Psoas force recruitment in full-body musculoskeletal movement simulations is restored with a geometrically informed cost function weighting

Simulations of musculoskeletal models are useful for estimating internal muscle and joint forces. However, predicted forces rely on optimization and modeling formulations. Geometric detail is important to predict muscle forces, and greater geometric complexity is required for muscles that have broad attachments or span many joints, as in the torso. However, the extent to which optimized muscle force recruitment is sensitive to these geometry choices is unclear. We developed level, uphill and downhill sloped walking simulations using a standard (uniformly weighted, "fatigue-like") cost function with lower limb and full-body musculoskeletal models to evaluate hip muscle recruitment with different geometric representations of the psoas muscle under walking conditions with varying hip moment demands. We also tested a novel cost function formulation where muscle activations were weighted according to the modeled geometric detail in the full-body model. Total psoas force was less and iliacus, rectus femoris, and other hip flexors' force was greater when psoas was modeled with greater geometric detail compared to other hip muscles for all slopes. The proposed weighting scheme restored hip muscle force recruitment without sacrificing detailed psoas geometry. In addition, we found that lumbar, but not hip, joint contact forces were influenced by psoas force recruitment. Our results demonstrate that static optimization dependent simulations using models comprised of muscles with different amounts of geometric detail bias force recruitment toward muscles with less geometric detail. Muscle activation weighting that accounts for differences in geometric complexity across muscles corrects for this recruitment bias.

The biomechanical influence of transtibial Bone-Anchored limbs during walking

Individuals with unilateral transtibial amputation (TTA) using socket prostheses demonstrate asymmetric joint biomechanics during walking, which increases the risk of secondary comorbidities (e.g., low back pain (LBP), osteoarthritis (OA)). Bone-anchored limbs are an alternative to socket prostheses, yet it remains unknown how they influence multi-joint loading. Our objective was to determine the influence of bone-anchored limb use on multi-joint biomechanics during walking. Motion capture data (kinematics, ground reaction forces) were collected during overground walking from ten participants with unilateral TTA prior to (using socket prostheses) and 12-months after bone-anchored limb implantation. Within this year, each participant completed a rehabilitation protocol that guided progression of loading based on patient pain response and optimized biomechanics. Musculoskeletal models were developed at each testing timepoint (baseline or 12-months after implantation) and used to calculate joint kinematics, internal joint moments, and joint reaction forces (JRFs). Analyses were performed during three stance periods on each limb. The between-limb normalized symmetry index (NSI) was calculated for joint moments and JRF impulses. Discrete (range of motion (ROM), impulse NSI) dependent variables were compared before and after implantation using paired t-tests with Bonferroni-Holm corrections while continuous (ensemble averages of kinematics, moments, JRFs) were compared using statistical parametric mapping (p < 0.05). When using a bone-anchored limb, frontal plane pelvic (residual: pre = 9.6 ± 3.3°, post = 6.3 ± 2.5°, p = 0.004; intact: pre = 10.2 ± 3.9°, post = 7.9 ± 2.6°, p = 0.006) and lumbar (residual: pre = 15.9 ± 7.0°, post = 10.6 ± 2.5°, p = 0.024, intact: pre = 17.1 ± 7.0°, post = 11.4 ± 2.8°, p = 0.014) ROM was reduced compared to socket prosthesis use. The intact limb hip extension moment impulse increased (pre = -11.0 ± 3.6 Nm*s/kg, post = -16.5 ± 4.4 Nm*s/kg, p = 0.005) and sagittal plane hip moment impulse symmetry improved (flexion: pre = 23.1 ± 16.0 %, post = -3.9 ± 19.5 %, p = 0.004, extension: pre = 29.2 ± 20.3 %, post = 8.7 ± 22.9 %, p = 0.049). Residual limb knee extension moment impulse decreased compared to baseline (pre = 15.7 ± 10.8 Nm*s/kg, post = 7.8 ± 3.9 Nm*s/kg, p = 0.030). These results indicate that bone-anchored limb implantation alters multi-joint biomechanics, which may impact LBP or OA risk factors in the TTA population longitudinally.

Estimation of lower back muscle force in a lifting task using wearable IMUs

Low back pain is commonly reported in occupational settings due to factors such as heavy lifting and poor ergonomic practices, often resulting in significant healthcare expenses and lowered productivity. Assessment tools for human motion and ergonomic risk at the workplace are still limited. Therefore, this study aimed to assess lower back muscle and joint reaction forces in laboratory conditions using wearable inertial measurement units (IMUs) during weight lifting, a frequently high-risk workplace task. Ten able-bodied participants were instructed to lift a 28 lbs. box while surface electromyography sensors, IMUs, and a camera-based motion capture system recorded their muscle activity and body motion. The data recorded by IMUs and motion capture system were used to estimate lower back muscle and joint reaction forces via musculoskeletal modeling. Lower back muscle patterns matched well with electromyography recordings. The normalized mean absolute differences between muscle forces estimated based on measurements of IMUs and cameras were less than 25 %, and the statistical parametric mapping results indicated no significant difference between the forces estimated by both systems. However, abrupt changes in motion, such as lifting initiation, led to significant differences (p < 0.05) between the muscle forces. Furthermore, the maximum L5-S1 joint reaction force estimated using IMU data was significantly lower (p < 0.05) than those estimated by cameras during weight lifting and lowering. The study showed how kinematic errors from IMUs propagated through the musculoskeletal model and affected the estimations of muscle forces and joint reaction forces. Our findings showed the potential of IMUs for in-field ergonomic risk evaluations.

Contribution of Soft Tissue Passive Forces in Thumb Carpometacarpal Joint Distraction

Thumb carpometacarpal joint space changes when the surrounding soft tissues including the capsule, ligaments, and tendons are stretched or pulled away. When at rest, joint forces originate from passive contraction of muscles and the involvement of joint capsule and ligaments. Previous biomechanical models of hand and finger joints have mostly focused on the assessment of joint properties when muscles were active. This study aims to present an experimental-numerical biomechanical model of thumb carpometacarpal joint to explore the contribution of tendons, ligaments, and other soft tissues in the passive forces during distraction. Five fresh cadaveric specimens were tested using a distractor device to measure the applied forces upon gradual distraction of the intact joint. The subsequent step involved inserting a minuscule sensor into the joint capsule through a small incision, while preserving the integrity of tendons and ligaments, in order to accurately measure the fundamental intra-articular forces. A numerical model was also used to calculate the passive forces of tendons and ligaments. Before the separation of bones, the forces exerted by tendons and ligaments were relatively small compared to the capsule force, which accounted for approximately 92% of the total applied force. Contribution of tendons and ligaments, however, increased by further distraction. The passive force contribution by tendons at 2-mm distraction was determined less than 11%, whereas it reached up to 74% for the ligaments. The present study demonstrated that the ligament-capsule complex plays significant contribution in passive forces of thumb carpometacarpal joint during distraction.

Explainable AI Elucidates Musculoskeletal Biomechanics: A Case Study Using Wrist Surgeries

As datasets increase in size and complexity, biomechanists have turned to artificial intelligence (AI) to aid their analyses. This paper explores how explainable AI (XAI) can enhance the interpretability of biomechanics data derived from musculoskeletal simulations. We use machine learning to classify the simulated lateral pinch data as belonging to models with healthy or one of two types of surgically altered wrists. This simulation-based classification task is analogous to using biomechanical movement and force data to clinically diagnose a pathological state. The XAI describes which musculoskeletal features best explain the classifications and, in turn, the pathological states, at both the local (individual decision) level and global (entire algorithm) level. We demonstrate that these descriptions agree with assessments in the literature and additionally identify the blind spots that can be missed with traditional statistical techniques.

Full-length radiograph based automatic musculoskeletal modeling using convolutional neural network

Full-length radiographs contain information from which many anatomical parameters of the pelvis, femur, and tibia may be derived, but only a few anatomical parameters are used for musculoskeletal modeling. This study aimed to develop a fully automatic algorithm to extract anatomical parameters from full-length radiograph to generate a musculoskeletal model that is more accurate than linear scaled one. A U-Net convolutional neural network was trained to segment the pelvis, femur, and tibia from the full-length radiograph. Eight anatomic parameters (six for length and width, two for angles) were automatically extracted from the bone segmentation masks and used to generate the musculoskeletal model. Sørensen-Dice coefficient was used to quantify the consistency of automatic bone segmentation masks with manually segmented labels. Maximum distance error, root mean square (RMS) distance error and Jaccard index (JI) were used to evaluate the geometric accuracy of the automatically generated pelvis, femur and tibia models versus CT bone models. Mean Sørensen-Dice coefficients for the pelvis, femur and tibia 2D segmentation masks were 0.9898, 0.9822 and 0.9786, respectively. The algorithm-driven bone models were closer to the 3D CT bone models than the scaled generic models in geometry, with significantly lower maximum distance error (28.3 % average decrease from 24.35 mm) and RMS distance error (28.9 % average decrease from 9.55 mm) and higher JI (17.2 % average increase from 0.46) (P < 0.001). The algorithm-driven musculoskeletal modeling (107.15 ± 10.24 s) was faster than the manual process (870.07 ± 44.79 s) for the same full-length radiograph. This algorithm provides a fully automatic way to generate a musculoskeletal model from full-length radiograph that achieves an approximately 30 % reduction in distance errors, which could enable personalized musculoskeletal simulation based on full-length radiograph for large scale OA populations.

