , muscle moment arm R and moment arm slope Calculate the muscle stiffness K of each muscle j , and then calculate the joint stiffness K.

[0245] Based on the above introduction to the musculoskeletal anatomical statistical model, in one aspect, the present invention provides a method for estimating ankle joint stiffness based on the musculoskeletal anatomical statistical model, comprising the following steps:

[0246] S1: Measure human limb characteristic parameters and surface electromyography signals; the human limb characteristic parameters include ankle joint angle, knee joint angle, height, and weight; the surface electromyography signals include the surface electromyography signals of the tibialis anterior, medial gastrocnemius, lateral gastrocnemius, and soleus muscles that drive the corresponding joints;

[0247] S2: After high-pass filtering to remove motion artifacts, the surface electromyographic signals from each muscle are subjected to full-wave rectification and normalization to complete the electromyographic signal preprocessing process; the preprocessed surface electromyographic signals are input into the muscle activation dynamics model to calculate the activation signals of each muscle;

[0248] S3: Inputting the characteristic parameters of human limbs into the musculoskeletal statistical model, and obtaining the microscopic parameters of muscle structure through the musculoskeletal statistical model;

[0249] S4: Based on the muscle activation signals and the microscopic parameters of each muscle structure, the muscle force and muscle-tendon unit force of each muscle at a specific muscle activation level are calculated using the muscle force solution model;

[0250] S5: Based on the microscopic parameters of each muscle force and muscle structure, the stiffness of each muscle-tendon unit is calculated using the muscle-tendon unit stiffness model;

[0251] S6: Based on the muscle-tendon unit force, the microscopic parameters of the muscle structure, and the stiffness of the muscle-tendon unit, the joint stiffness of the ankle joint driven by the tibialis anterior, medial gastrocnemius, lateral gastrocnemius, and soleus muscles was calculated using the muscle-tendon-joint mapping model.

[0252] In another aspect, the present invention provides an ankle joint stiffness estimation system based on a musculoskeletal anatomical statistical model, comprising:

[0253] A parameter measurement module is used to measure human limb characteristic parameters and surface electromyography signals; wherein the human limb characteristic parameters include: ankle joint angle, knee joint angle, height and weight; the surface electromyography signals include: surface electromyography signals of the tibialis anterior muscle, medial gastrocnemius muscle, lateral gastrocnemius muscle and soleus muscle that drive the corresponding joints;

[0254] The parameter preprocessing module is used to perform high-pass filtering on the surface electromyographic signals from each muscle to remove motion artifacts, and then perform full-wave rectification and normalization to complete the preprocessing of the electromyographic signals;

[0255] A muscle activation signal calculation module is used to input the pre-processed surface electromyography signal into the muscle activation dynamics model to calculate each muscle activation signal;

[0256] A micro-parameter acquisition module is used to input the characteristic parameters of human limbs into the musculoskeletal statistical model and obtain the micro-parameters of muscle structure through the musculoskeletal statistical model;

[0257] The muscle force calculation module is used to calculate the muscle force and muscle-tendon unit force of each muscle at a specific muscle activation level through a muscle force solution model based on the muscle activation signals and the microscopic parameters of each muscle structure;

[0258] The muscle-tendon unit stiffness calculation module is used to calculate the stiffness of each muscle-tendon unit based on the microscopic parameters of each muscle force and muscle structure through the muscle-tendon unit stiffness model;

[0259] The joint stiffness evaluation module is used to calculate the joint stiffness of the ankle joint driven by the tibialis anterior, medial gastrocnemius, lateral gastrocnemius and soleus muscles based on the muscle-tendon unit force, microscopic parameters of the muscle structure and the stiffness of the muscle-tendon unit through the muscle-tendon-joint mapping model.