Muscle contributions to reduced ankle joint contact force during drop vertical jumps in patients with chronic ankle instability

Chronic ankle instability is a condition linked to progressive early ankle joint degeneration. Patients with chronic ankle instability exhibit altered biomechanics during gait and jump landings and these alterations are believed to contribute to aberrant joint loading and subsequent joint degeneration. Musculoskeletal modeling has the capacity to estimate joint loads from individual muscle forces. However, the influence of chronic ankle instability on joint contact forces remains largely unknown. The objective of this study was to compare tri-axial (i.e., compressive, anterior-posterior, and medial-lateral) ankle joint contact forces between those with and without chronic ankle instability during the ground contact phase of a drop vertical jump. Fifteen individuals with and 15 individuals without chronic ankle instability completed drop vertical jump maneuvers in a research laboratory. We used those data to drive three-dimensional musculoskeletal simulations and estimate muscle forces and tri-axial joint contact force variables (i.e., peak and impulse). Compared to those without chronic ankle instability, the ankles of patients with chronic ankle instability underwent lower compressive ankle joint contact forces as well as lower anterior-posterior and medial-lateral shearing forces during the weight acceptance phase of landing (p <.05). These findings suggest that patients with chronic ankle instability exhibit lower ankle joint loading patterns than uninjured individuals during a drop vertical jump, which may be considered in rehabilitation to potentially reduce the risk of early onset of ankle joint degeneration.

Speed and surface steepness affect internal tibial loading during running

BACKGROUND

Internal tibial loading is influenced by modifiable factors with implications for the risk of stress injury. Runners encounter varied surface steepness (gradients) when running outdoors and may adapt their speed according to the gradient. This study aimed to quantify tibial bending moments and stress at the anterior and posterior peripheries when running at different speeds on surfaces of different gradients.

METHODS

Twenty recreational runners ran on a treadmill at 3 different speeds (2.5 m/s, 3.0 m/s, and 3.5 m/s) and gradients (level: 0%; uphill: +5%, +10%, and +15%; downhill: -5%, -10%, and -15%). Force and marker data were collected synchronously throughout. Bending moments were estimated at the distal third centroid of the tibia about the medial-lateral axis by ensuring static equilibrium at each 1% of stance. Stress was derived from bending moments at the anterior and posterior peripheries by modeling the tibia as a hollow ellipse. Two-way repeated-measures analysis of variance were conducted using both functional and discrete statistical analyses.

RESULTS

There were significant main effects for running speed and gradient on peak bending moments and peak anterior and posterior stress. Higher running speeds resulted in greater tibial loading. Running uphill at +10% and +15% resulted in greater tibial loading than level running. Running downhill at -10% and -15% resulted in reduced tibial loading compared to level running. There was no difference between +5% or -5% and level running.

CONCLUSION

Running at faster speeds and uphill on gradients ≥+10% increased internal tibial loading, whereas slower running and downhill running on gradients ≥-10% reduced internal loading. Adapting running speed according to the gradient could be a protective mechanism, providing runners with a strategy to minimize the risk of tibial stress injuries.

Concentric and eccentric hip musculotendon work depends on backpack loads and walking slopes

Hip muscle weakness is associated with low back and leg injuries. In addition, hiking with heavy loads is linked to high incidence of overuse injuries. Walking with heavy loads on slopes alters hip biomechanics compared to unloaded walking, but individual muscle mechanical work in these challenging conditions is unknown. Using movement simulations, we quantified hip muscle concentric and eccentric work during walking on 0° and ±10° slopes with, and without 40% bodyweight added loads, and with and without a hip belt. For gluteus maximus, psoas, iliacus, gluteus medius, and biceps femoris long head, both concentric and eccentric work were greatest during uphill walking. For rectus femoris and semimembranosus, concentric work was greatest during uphill and eccentric work was greatest during downhill walking. Loaded walking had greater concentric and eccentric work from rectus femoris, biceps femoris long head, and gluteus maximus. Psoas concentric work was greatest while carrying loads regardless of hip belt usage, but eccentric work was only greater than unloaded walking when using a hip belt. Loaded and uphill walking had high concentric work from gluteus maximus, and high eccentric work from gluteus medius and biceps femoris long head. Carrying heavy loads uphill may lead to excessive hip muscle fatigue and heightened injury risk. Effects of the greater eccentric work from hip flexors when wearing a hip belt on lumbar spine forces and pelvic stability should be investigated. Military and other occupational groups who carry heavy backpacks with hip belts should maintain eccentric strength of hip flexors and hamstrings.

Effect of personalized spinal profile on its biomechanical response in an EMG-assisted optimization musculoskeletal model of the trunk

Recent developments in musculoskeletal (MS) modeling have been geared towards model customization. Personalization of the spine profile could affect estimates of spinal loading and stability, particularly in the upright standing posture where large inter-subject variations in the lumbar lordosis have been reported. This study investigates the biomechanical consequences of changes in the spinal profile. In 31 participants (healthy and with back pain), (1) the spine external profile was measured, (2) submaximal contractions were recorded in a dynamometer to calibrate the EMG-driven MS model and finally (3) static lifting in the upright standing challenging spine stability while altering load position and magnitude were considered. EMG signals of 12 trunk muscles and angular kinematics of 17 segments were recorded. For each participant, the MS model was constructed using either a generic or a personalized spinal profile and 17 biomechanical outcomes were computed, including individual muscle forces, ratios of muscle group forces, spinal loading and stability parameters. According to the ANOVA results and corresponding effect sizes, personalizing the spine profile induced medium and large effects on about half MS model outcomes related to the trunk muscle forces and negligible to small effects on spinal loading and stability as more aggregate outcomes. These effects are explained by personalized spine profiles that were a little more in extension as well as more pronounced spine curvatures (lordosis and kyphosis). These findings suggest that spine profile personalization should be considered in MS spine modeling as it may impact muscle force prediction and spinal loading.

Predictions of thumb, hand, and arm muscle parameters derived using force measurements of varying complexity and neural networks

Subject-specific musculoskeletal models are a promising avenue for personalized healthcare. However, current methods for producing personalized models require dense, biomechanical datasets that include expensive and time-consuming physiological measurements. For personalized models to be clinically useful, we must be able to rapidly generate models from simple, easy to collect data. In this context, the objective of this paper is to evaluate if and how simple data, namely height/weight and pinch force data, can be used to achieve model personalization via machine learning. Using simulated lateral pinch force measurements from a synthetic population of 40,000 randomly generated subjects, we train neural networks to estimate four Hill-type muscle model parameters and bone density. We compare parameter estimates to the true parameters of 10,000 additional synthetic subjects. We also generate new personalized models using the parameter estimates and perform new lateral pinch simulations to compare predicted forces using these personalized models to those generated using a baseline model. We demonstrate that increasing force measurement complexity reduces the root-mean-square error in the majority of parameter estimates. Additionally, musculoskeletal models using neural network-based parameter estimates provide up to an 80% reduction in absolute error in simulated forces when compared to a generic model. Thus, easily obtained force measurements may be suitable for personalizing models of the thumb, although extending the method to more tasks and models involving other joints likely requires additional measurements.

Validity of evaluating spinal kinetics without participant-specific kinematics

Musculoskeletal models are commonly used to estimate in vivo spinal loads under various loading conditions. Typically, participant-specific measured kinematics (PSMK) are coupled with participant-specific models, but obtaining PSMK data can be costly and infeasible in large studies or clinical practice. Thus, we evaluated two alternative methods to estimate spinal loads without PSMK: 1) ensemble average kinematics (EAK) based on kinematics from all participants; and 2) using separately measured individual kinematics (SMIK) from multiple other participants as inputs, then averaging the resulting loads. This study compares the dynamic spine loading patterns and peak loads in older adults performing five lifting tasks using PSMK, EAK and SMIK. Median root mean square errors of EAK and SMIK methods versus PSMK ranged from 18 to 72% body weight for compressive loads and from 2 to 25% body weight for shear loads, with median cross-correlations ranging from 0.931 to 0.991. The root mean square errors and cross-correlations between repeated PSMK trials fell within similar ranges. Compressive peak loads evaluated by EAK and SMIK were not different than PSMK in 12 of 15 cases, while by comparison repeated PSMK trials were not different in 13 of 15 cases. Overall, the resulting spine loading magnitudes and profiles using EAK or SMIK were not notably different than using a PSMK approach, and differences were not greater than between two PSMK trials. Thus, these findings indicate that these approaches may be used to make reasonable estimates of dynamic spinal loading without direct measurement of participant kinematics.

Accurately and effectively predict the ACL force: Utilizing biomechanical landing pattern before and after-fatigue

BACKGROUND AND OBJECTIVE

As a fundamental exercise technique, landing can commonly be associated with anterior cruciate ligament (ACL) injury, especially during after-fatigue single-leg landing (SL). Presently, the inability to accurately detect ACL loading makes it difficult to recognize the risk degree of ACL injury, which reduces the effectiveness of injury prevention and sports monitoring. Increased risk of ACL injury during after-fatigue SL may be related to changes in ankle motion patterns. Therefore, this study aims to develop a highly accurate and easily implemented ACL force prediction model by combining deep learning and the explored relationship between ACL force and ankle motion pattern.