[0260] The estimated value of the human ankle joint stiffness was calculated using the above anatomical statistical model at a specific muscle activation level; the identified value and estimated value of the ankle joint stiffness were combined and the above data were plotted in the figure. The results are shown in the figure. Figure 3 As shown. Figure 3 It can be found that under three different tibialis anterior muscle activation levels, the anatomical statistical model can accurately estimate the human ankle joint stiffness from the surface electromyography signal.

[0261] It will be easily understood by those skilled in the art that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for estimating ankle joint stiffness based on a musculoskeletal anatomical statistical model, characterized in that: The following steps are involved: S1: Measure human limb characteristic parameters and surface electromyography signals; the human limb characteristic parameters include ankle joint angle, height, and weight; the surface electromyography signals include the surface electromyography signals of the tibialis anterior, medial gastrocnemius, lateral gastrocnemius, and soleus muscles that drive the corresponding joints; S2: After high-pass filtering to remove motion artifacts, the surface electromyographic signals from each muscle are subjected to full-wave rectification and normalization to complete the electromyographic signal preprocessing process; the preprocessed surface electromyographic signals are input into the muscle activation dynamics model to calculate the activation signals of each muscle; S3: Inputting the characteristic parameters of human limbs into the musculoskeletal statistical model, and obtaining the microscopic parameters of muscle structure through the musculoskeletal statistical model; S4: Based on the muscle activation signals and the microscopic parameters of each muscle structure, the muscle force and muscle-tendon unit force of each muscle at a specific muscle activation level are calculated using the muscle force solution model; S5: Based on the microscopic parameters of each muscle force and muscle structure, the stiffness of each muscle-tendon unit is calculated using the muscle-tendon unit stiffness model; S6: calculating the joint stiffness of the ankle joint driven by the tibialis anterior muscle, the medial gastrocnemius muscle, the lateral gastrocnemius muscle, and the soleus muscle through a muscle-tendon-joint mapping model based on the muscle-tendon unit force, the microscopic parameters of the muscle structure, and the stiffness of the muscle-tendon unit; Among them, the musculoskeletal anatomical statistical model includes: muscle force solution model, musculoskeletal statistical model, muscle activation dynamics model, muscle tendon unit stiffness model and muscle tendon-joint mapping model.

2. The ankle joint stiffness estimation method according to claim 1, characterized in that: The musculoskeletal statistical model is: , , , in, in, in, is the height of the subject; is the subject's gender parameter. When the subject is male 0 when the subject is female is 1; is the length of the subject's thigh; is the length of the subject's calf; 、 、 、 Represents the optimal muscle fiber length of the soleus, medial gastrocnemius, lateral gastrocnemius and tibialis anterior muscles respectively ; 、 、 and The optimal pennation angles are for the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively; 、 、 and represent the pennation angles of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively; 、 、 、 The resting lengths of the muscle-tendon unit of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively ; 、 、 、 The muscle-tendon unit lengths of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively , The initial angle of the ankle joint and angle changes Related functions; 、 、 、 Muscle moment arms for the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles , 、 、 、 The moment arm slopes of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are ; is tendon stiffness; is the normalized tendon stiffness; 、 、 is the proportionality coefficient, is the width of the muscle; Activate signals for muscles; , , , is a mathematical parameter; muscle microscopic parameters are calculated based on human limb parameters, including the resting length of the muscle-tendon unit , optimal muscle fiber length , optimal feathering angle of muscle , the pennate angle of the muscle , muscle moment arm and the moment arm slope ; Among them, human limb parameters include: height ,weight , ankle joint angle ; For optimal muscle strength.