METHODS

First, 56 subjects' during before and after-fatigue SL data were collected to explore the relationship between the ankle initial contact angle (AIC), ankle range of motion (AROM) and peak ACL force (PAF). Then, the musculoskeletal model was developed to simulate and calculate the ACL force. Finally, the ACL force prediction model was constructed by combining the explored relationship and sparrow search algorithm (SSA) to optimize the extreme learning machine (ELM) and long short-term memory (LSTM).

RESULTS

There was almost a stronger linear relationship between the PAF and AIC (R = -0.70), AROM (R2 = -0.61). By substituting AIC and AROM as independent variables in the SSA-ELM prediction model, the model shows excellent prediction performance because of very strong correlation (R2 = 0.9992,  MSE = 0.0023,  RMSE = 0.0474). Based on the equal scaling by combining results of SSA-ELM and SSA-LSTM, the prediction model achieves excellent performance in ACL force prediction of the overall waveform (R2 = 0.9947,  MSE = 0.0076,  RMSE = 0.0873).

CONCLUSION

By increasing the AIC and AROM during SL, the lower limb joint energy dissipation can be increased and the PAF reduced, thus reducing the impact loads on the lower limb joints and reducing ACL injuries. The proposed ACL dynamic load force prediction model has low input variable demands (sagittal joint angles), excellent generalization capabilities and superior performance in terms of high accuracy. In the future, we plan to use it as an accurate ACL injury risk assessment tool to promote and apply it to a wider range of sports training and injury monitoring.

Computational modelling of the fossa component fixation associated with alloplastic total temporomandibular joint replacements

The alloplastic total temporomandibular joint (TMJ) replacement is a complex surgical approach to end-stage TMJ disorders. The fixation of TMJ prostheses remains a critical issue for implant design and performance. For the fossa component, it is generally considered to use fixation screws to achieve tripod stability. However, the fossa may still come loose, and the mechanism remains unknown. A computational framework, consisting of a musculoskeletal model for calculating muscle and TMJ forces, and a finite element model for the fossa fixation simulation, was developed. A polyethylene (PE) fossa with stock prosthesis design was analyzed to predict contact pressures at the fixation interfaces, and stresses/strains in the fossa implant and bone during the static loading of normal chewing bite and maximum-force bite. The predicted maximum von Mises stresses were 33 MPa and 44 MPa for the bone, 13 MPa and 28 MPa for the PE fossa, and 131 MPa and 244 MPa for the screws, for the normal and maximum bites, respectively; the peak minimum principal strain was in the range of -2514 ∼ -3545 με for the bone. The results show that the sufficient initial mechanical strength of the fossa component fixation can be established using the screws in combination with bone support. The functional loads applied through the prosthetic TMJ bearing can be largely transferred to supporting bone without causing high level stresses. Tightening fixation screws with a pretension of 100 N can reduce transverse load to the screws and help prevent screw loosening. Further research is recommended to accurately quantify the transverse load and its influence on screw loosening during dynamic loading, and the frictional properties at the bone-implant interface.

Biomechanical modeling for the estimation of muscle forces: toward a common language in biomechanics, medical engineering, and neurosciences

Different research fields, such as biomechanics, medical engineering or neurosciences take part in the development of biomechanical models allowing for the estimation of individual muscle forces involved in motor action. The heterogeneity of the terminology used to describe these models according to the research field is a source of confusion and can hamper collaboration between the different fields. This paper proposes a common language based on lexical disambiguation and a synthesis of the terms used in the literature in order to facilitate the understanding of the different elements of biomechanical modeling for force estimation, without questioning the relevance of the terms used in each field or the different model components or their interest. We suggest that the description should start with an indication of whether the muscle force estimation problem is solved following the physiological movement control (from the nervous drive to the muscle force production) or in the opposite direction. Next, the suitability of the model for force production estimation at a given time or for monitoring over time should be specified. Authors should pay particular attention to the method description used to find solutions, specifying whether this is done during or after data collection, with possible method adaptations during processing. Finally, the presence of additional data must be specified by indicating whether they are used to drive, assist, or calibrate the model. Describing and classifying models in this way will facilitate the use and application in all fields where the estimation of muscle forces is of real, direct, and concrete interest.

Individual Muscle Force Differences During Loaded Hexbar Jumps: A Statistical Parametric Mapping Analysis

Knowledge of individual muscle force during strength and conditioning exercises provides deeper understanding of how specific training decisions relate to desired training outcomes. The purpose of this study was to estimate individual muscle forces during hexbar jumps with 0%, 20%, 40%, and 60% of the hexbar deadlift 1-repetition maximum utilizing in vivo motion capture and computational modeling techniques of male participants. Muscle forces for the gluteus maximus, biceps femoris, rectus femoris, vastus intermedius, gastrocnemius, and soleus were estimated via static optimization. Changes in muscle forces over the concentric phase were analyzed across loading conditions using statistical parametric mapping, impulse, and peak values. Conclusions about the effects of load differ between the three analysis methods; therefore, careful selection of analysis method is essential. Peaks may be inadequate in assessing differences in muscle force during dynamic movements. If SPM, assessing point-by-point differences, is combined with impulse, where time of force application is considered, both timepoint and overall loading can be analyzed. The response of individual muscle forces to increases in external load, as assessed by impulse and SPM, includes increased total muscle output, proportionally highest at 20%1RM, and increased absolute force for the vasti and plantarflexors during the concentric phase of hexbar jumps.

Added mass increases Achilles tendon stress in female runners

CONTEXT

Achilles tendon (AT) injuries are common in female runners and military personnel where increased AT loading may be a contributing factor. Few studies have examined AT stress during running with added mass. The purpose was to examine the stress, strain, and force placed on the AT, kinematics and temporospatial variable in running with different amounts of added mass.

DESIGN

Repeated measure design METHODS: Twenty-three female runners with a rear-foot strike pattern were participants. AT stress, strain, and force were measured during running using a musculoskeletal model that used kinematic (180 Hz) and kinetic data (1800 Hz) as input. Ultrasound data were used to measure AT cross sectional area. A repeated measures multivariate analysis of variance (α = 0.05) was used on AT loading variables, kinematics and temporospatial variables.

RESULTS

Peak AT stress, strain, and force were greatest during the 9.0 kg added load running condition (p < .0001). There was a 4.3% and 8.8% increase in AT stress and strain during the 4.5 kg and 9.0 kg added load conditions, respectively, compared to baseline. Kinematics at the hip and knee changed with added load but not at the ankle. Small changes in temporospatial variables were seen.

CONCLUSION

Added load increased stress on the AT during running. There may be an increased risk for AT injury with added load. Individuals may consider slowly progressing training with added load to allow for increased AT loading.

Prediction of gait kinetics using Markerless-driven musculoskeletal modeling

Video-based motion analysis systems are emerging in the biomechanics research community, yet there is limited exploration of kinetics prediction using RGB-markerless kinematics and musculoskeletal modeling. This project aimed to provide ground reaction force (GRF) and ground reaction moment (GRM) predictions during over-ground gait by introducing RGB-markerless kinematics into a musculoskeletal modeling framework. Full-body markerless kinematic inputs and musculoskeletal modeling were used to obtain GRF and GRM predictions which were compared to measured force plate values. The markerless-driven predictions yielded average root mean-squared error (RMSE) in the stance phase of 0.035 ± 0.009 N∙BW-1, 0.070 ± 0.014 N∙BW-1, and 0.155 ± 0.041 N∙BW-1 in the mediolateral (ML), anteroposterior (AP), and vertical (V) GRFs. This was accompanied by moderate to high correlations and interclass correlation coefficients (ICC) indicating moderate to good agreement between measured and predicted values (95% Confidence Inervals: ML = [0.479, 0.717], AP = [0.714, 0.856], V = [0.803, 0.905]). For ground reaction moments (GRM), average RMSE was 0.029 ± 0.013 Nm∙BWH-1, 0.014 ± 0.005 Nm∙BWH-1, and 0.005 ± 0.002 Nm∙BWH-1 in the sagittal, frontal, and transverse planes. Pearson correlations and ICCs indicated poor agreement between systems for GRMs (95% Confidence Intervals: Sagittal = [0.314, 0.608], Frontal = [0.006, 0.373], Transverse = [0.269, 0.570]). Currently, RMSE is larger than target thresholds set from studies using Kinect, inertial, or marker-based kinematic drivers; but methodological considerations highlighted in this work may help guide follow-up iterations. At this point, further use in research or clinical practice is cautioned until methodological considerations are addressed, although results are promising at this point.

Ten steps to becoming a musculoskeletal simulation expert: A half-century of progress and outlook for the future

Over the past half-century, musculoskeletal simulations have deepened our knowledge of human and animal movement. This article outlines ten steps to becoming a musculoskeletal simulation expert so you can contribute to the next half-century of technical innovation and scientific discovery. We advocate looking to the past, present, and future to harness the power of simulations that seek to understand and improve mobility. Instead of presenting a comprehensive literature review, we articulate a set of ideas intended to help researchers use simulations effectively and responsibly by understanding the work on which today's musculoskeletal simulations are built, following established modeling and simulation principles, and branching out in new directions.