3. The ankle joint stiffness estimation method according to claim 1 or 2, characterized in that: The muscle activation dynamics model is: in, is a neural activation model; It is a nerve activation signal; is the surface electromyographic signal; It reflects the The activation signal of the muscle activation intensity, It is The activation rate of each muscle; is the activation rate constant, corresponding to =1; is the inactivation rate constant, corresponding to =0; is the muscle activation time constant; is the inactivation time constant; is the muscle activation-inactivation ratio coefficient; according to the surface electromyography signal Calculating muscle activation signals ; The muscle force solution model is: in, For muscle strength, is the proportionality coefficient, is the width of the muscle, The optimal feathering angle for the muscle, The pennate angle of the muscle, For the optimal muscle fiber length, For optimal muscle strength; 、 、 、 The optimal muscle strength of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are ; 、 、 、 Muscle cross-sectional areas of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively ; 、 、 、 The muscle volumes of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are ; is the subject's height, is the subject's weight; according to the optimal muscle fiber length , optimal feathering angle of muscle , muscle pennate angle and muscle activation signals Calculating muscle force .

4. The ankle joint stiffness estimation method according to claim 1 or 3, characterized in that: The muscle-tendon unit stiffness model is: in, ， , , , in, is the proportionality coefficient, is the passive stiffness of the muscle-tendon unit, is the muscle-tendon unit stiffness, is muscle stiffness, is the tendon stiffness, 、 、 and Tendon stiffness of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively; 、 、 and The resting lengths of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are ; is the optimal muscle fiber length; 、 、 and The optimal pennation angles are for the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively; is a mathematical parameter; based on the resting length of the muscle-tendon unit , optimal muscle fiber length and muscle strength Calculating muscle-tendon unit stiffness .

5. The method for estimating human joint stiffness according to claim 1 or 4, wherein: The muscle-tendon-joint mapping model is: Among them, when multiple muscles act on the same joint and drive the same joint movement, the subscript added to the mathematical variable j Indicates the muscle mass; For the Muscle stiffness of the muscle, For the Muscle-tendon unit stiffness, For the The tendon stiffness of the muscle For the The amount of muscle deformation, is the ankle joint angle; For the The stiffness of the muscle; is the joint stiffness; For the The moment arm of a muscle; For the The slope of the moment arm of the muscle mass; For the muscle strength of a muscle; according to muscle strength , muscle-tendon unit stiffness , muscle moment arm and the moment arm slope Calculate muscle stiffness for each muscle , and then calculate the joint stiffness .

6. An ankle joint stiffness estimation system based on a musculoskeletal anatomical statistical model, characterized in that: include: A parameter measurement module is used to measure human limb characteristic parameters and surface electromyography signals; wherein the human limb characteristic parameters include: ankle joint angle, height and weight; the surface electromyography signals include: surface electromyography signals of the tibialis anterior muscle, medial gastrocnemius muscle, lateral gastrocnemius muscle and soleus muscle that drive the corresponding joints; The parameter preprocessing module is used to perform high-pass filtering on the surface electromyographic signals from each muscle to remove motion artifacts, and then perform full-wave rectification and normalization to complete the preprocessing of the electromyographic signals; A muscle activation signal calculation module is used to input the pre-processed surface electromyography signal into the muscle activation dynamics model to calculate each muscle activation signal; A micro-parameter acquisition module is used to input the characteristic parameters of human limbs into the musculoskeletal statistical model and obtain the micro-parameters of muscle structure through the musculoskeletal statistical model; The muscle force calculation module is used to calculate the muscle force and muscle-tendon unit force of each muscle at a specific muscle activation level through a muscle force solution model based on the muscle activation signals and the microscopic parameters of each muscle structure; The muscle-tendon unit stiffness calculation module is used to calculate the stiffness of each muscle-tendon unit based on the microscopic parameters of each muscle force and muscle structure through the muscle-tendon unit stiffness model; The joint stiffness evaluation module is used to calculate the joint stiffness of the ankle joint driven by the tibialis anterior, medial gastrocnemius, lateral gastrocnemius and soleus muscles based on the muscle-tendon unit force, microscopic parameters of the muscle structure and the stiffness of the muscle-tendon unit through the muscle-tendon-joint mapping model.