Unilateral transfemoral osseointegrated prostheses improve joint loading during walking

People with unilateral transfemoral amputation using socket prostheses are at increased risk for developing osteoarthritis in both the residual hip and intact lower-limb joints. Osseointegrated prostheses are a surgical alternative to socket prostheses that directly attach to the residual femur via a bone-anchored implant, however their multi-joint loading effect is largely unknown. Our objective was to establish how osseointegrated prostheses influence joint loading during walking. Motion capture data (kinematics, ground reaction forces) were collected from 12 participants at baseline, with socket prostheses, and 12-months after prosthesis osseointegration during overground walking at self-selected speeds. Subject-specific musculoskeletal models were developed at each timepoint relative to osseointegration. Internal joint moments were calculated using inverse dynamics, muscle and joint reaction forces (JRFs) were estimated with static optimization. Changes in internal joint moments, JRFs, and joint loading-symmetry were compared using statistical parametric mapping (p≤ 0.05) before and after osseointegration. Amputated limb hip flexion moments and anterior JRFs decreased during terminal stance (p = 0.002, <0.001; respectively), while amputated limb hip abduction moments increased during mid-stance (p < 0.001), amputated hip rotation moment changed from internal to external throughout early stance (p < 0.001). Intact limb hip extension and knee flexion moments (p = 0.028, 0.032; respectively), superior and resultant knee JRFs (p = 0.046, 0.049; respectively) decreased during the loading response following prosthesis osseointegration. These results may indicate that the direct loading transmission of these novel prostheses create a more typical mechanical environment in bilateral joints, which is comparable with loading observed in able-bodied individuals and could decrease the risk of development or progression of osteoarthritis.

The influence of induced gait asymmetry on joint reaction forces

Chronic injury- or disease-induced joint impairments result in asymmetric gait deviations that may precipitate changes in joint loading associated with pain and osteoarthritis. Understanding the impact of gait deviations on joint reaction forces (JRFs) is challenging because of concurrent neurological and/or anatomical changes and because measuring JRFs requires medically invasive instrumented implants. Instead, we investigated the impact of joint motion limitations and induced asymmetry on JRFs by simulating data recorded as 8 unimpaired participants walked with bracing to unilaterally and bilaterally restrict ankle, knee, and simultaneous ankle + knee motion. Personalized models, calculated kinematics, and ground reaction forces (GRFs) were input into a computed muscle control tool to determine lower limb JRFs and simulated muscle activations guided by electromyography-driven timing constraints. Unilateral knee restriction increased GRF peak and loading rate ipsilaterally but peak values decreased contralaterally when compared to walking without joint restriction. GRF peak and loading rate increased with bilateral restriction compared to the contralateral limb of unilaterally restricted conditions. Despite changes in GRFs, JRFs were relatively unchanged due to reduced muscle forces during loading response. Thus, while joint restriction results in increased limb loading, reductions in muscle forces counteract changes in limb loading such that JRFs were relatively unchanged.

Can a Musculoskeletal Model Adapted to Knee Implant Geometry Improve Prediction of 3D Contact Forces and Moments?

Tibiofemoral contact loads are crucial parameters in the onset and progression of osteoarthrosis. While contact loads are frequently estimated from musculoskeletal models, their customization is often limited to scaling musculoskeletal geometry or adapting muscle lines. Moreover, studies have usually focused on superior-inferior contact force without investigating three-dimensional contact loads. Using experimental data from six patients with instrumented total knee arthroplasty (TKA), this study customized a lower limb musculoskeletal model to consider the positioning and the geometry of the implant at knee level. Static optimization was performed to estimate tibiofemoral contact forces and contact moments as well as musculotendinous forces. Predictions from both a generic and a customized model were compared to the instrumented implant measurements. Both models accurately predict superior-inferior (SI) force and abduction-adduction (AA) moment. Notably, the customization improves prediction of medial-lateral (ML) force and flexion-extension (FE) moments. However, there is subject-dependent variability in the prediction of anterior-posterior (AP) force. The customized models presented here predict loads on all joint axes and in most cases improve prediction. Unexpectedly, this improvement was more limited for patients with more rotated implants, suggesting a need for further model adaptations such as muscle wrapping or redefinition of hip and ankle joint centers and axes.

Video-driven simulation of lower limb mechanical loading during aquatic exercises

Understanding the mechanical demands of an exercise on the musculoskeletal system is crucial to prescribe effective training or therapeutic interventions. Yet, that knowledge is currently limited in water, mostly because of the difficulty in evaluating external resistance. Here I reconcile recent advances in 3D markerless pose and mesh estimation, biomechanical simulations, and hydrodynamic modeling, to predict lower limb mechanical loading during aquatic exercises. Simulations are driven exclusively from a single video. Fluid forces were estimated within 12.5±4.1% of the peak forces determined through computational fluid dynamics analyses, at a speed three orders of magnitude greater. In silico hip and knee resultant joint forces agreed reasonably well with in vivo instrumented implant recordings (R²=0.74) downloaded from the OrthoLoad database, both in magnitude (RMSE =251±125 N) and direction (cosine similarity = 0.92±0.09). Hip flexors, glutes, adductors, and hamstrings were the main contributors to hip joint compressive forces (40.4±12.7%, 25.6±9.7%, 14.2±4.8%, 13.0±8.2%, respectively), while knee compressive forces were mostly produced by the gastrocnemius (39.1±15.9%) and vasti (29.4±13.7%). Unlike dry-land locomotion, non-hip- and non-knee-spanning muscles provided little to no offloading effect via dynamic coupling. This noninvasive method has the potential to standardize the reporting of exercise intensity, inform the design of rehabilitation protocols and improve their reproducibility.

Differences in hip musculoskeletal loads between limbs during daily activities in patients with 3D-printed hemipelvic reconstructions following tumor surgery

BACKGROUND

Anatomical custom-made prostheses, thanks to computer-aided design and 3D-printing technology, help improve osseointegration and reduce mechanical complications in bone reconstructions following bone tumors. A recent quantitative analysis of long-term recovery in patients with 3D-printed reconstructions following pelvic tumor surgery showed asymmetries in ground reaction forces between limbs during different motor activities, while standing very good motor performance and quality of life.

RESEARCH QUESTION

We analyzed hip contact forces and muscle forces in that cohort of six patients with an innovative custom-made reconstruction of the hemipelvis, and we tested the hypothesis that asymmetries in ground reaction forces would result in more marked differences in musculoskeletal forces.

METHODS

State-of-the-art musculoskeletal modeling in an optimization-based inverse-dynamics workflow was used to calculate hip contact forces and muscle forces during five motor activities, and the differences between limbs were statistically evaluated across the motor activity cycles and on the force peaks.

RESULTS

The musculoskeletal loads were found to be not symmetric, as hip loads were generally higher in the contralateral limb. We found significant differences in considerable portions of the motor activities cycles except squat, load symmetry indices indicating a load increase (median up to 25%) on the contralateral limb, especially during stair descent and chair rise/sit, and significantly higher values in the contralateral limb at force peaks.

SIGNIFICANCE

We confirmed the hypothesis that residual asymmetries found in ground reaction forces were amplified when hip musculoskeletal loads were investigated, reflecting a shift of the loads toward the intact limb. Despite the general trend of higher loads found in the contralateral hip, this cannot be considered a risk of overloading, as both hips supported loads in a physiological range or lower, indicating a likely optimal recovery.

Imaging-based musculoskeletal models alter muscle and joint contact forces but do not improve the agreement with experimentally measured electromyography signals in children with cerebral palsy

BACKGROUND

Musculoskeletal simulations are used to estimate muscle-tendon and joint contact forces (JCF). Personalizing the model's femoral geometry has been shown to improve the accuracy of JCF calculations. It is, however, unknown if the personalized geometry improves the agreement between estimated muscle activations and experimentally measured electromyography (EMG) signals.

RESEARCH QUESTION

Does personalizing the musculoskeletal geometry improve the agreement between estimated muscle activations and EMG signals in terms of timing?

METHODS

We retrospectively analysed data from Bosmans et al. [5], which included three-dimensional motion capture data, EMG signals of eight lower limb muscles on each leg, and magnetic resonance imaging (MRI) data from seven children with cerebral palsy. For each patient we created a generic-scaled model and MRI-based model, which accounted for the subject-specific musculoskeletal geometry. We calculated muscle activations, muscle-tendon forces and JCF. Muscle activations were compared to the EMG signals using coefficient of determination and cosines similarity.

RESULTS

MRI-based models altered the magnitude of muscle activations and had a large impact on JCF but did not change the muscle activations profiles and therefore did not improve the agreement with EMG signals.

SIGNIFICANCE

MRI-based models do not alter the shape of muscle activations. Hence, if detailed muscle activations are a desired output of the simulations, EMG-informed modeling approaches should be used for musculoskeletal simulation in children with cerebral palsy. Furthermore, our study highlighted that altered JCF does not necessarily mean accurate muscle activations. To improve patient-specific simulations, future work should focus on developing methods to estimate cost functions representative for the neural control of children with cerebral palsy.

The functional role of hip muscles during gait in patients with increased femoral anteversion

BACKGROUND

Femoral anteversion affects the lever arm and moment-generating capacity of the hip abductors, while an increased hip internal rotation during walking was proposed to be a compensatory mechanism to restore the abductive lever arm. Children with isolated increased femoral anteversion, however, do not always present a deficit in the net hip abduction moment during gait, suggesting that a more comprehensive understanding of the effect of morphology and motion on muscle forces and moments is needed to aid clinical decision making.

RESEARCH QUESTION

Are muscle contributions to hip joint moments and muscle forces altered in patients with increased femoral anteversion and internally rotated gait pattern compared to a control group of typically developing children? And how would the functional role of the muscle be altered if the patients walked straight?

METHODS

This follow-up study compared patients with increased femoral anteversion (n = 42, 12.8 ± 1.9 years, femoral anteversion: 39.6 ± 6.9°) to controls (n = 9, 12.0 ± 3.0 years, femoral anteversion: 18.7 ± 4.1°). Muscle forces and moment contributions were calculated using personalized musculoskeletal models. Additionally, a hypothetical scenario, in which the gait of the controls was modelled with an anteverted femoral morphology, was used to understand what would happen if the patients walked straight.

RESULTS

Gluteus medius abductive contribution was lower in patients compared to controls, despite a comparable net abduction moment around the hip. Patients presented lower muscle forces. However, if modelled to walk straight, they would require higher forces as well as a larger co-contraction of both hip internal and external rotators in the transversal plane.

SIGNIFICANCE

This study suggests that patients with increased femoral anteversion walking with an internally rotated gait pattern present lower muscle forces, but when modelled to walk straight muscle forces increase. The current results provide important information to better understand this condition and improve treatment recommendations in these patients.

Ankle joint contact force profiles differ between those with and without chronic ankle instability during walking

BACKGROUND

Individuals with chronic ankle instability (CAI) exhibit aberrant gait biomechanics relative to uninjured controls. Altered gait biomechanics likely contribute aberrant joint loading and subsequent early onset ankle joint degeneration. Joint (i.e. cartilage) loading cannot be directly measured without invasive procedures but can be estimated via joint contact forces (JCF) generated from musculoskeletal modeling. However, no investigation has quantified JCF in those with CAI during walking despite the link between ligamentous injury and ankle post-traumatic ankle osteoarthritis.

RESEARCH QUESTION

Do patients with CAI exhibit altered ankle compressive and shear JCF profiles during the stance phase of walking compared to those without CAI?

METHODS

Ten individuals with CAI and 10 individuals without a history of ankle sprain completed a gait assessment at their self-selected speed on an instrumented treadmill. Musculoskeletal modeling was applied to estimate ankle JCF variables within a generic model. Variables included the peak, impulse, and loading rates for compressive, anteroposterior shear, and mediolateral shear JCF.

RESULTS

Those with CAI had significantly different JCF forces, relative to uninjured controls, in all directions. More specifically, lower compressive peak and impulse values were noted while higher anteroposterior shearing forces (1 st peak, impulse, loading late) were observed in those with CAI. Those with CAI also demonstrated higher mediolateral shearing forces (1 st peak and impulse).

SIGNIFICANCE

Our finding suggests that those with CAI exhibit different ankle joint loading patterns than uninjured controls. Directionality of the identified differences depends on the axis of movement.

Association between sagittal alignment and loads at the adjacent segment in the fused spine: a combined clinical and musculoskeletal modeling study of 205 patients with adult spinal deformity

PURPOSE

Sagittal malalignment is a risk factor for mechanical complications after surgery for adult spinal deformity (ASD). Spinal loads, modulated by sagittal alignment, may explain this relationship. The aims of this study were to investigate the relationships between: (1) postoperative changes in loads at the proximal segment and realignment, and (2) absolute postoperative loads and postoperative alignment measures.

METHODS

A previously validated musculoskeletal model of the whole spine was applied to study a clinical sample of 205 patients with ASD. Based on clinical and radiographic data, pre-and postoperative patient-specific alignments were simulated to predict loads at the proximal segment adjacent to the spinal fusion.

RESULTS

Weak-to-moderate associations were found between pre-to-postop changes in lumbar lordosis, LL (r =  - 0.23, r =  - 0.43; p < 0.001), global tilt, GT (r = 0.26, r = 0.38; p < 0.001) and the Global Alignment and Proportion score, GAP (r = 0.26, r = 0.37; p < 0.001), and changes in compressive and shear forces at the proximal segment. GAP score parameters, thoracic kyphosis measurements and the slope of upper instrumented vertebra were associated with changes in shear. In patients with T10-pelvis fusion, moderate-to-strong associations were found between postoperative sagittal alignment measures and compressive and shear loads, with GT showing the strongest correlations (r = 0.75, r = 0.73, p < 0.001).

CONCLUSIONS

Spinal loads were estimated for patient-specific full spinal alignment profiles in a large cohort of patients with ASD pre-and postoperatively. Loads on the proximal segments were greater in association with sagittal malalignment and malorientation of proximal vertebra. Future work should explore whether they provide a causative mechanism explaining the associated risk of proximal junction complications.

Estimating lumbar spine loading when using back-support exoskeletons in lifting tasks

Low-back pain (LBP) continues as the leading cause of work-related musculoskeletal disorders, and the high LBP burden is attributed largely to physical risk factors prevalent in manual material handling tasks. Industrial back-support exoskeletons (BSEs) are a promising ergonomic intervention to help control/prevent exposures to such risk factors. While earlier research has demonstrated beneficial effects of BSEs in terms of reductions in superficial back muscle activity, limited evidence is available regarding the impacts of these devices on spine loads. We evaluated the effects of two passive BSEs (BackX™ AC and Laevo™ V2.5) on lumbosacral compression and shear forces during repetitive lifting using an optimization-based model. Eighteen participants (gender-balanced) completed four minutes of repetitive lifting in nine different conditions, involving symmetric and asymmetric postures when using the BSEs (along with no BSE as a control condition). Using both BSEs reduced estimated peak compression and anteroposterior shear forces (by ∼8-15%). Such reductions, however, were task-specific and depended on the BSE design. Laevo™ use reduced mediolateral shear forces during asymmetric lifting (by ∼35%). We also found that reductions in composite measures of trunk muscle activity may not correspond well with changes in spine forces when using a BSE. These results can help guide the proper selection and application of BSEs during repetitive lifting tasks. Future work is recommended to explore the viability of different biomechanical models to assess changes in spine mechanical loads when using BSEs and whether reasonable estimates would be obtained using such models.

Contribution of lower extremity muscles to center of mass acceleration during walking: Effect of body weight

Overweight or obesity is known to be associated with altered activations of lower extremity muscles. Such changes in muscular function may lead to the development of mobility impairments or joint diseases. However, little is known about how individual lower extremity muscles contribute to the whole-body center of mass (COM) control during walking and the effect of body weight. This study examined the contribution of individual lower extremity muscle force to the COM accelerations during walking in overweight and non-overweight individuals. Musculoskeletal simulations were performed for the stance phase of walking with data collected from 11 overweight and 13 non-overweight adults to estimate lower extremity muscle forces and their contributions to the COM acceleration. Mean time-series data from each parameter were compared between body size groups using Statistical Parametric Mapping. Compared to the non-overweight group, the overweight group revealed a greater gastrocnemius contribution to the mediolateral (p = 0.006) and vertical (p < 0.001) COM accelerations during mid-stance, and had a lower vastus contribution to the anteroposterior COM acceleration (p < 0.001) during pre-swing. Increased contributions from the large posterior calf muscles to the mediolateral COM acceleration may be related to efforts to alleviate COM sway in overweight individuals.

On the estimation of hip joint loads through musculoskeletal modeling

Noninvasive estimation of joint loads is still an open challenge in biomechanics. Although musculoskeletal modeling represents a solid resource, multiple improvements are still necessary to obtain accurate predictions of joint loads and to translate such potential into practical utility. The present study, focused on the hip joint, is aimed at reviewing the state-of-the-art literature on the estimation of hip joint reaction forces through musculoskeletal modeling. Our literature inspection, based on well-defined selection criteria, returned seventeen works, which were compared in terms of methods and results. Deviations between predicted and in vivo measured hip joint loads, taken from the OrthoLoad database, were assessed through quantitative deviation indices. Despite the numerous modeling and computational improvements made over the last two decades, predicted hip joint loads still deviate from their experimental counterparts and typically overestimate them. Several critical aspects have emerged that affect muscle force estimation, hence joint loads. Among them, the physical fidelity of the musculoskeletal model, with its parameters and geometry, plays a crucial role. Also, predicted joint loads are markedly affected by the selected muscle recruitment strategy, which reflects the underlying motor control policy. Practical guidelines for researchers interested in noninvasive estimation of hip joint loads are also provided.