7. The ankle joint stiffness estimation system according to claim 6, characterized in that: The musculoskeletal statistical model is: , , , in, in, in, is the height of the subject; is the subject's gender parameter. When the subject is male 0 when the subject is female is 1; is the length of the subject's thigh; is the length of the subject's calf; 、 、 、 Represents the optimal muscle fiber length of the soleus, medial gastrocnemius, lateral gastrocnemius and tibialis anterior muscles respectively ; 、 、 and The optimal pennation angles are for the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively; 、 、 and represent the pennation angles of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively; 、 、 、 The resting lengths of the muscle-tendon unit of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively ; 、 、 、 The muscle-tendon unit lengths of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively , The initial angle of the ankle joint and angle changes Related functions; 、 、 、 Muscle moment arms for the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles , 、 、 、 The moment arm slopes of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are ; is tendon stiffness; is the normalized tendon stiffness; 、 、 is the proportionality coefficient, is the width of the muscle; Activate signals for muscles; , , , is a mathematical parameter; muscle microscopic parameters are calculated based on human limb parameters, including the resting length of the muscle-tendon unit , optimal muscle fiber length , optimal feathering angle of muscle , the pennate angle of the muscle , muscle moment arm and the moment arm slope ; Among them, human limb parameters include: height ,weight , ankle joint angle ; For optimal muscle strength.

8. The ankle joint stiffness estimation system according to claim 6 or 7, characterized in that: The muscle activation dynamics model is: in, is a neural activation model; It is a nerve activation signal; is the surface electromyographic signal; It reflects the The activation signal of the muscle activation intensity, It is The activation rate of each muscle; is the activation rate constant, corresponding to =1; is the inactivation rate constant, corresponding to =0; is the muscle activation time constant; is the inactivation time constant; is the muscle activation-inactivation ratio coefficient; according to the surface electromyography signal Calculating muscle activation signals ; The muscle force solution module is: in, For muscle strength, is the proportionality coefficient, is the width of the muscle, The optimal feathering angle for the muscle, The pennate angle of the muscle, For the optimal muscle fiber length, For optimal muscle strength; 、 、 、 The optimal muscle strength of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are ; 、 、 、 Muscle cross-sectional areas of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively ; 、 、 、 The muscle volumes of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are ; is the subject's height, is the subject's weight; according to the optimal muscle fiber length , optimal feathering angle of muscle , muscle pennate angle and muscle activation signals Calculating muscle force .

9. The ankle joint stiffness estimation system according to claim 6 or 8, characterized in that: The muscle-tendon unit stiffness model is: in, ， , , , in, is the proportionality coefficient, is the passive stiffness of the muscle-tendon unit, is the muscle-tendon unit stiffness, is muscle stiffness, is the tendon stiffness, 、 、 and Tendon stiffness of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively; 、 、 and The resting lengths of the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles are ; is the optimal muscle fiber length; 、 、 and The optimal pennation angles are for the soleus, medial gastrocnemius, lateral gastrocnemius, and tibialis anterior muscles, respectively; is a mathematical parameter; based on the resting length of the muscle-tendon unit , optimal muscle fiber length and muscle strength Calculating muscle-tendon unit stiffness .

10. The ankle joint stiffness estimation system according to claim 6 or 8, characterized in that: The muscle-tendon-joint mapping model is: When multiple muscles act on the same joint and drive the same joint movement, the subscript added to the mathematical variable is Indicates the muscle mass; For the Muscle stiffness of the muscle, For the Muscle-tendon unit stiffness, For the The tendon stiffness of the muscle For the The amount of muscle deformation, is the ankle joint angle; For the The stiffness of the muscle; is the joint stiffness; For the The moment arm of a muscle; For the The slope of the moment arm of the muscle mass; For the muscle strength of a muscle; according to muscle strength , muscle-tendon unit stiffness , muscle moment arm and the moment arm slope Calculate muscle stiffness for each muscle , and then calculate the joint stiffness .

Images