A novel approach to quantify the assistive torque profiles generated by passive back-support exoskeletons

Industrial exoskeletons are a promising ergonomic intervention to reduce the risk of work-related musculoskeletal disorders by providing external physical support to workers. Passive exoskeletons, having no power supplies, are of particular interest given their predominance in the commercial market. Understanding the mechanical behavior of the torque generation mechanisms embedded in passive exoskeletons is, however, essential to determine the efficacy of these devices in reducing physical loads (e.g., in manual material handling tasks). We introduce a novel approach using a computerized dynamometer to quantify the assistive torque profiles of two passive back-support exoskeletons (BSEs) at different support settings and in both static and dynamic conditions. The feasibility of this approach was examined using both human subjects and a mannequin. Clear differences in assistive torque magnitudes were evident between the two BSEs, and both devices generated more assistive torques during trunk/hip flexion than extension. Assistive torques obtained from human subjects were often within similar ranges as those from the mannequin, though values were more comparable over a narrow range of flexion/extension angles due to practical limitations with the dynamometer and human subjects. Characterizing exoskeleton assistive torque profiles can help in better understanding how to select a torque profile for given task requirements and user anthropometry, and aid in predicting the potential impacts of exoskeleton use by incorporating measured torque profiles in a musculoskeletal modeling system. Future work is recommended to assess this approach for other occupational exoskeletons.

Electromyography-driven model-based estimation of ankle torque and stiffness during dynamic joint rotations in perturbed and unperturbed conditions

The simultaneous modulation of joint torque and stiffness enables humans to perform large repertoires of movements, while versatilely adapting to external mechanical demands. Multi-muscle force control is key for joint torque and stiffness modulation. However, the inability to directly measure muscle force in the intact moving human prevents understanding how muscle force causally links to joint torque and stiffness. Joint stiffness is predominantly estimated via joint perturbation-based experiments in combination with system identification techniques. However, these techniques provide joint-level stiffness estimations with no causal link to the underlying muscle forces. Moreover, the need for joint perturbations limits the generalizability and applicability to study natural movements. Here, we present an electromyography (EMG)-driven musculoskeletal modeling framework that can be calibrated to match reference joint torque and stiffness profiles simultaneously via a multi-term objective function. EMG-driven models calibrated on <2 s of reference torque and stiffness data could blindly estimate reference profiles across 100 s of data not used for calibration. Model calibrations using an objective function comprising torque and stiffness terms always provided less feasible solutions than an objective function comprising solely a torque term, thereby reducing the space of feasible muscle-tendon parameters. Results also showed the proposed framework's ability to estimate joint stiffness in unperturbed conditions, while capturing differences against stiffness profiles derived during perturbed conditions. The proposed framework may provide new ways for studying causal relationships between muscle force and joint torque and stiffness during movements in interaction with the environment, with broad implications across biomechanics, rehabilitation and robotics.

Automated estimation of ankle muscle EMG envelopes and resulting plantar-dorsi flexion torque from 64 garment-embedded electrodes uniformly distributed around the human leg

The design of personalized movement training and rehabilitation pipelines relies on the ability of assessing the activation of individual muscles concurrently with the resulting joint torques exerted during functional movements. Despite advances in motion capturing, force sensing and bio-electrical recording technologies, the estimation of muscle activation and resulting force still relies on lengthy experimental and computational procedures that are not clinically viable. This work proposes a wearable technology for the rapid, yet quantitative, assessment of musculoskeletal function. It comprises of (1) a soft leg garment sensorized with 64 uniformly distributed electromyography (EMG) electrodes, (2) an algorithm that automatically groups electrodes into seven muscle-specific clusters, and (3) a EMG-driven musculoskeletal model that estimates the resulting force and torque produced about the ankle joint sagittal plane. Our results show the ability of the proposed technology to automatically select a sub-set of muscle-specific electrodes that enabled accurate estimation of muscle excitations and resulting joint torques across a large range of biomechanically diverse movements, underlying different excitation patterns, in a group of eight healthy individuals. This may substantially decrease time needed for localization of muscle sites and electrode placement procedures, thereby facilitating applicability of EMG-driven modelling pipelines in standard clinical protocols.

Multi-level personalization of neuromusculoskeletal models to estimate physiologically plausible knee joint contact forces in children

Neuromusculoskeletal models are a powerful tool to investigate the internal biomechanics of an individual. However, commonly used neuromusculoskeletal models are generated via linear scaling of generic templates derived from elderly adult anatomies and poorly represent a child, let alone children with a neuromuscular disorder whose musculoskeletal structures and muscle activation patterns are profoundly altered. Model personalization can capture abnormalities and appropriately describe the underlying (altered) biomechanics of an individual. In this work, we explored the effect of six different levels of neuromusculoskeletal model personalization on estimates of muscle forces and knee joint contact forces to tease out the importance of model personalization for normal and abnormal musculoskeletal structures and muscle activation patterns. For six children, with and without cerebral palsy, generic scaled models were developed and progressively personalized by (1) tuning and calibrating musculotendon units' parameters, (2) implementing an electromyogram-assisted approach to synthesize muscle activations, and (3) replacing generic anatomies with image-based bony geometries, and physiologically and physically plausible muscle kinematics. Biomechanical simulations of gait were performed in the OpenSim and CEINMS software on ten overground walking trials per participant. A mixed-ANOVA test, with Bonferroni corrections, was conducted to compare all models' estimates. The model with the highest level of personalization produced the most physiologically plausible estimates. Model personalization is crucial to produce physiologically plausible estimates of internal biomechanical quantities. In particular, personalization of musculoskeletal anatomy and muscle activation patterns had the largest effect overall. Increased research efforts are needed to ease the creation of personalized neuromusculoskeletal models.

Biomechanical compensations during a stand-to-sit maneuver using transfemoral osseointegrated prostheses: A case series

BACKGROUND

Patients with transfemoral amputation and socket prostheses are at a heightened risk of developing musculoskeletal overuse injuries, commonly due to altered joint biomechanics. Osseointegrated prostheses, which involve direct anchorage of the prosthesis to the residual limb through a bone anchored prosthesis, are a novel alternative to sockets yet their biomechanical effect is largely unknown.

METHODS

Four patients scheduled to undergo unilateral transfemoral prosthesis osseointegration completed two data collections (baseline with socket prosthesis and 12-months after prosthesis osseointegration) in which whole-body kinematics and ground reaction forces were collected during stand-to-sit tasks. Trunk, pelvis, and hip kinematics, and the surrounding muscle forces, were calculated using subject-specific musculoskeletal models developed in OpenSim. Peak joint angles and muscle forces were compared between timepoints using Cohen's d effect sizes.

FINDINGS

Compared to baseline with socket prostheses, patients with osseointegrated prostheses demonstrated reduced lateral trunk bending (d = 1.46), pelvic obliquity (d = 1.09), and rotation (d = 1.77) toward the amputated limb during the stand to sit task. This was accompanied by increased amputated limb hip flexor, abductor, and rotator muscle forces (d> > 0.8).

INTERPRETATION

Improved lumbopelvic movement patterns and stabilizing muscle forces when using an osseointegrated prosthesis indicate that this novel prosthesis type likely reduces the risk of the development and/or progression of overuse injuries, such as low back pain and osteoarthritis. We attribute the increased muscle hip muscle forces to the increased load transmission between the osseointegrated prosthesis and residual limb, which allows a greater eccentric ability of the amputated limb to control lowering during the stand-to-sit task.

Sensitivity analysis guided improvement of an electromyogram-driven lumped parameter musculoskeletal hand model

EMG-driven neuromusculoskeletal models have been used to study many impairments and hold great potential to facilitate human-machine interactions for rehabilitation. A challenge to successful clinical application is the need to optimize the model parameters to produce accurate kinematic predictions. In order to identify the key parameters, we used Monte-Carlo simulations to evaluate the sensitivities of wrist and metacarpophalangeal (MCP) flexion/extension prediction accuracies for an EMG-driven, lumped-parameter musculoskeletal model. Four muscles were modeled with 22 total optimizable parameters. Model predictions from EMG were compared with measured joint angles from 11 able-bodied subjects. While sensitivities varied by muscle, we determined muscle moment arms, maximum isometric force, and tendon slack length were highly influential, while passive stiffness and optimal fiber length were less influential. Removing the two least influential parameters from each muscle reduced the optimization search space from 22 to 14 parameters without significantly impacting prediction correlation (wrist: 0.90 ± 0.05 vs 0.90 ± 0.05, p = 0.96; MCP: 0.74 ± 0.20 vs 0.70 ± 0.23, p = 0.51) and normalized root mean square error (wrist: 0.18 ± 0.03 vs 0.19 ± 0.03, p = 0.16; MCP: 0.18 ± 0.06 vs 0.19 ± 0.06, p = 0.60). Additionally, we showed that wrist kinematic predictions were insensitive to parameters of the modeled MCP muscles. This allowed us to develop a novel optimization strategy that more reliably identified the optimal set of parameters for each subject (27.3 ± 19.5%) compared to the baseline optimization strategy (6.4 ± 8.1%; p = 0.004). This study demonstrated how sensitivity analyses can be used to guide model refinement and inform novel and improved optimization strategies, facilitating implementation of musculoskeletal models for clinical applications.

Ground reaction force and joint moment estimation during gait using an Azure Kinect-driven musculoskeletal modeling approach

BACKGROUND

Gait analysis is burdened by time and equipment costs, interpretation, and accessibility of three-dimensional motion analysis systems. Evidence suggests growing adoption of gait testing in the shift toward evidence-based medicine. Further developments addressing these barriers will aid its efficacy in clinical practice. Previous research aiming to develop gait analysis systems for kinetics estimation using the Kinect V2 have provided promising results yet modified approaches using the latest hardware may further aid kinetics estimation accuracy RESEARCH QUESTION: Can a single Azure Kinect sensor combined with a musculoskeletal modeling approach provide kinetics estimations during gait similar to those obtained from marker-based systems with embedded force platforms?

METHODS

Ten subjects were recruited to perform three walking trials at their normal speed. Trials were recorded using an eight-camera optoelectronic system with two embedded force plates and a single Azure Kinect sensor. Marker and depth data were both used to drive a musculoskeletal model using the AnyBody Modeling System. Predicted kinetics from the Azure Kinect-driven model, including ground reaction force (GRF) and joint moments, were compared to measured values using root meansquared error (RMSE), normalized RMSE, Pearson correlation, concordance correlation, and statistical parametric mapping RESULTS: High to very high correlations were observed for anteroposterior GRF (ρ = 0.889), vertical GRF (ρ = 0.940), and sagittal hip (ρ = 0.805) and ankle (ρ = 0.876) moments. RMSEs were 1.2 ± 2.2 (%BW), 3.2 ± 5.7 (%BW), 0.7 ± 0.1.3 (%BWH), and 0.6 ± 1.0 (%BWH) SIGNIFICANCE: The proposed approach using the Azure Kinect provided higher accuracy compared to previous studies using the Kinect V2 potentially due to improved foot tracking by the Azure Kinect. Future studies should seek to optimize ground contact parameters and focus on regions of error between predicted and measured kinetics highlighted currently for further improvements in kinetic estimations.

Design of a control framework for lower limb exoskeleton rehabilitation robot based on predictive assessment

BACKGROUND

Patients suffering from lower limb dyskinesia, especially in early stages of rehabilitation, have weak residual muscle strength in affected limb and require passive training by the lower limb rehabilitation robot. Anatomy indicates that the biceps femoris short head muscle has a strong influence on knee motion at the swing phase of walking. We sought to explore how it would influence on gait cycle in optimization framework. However, the training trajectory of conventional rehabilitation robots performing passive training usually follows gait planning based on general human gait data, which cannot simultaneously ensure both effective rehabilitation of affected limbs with varying severity pathological gait and comfort of the wearer within a safe motion trajectory.

METHODS

To elucidate the effects of weakness and contracture, we systematically introduced isolated defects into the musculoskeletal model and generated walking simulations to predict the gait adaptation due to these defects. An impedance control model of the rehabilitation robot is developed. Knee joint parameters optimized by predictive forward dynamics simulation are adopted as the expected values for the robot controller to achieve customized adjustment of the robot motion trajectory.

FINDINGS

Severe muscle contracture leads to severe knee flexion; severe muscle weakness induces a significant posterior tilt of the upper trunk, which hinders walking speed.

INTERPRETATION

Our simulation results attempt to reveal pathological gait features, which may help to reproduce the simulation of pathological gait. Furthermore, the robot simulation results show that the robot system achieves a speedy tracking by setting a larger stiffness value. The model also allows the implementation of different levels of damping or elasticity effects.

TRIAL REGISTRATION

The method proposed in this paper is an initial basic study that did not reach clinical trials and therefore retains retrospectively registered.

Skin marker-based subject-specific spinal alignment modeling: A feasibility study

Musculoskeletal models have the potential to improve diagnosis and optimize clinical treatment by predicting accurate outcomes on an individual basis. However, the subject-specific modeling of spinal alignment is often strongly simplified or is based on radiographic assessments, exposing subjects to unnecessary radiation. We therefore developed and introduced a novel skin marker-based approach for modeling subject-specific spinal alignment and evaluated its feasibility by comparing the predicted L1/L2 spinal loads during various functional activities with the loads predicted by the generically scaled models as well as with in vivo measured data obtained from the OrthoLoad database. Spinal loading simulations resulted in considerably higher compressive forces for both scaling approaches over all simulated activities, and AP shear forces that were closer or similar to the in vivo data for the subject-specific approach during upright standing activities and for the generic approach during activities that involved large flexions. These results underline the feasibility of the proposed method and associated workflow for inter- and intra-subject investigations using musculoskeletal simulations. When implemented into standard model scaling workflows, it is expected to improve the accuracy of muscle activity and joint loading simulations, which is crucial for investigations of treatment effects or pathology-dependent deviations.

Computational model predicts risk of spinal screw loosening in patients

PURPOSE

Pedicle screw loosening is a frequent complication in lumbar spine fixation, most commonly among patients with poor bone quality. Determining patients at high risk for insufficient implant stability would allow clinicians to adapt the treatment accordingly. The aim of this study was to develop a computational model for quantitative and reliable assessment of the risk of screw loosening.

METHODS

A cohort of patient vertebrae with diagnosed screw loosening was juxtaposed to a control group with stable fusion. Imaging data from the two cohorts were used to generate patient-specific biomechanical models of lumbar instrumented vertebral bodies. Single-level finite element models loading the screw in axial or caudo-cranial direction were generated. Further, multi-level models incorporating individualized joint loading were created.

RESULTS

The simulation results indicate that there is no association between screw pull-out strength and the manifestation of implant loosening (p = 0.8). For patient models incorporating multiple instrumented vertebrae, CT-values and stress in the bone were significantly different between loose screws and non-loose screws (p = 0.017 and p = 0.029, for CT-values and stress, respectively). However, very high distinction (p = 0.001) and predictability (R²_Pseudo = 0.358, AUC = 0.85) were achieved when considering the relationship between local bone strength and the predicted stress (loading factor). Screws surrounded by bone with a loading factor higher than 25% were likely to be loose, while the chances of screw loosening were close to 0 with a loading factor below 15%.

CONCLUSION

The use of a biomechanics-based score for risk assessment of implant fixation failure might represent a paradigm shift in addressing screw loosening after spondylodesis surgery.

Effect of Assistance Using a Bilateral Robotic Knee Exoskeleton on Tibiofemoral Force Using a Neuromuscular Model

Tibiofemoral compression forces present during locomotion can result in high stress and risk damage to the knee. Powered assistance using a knee exoskeleton may reduce the knee load by reducing the work required by the muscles. However, the exact effect of assistance on the tibiofemoral force is unknown. The goal of this study was to investigate the effect of knee extension assistance during the early stance phase on the tibiofemoral force. Nine able-bodied adults walked on an inclined treadmill with a bilateral knee exoskeleton with assistance and with no assistance. Using an EMG-informed neuromusculoskeletal model, muscle forces were estimated, then utilized to estimate the tibiofemoral contact force. Results showed a 28% reduction in the knee moment, which resulted in approximately a 15% decrease in knee extensor muscle activation and a 20% reduction in subsequent muscle force, leading to a significant 10% reduction in peak and 9% reduction in average tibiofemoral contact force during the early stance phase (p < 0.05). The results indicate the tibiofemoral force is highly dependent on the knee kinetics and quadricep muscle activation due to their influence on knee extensor muscle forces, the primary contributor to the knee load.

Visual programming for accessible interactive musculoskeletal models

Objective

Musculoskeletal modeling and simulation are powerful research and education tools in engineering, neuroscience, and rehabilitation. Interactive musculoskeletal models (IMMs) can be controlled by muscle activity recorded with electromyography (EMG). IMMs are typically coded using textual programming languages that present barriers to understanding for non-experts. The goal of this project was to use a visual programming language (Simulink) to create and test an IMM that is accessible to non-specialists for research and educational purposes.

Results

The developed IMM allows users to practice a goal-directed task with different control modes (keyboard, mouse, and EMG) and actuator types (muscle model, force generator, and torque generator). Example data were collected using both keyboard and EMG control. One male participant in his early 40’s performed a goal-directed task for four sequential trials using each control mode. For EMG control, the participant used a low-cost EMG system with consumer-grade EMG sensors and an Arduino microprocessor. The participant successfully performed the task with both control modes, but the inability to grade muscle model excitation and co-activate antagonist muscles limited performance with keyboard control. The IMM developed for this project serves as a foundation that can be further tailored to specific research and education needs.

Femoral version deformities alter joint reaction forces in dysplastic hips during gait

Developmental dysplasia of the hip (DDH) causes hip instability and early-onset osteoarthritis. The focus on pathomechanics in DDH has centered on the shallow acetabulum, however there is growing awareness of the role of femoral deformities in joint damage. The objective of this study was to determine the influence of femoral version (FV) on the muscle and joint reaction forces (JRFs) of dysplastic hips during gait. Magnetic resonance images, in-vivo gait data, and musculoskeletal models were used to calculate JRFs and simulate changes due to varying FV deformities. Rotation about the long axis of the femur was added in the musculoskeletal models to simulate FV values from -5° (relative retroversion) to + 35° (increased anteversion). In our simulations, FV deformities caused the largest changes to the anteroposterior and resultant JRFs. From a normal FV of 15°, a 15° increase in femoral anteversion caused JRFs to be less posterior in early stance (Δ = 0.43 ± 0.22 xbodyweight) and more anterior in late stance (Δ = 0.60 ± 14 xbodyweight). Relative retroversion caused anteroposterior changes that were similar to anteversion in early stance but opposite in late stance. Resultant JRFs experienced the largest changes during late stance where anteversion raised the peak by 0.48 ± 0.15 xbodyweight and relative retroversion lowered the peak by 0.32 ± 0.30 xbodyweight. Increasing anteversion increased hip flexor and abductor muscle forces, which caused the changes in JRFs. Identifying how FV deformities influence hip joint loading can elucidate their role in the mechanisms of hip degeneration in patients with DDH.

Changes in segment coordination variability and the impacts of the lower limb across running mileages in half marathons: Implications for running injuries

BACKGROUND

Segment coordination variability (CV) is a movement pattern associated with running-related injuries. It can also be adversely affected by a prolonged run. However, research on this topic is currently limited. The purpose of this study was to investigate the effects of a prolonged run on segment CV and vertical loading rates during a treadmill half marathon.

METHODS

Fifteen healthy runners ran a half marathon on an instrumental treadmill in a biomechanical laboratory. Synchronized kinematic and kinetic data were collected every 2 km (from 2 km until 20 km), and the data were processed by musculoskeletal modeling. Segment CVs were computed from the angle-angle plots of selected pelvis-thigh, thigh-shank, and shank-rearfoot couplings using a modified vector coding technique. The loading rate of vertical ground reaction force was also calculated. A one-way MANOVA with repeated measures was performed on each of the outcome variables to examine the main effect of running mileage.

RESULTS

Significant effects of running mileage were found on segment CVs (p ≤ 0.010) but not on loading rate (p = 0.881). Notably, during the early stance phase, the CV of pelvis frontal vs. thigh frontal was significantly increased at 20 km compared with the CV at 8 km (g = 0.59, p = 0.022). The CV of shank transverse vs. rearfoot frontal decreased from 2 km to 8 km (g = 0.30, p = 0.020) but then significantly increased at both 18 km (g = 0.05, p < 0.001) and 20 km (g = 0.36, p < 0.001).

CONCLUSION

At the early stance, runners maintained stable CVs on the sagittal plane, which could explain the unchanged loading rate throughout the half marathon. However, increased CVs on the frontal/transverse plane may be an early sign of fatigue and indicative of possible injury risk. Further studies are necessary for conclusive statements in this regard.

Plasticity of the gastrocnemius elastic system in response to decreased work and power demand during growth

Elastic energy storage and release can enhance performance that would otherwise be limited by the force-velocity constraints of muscle. Although functional influence of a biological spring depends on tuning between components of an elastic system (the muscle, spring-driven mass and lever system), we do not know whether elastic systems systematically adapt to functional demand. To test whether altering work and power generation during maturation alters the morphology of an elastic system, we prevented growing guinea fowl (Numida meleagris) from jumping. We compared the jump performance of our treatment group at maturity with that of controls and measured the morphology of the gastrocnemius elastic system. We found that restricted birds jumped with lower jump power and work, yet there were no significant between-group differences in the components of the elastic system. Further, subject-specific models revealed no difference in energy storage capacity between groups, though energy storage was most sensitive to variations in muscle properties (most significantly operating length and least dependent on tendon stiffness). We conclude that the gastrocnemius elastic system in the guinea fowl displays little to no plastic response to decreased demand during growth and hypothesize that neural plasticity may explain performance variation.

Plasticity of the gastrocnemius elastic system in response to decreased work and power demand during growth

Elastic energy storage and release can enhance performance that would otherwise be limited by the force-velocity constraints of muscle. Although functional influence of a biological spring depends on tuning between components of an elastic system (the muscle, spring-driven mass and lever system), we do not know whether elastic systems systematically adapt to functional demand. To test whether altering work and power generation during maturation alters the morphology of an elastic system, we prevented growing guinea fowl (Numida meleagris) from jumping. We compared the jump performance of our treatment group at maturity with that of controls and measured the morphology of the gastrocnemius elastic system. We found that restricted birds jumped with lower jump power and work, yet there were no significant between-group differences in the components of the elastic system. Further, subject-specific models revealed no difference in energy storage capacity between groups, though energy storage was most sensitive to variations in muscle properties (most significantly operating length and least dependent on tendon stiffness). We conclude that the gastrocnemius elastic system in the guinea fowl displays little to no plastic response to decreased demand during growth and hypothesize that neural plasticity may explain performance variation.

Biomechanical assessment of patellar tendon advancement in patients with cerebral palsy and crouch gait

BACKGROUND

Patellar height is a valuable measure to evaluate the effect of patellar tendon advancement (PTA) on knee function. In the literature, there is no validated procedure to measure the patellar height. In this study we aimed to (1) determine the patella position through musculoskeletal modeling, (2) investigate the effects of two surgical procedures applied for PTA, and (3) assess the effect of PTA in combination with single-event multilevel surgery (SEMLS) on the knee kinematics of patients with cerebral palsy (CP) and crouch gait.

METHOD

Three-dimensional gait and X-ray data of children with CP and crouch gait were retrospectively analyzed if they had received a SEMLS in combination with PTA (PTA group, n = 18) or without PTA (NoPTA group, n = 18). A computational musculoskeletal model was used to quantify patella position, knee extension moment arm, and knee kinematics pre- and postoperatively.

RESULTS

Patellar height significantly decreased in the PTA group (P = 0.004), while there was no difference in the NoPTA group (P > 0.05). The bony procedure for PTA provided a better Insall-Salvati ratio than the soft tissue procedure. The peak knee extension moment arm significantly increased in the PTA group (P = 0.008). In terms of postoperative knee joint kinematics, the PTA group was closer to typically developed children than the NoPTA group.

CONCLUSION

Musculoskeletal modeling was found to be an effective tool for the determination of the patellar height. PTA improved the patella position, knee extension moment arm, and knee kinematics and was an effective procedure for the surgical management of crouch gait in patients with CP.

Acute Vibration Feedback During Gait Reduces Mechanical Ankle Joint Loading in Chronic Ankle Instability Patients

BACKGROUND

Individuals with chronic ankle instability (CAI) exhibit altered vertical ground reaction forces (vGRF), a laterally shifted center of pressure, and an inverted foot position during walking. These neuromechanical alterations are linked with altered ankle joint loading in this population. Vibration-based gait retraining improves center of pressure positioning but effects on neuromechanical variables influencing joint loading remains unknown.

RESEARCH QUESTION

Do patients with CAI exhibit altered vGRF and ankle joint contact forces (JCF) after receiving a single session of vibration-based gait retraining?

METHODS

Ten individuals with CAI underwent a single session of vibration-based gait retraining. Kinematic and kinetic data were collected during walking on an instrumental treadmill with force plates embedded in it. Following a baseline gait assessment without feedback, participants walked at a self-selected speed for 10 minutes while receiving feedback. Data was collected during an early (1 st and 2 nd minute) and late adaptation phase (9 th and 10 th minute) and, compared to baseline values. Impact and propulsive vGRF variables (i.e. peak, time to peak, and loading rate) were obtained. Musculoskeletal modeling was used to calculate ankle JCF variables (peak, impulse, and loading rate) during stance phase.

RESULTS

Propulsive vGRF and ankle JCF outcomes were significantly reduced during the early and late adaptation phases (p ≤ 0.039).

SIGNIFICANCE

These results indicate that vibration-based gait retraining can immediately reduce propulsive vGRF and ankle JCF and may represent a modality that could help restore appropriate ankle joint loading patterns in those with CAI.

Deep reinforcement learning for modeling human locomotion control in neuromechanical simulation

Modeling human motor control and predicting how humans will move in novel environments is a grand scientific challenge. Researchers in the fields of biomechanics and motor control have proposed and evaluated motor control models via neuromechanical simulations, which produce physically correct motions of a musculoskeletal model. Typically, researchers have developed control models that encode physiologically plausible motor control hypotheses and compared the resulting simulation behaviors to measurable human motion data. While such plausible control models were able to simulate and explain many basic locomotion behaviors (e.g. walking, running, and climbing stairs), modeling higher layer controls (e.g. processing environment cues, planning long-term motion strategies, and coordinating basic motor skills to navigate in dynamic and complex environments) remains a challenge. Recent advances in deep reinforcement learning lay a foundation for modeling these complex control processes and controlling a diverse repertoire of human movement; however, reinforcement learning has been rarely applied in neuromechanical simulation to model human control. In this paper, we review the current state of neuromechanical simulations, along with the fundamentals of reinforcement learning, as it applies to human locomotion. We also present a scientific competition and accompanying software platform, which we have organized to accelerate the use of reinforcement learning in neuromechanical simulations. This “Learn to Move” competition was an official competition at the NeurIPS conference from 2017 to 2019 and attracted over 1300 teams from around the world. Top teams adapted state-of-the-art deep reinforcement learning techniques and produced motions, such as quick turning and walk-to-stand transitions, that have not been demonstrated before in neuromechanical simulations without utilizing reference motion data. We close with a discussion of future opportunities at the intersection of human movement simulation and reinforcement learning and our plans to extend the Learn to Move competition to further facilitate interdisciplinary collaboration in modeling human motor control for biomechanics and rehabilitation research