Most children with hemiplegic cerebral palsy (HCP), one of the most prevalent subtypes of cerebral palsy, struggle with grasping and manipulating objects. This impairment may arise from a diminished capacity to properly direct forces created with the finger pad due to aberrant force application. Children with HCP were asked to create maximal force with the index finger pad in the palmar (normal) direction with both the paretic and non-paretic hands. The resulting forces and finger postures were then applied to a computational musculoskeletal model of the hand to estimate the corresponding muscle activation patterns. Subjects tended to create greater shear force relative to normal force with the paretic hand (p < 0.05). The resultant force was directed 33.6°±10.8° away from the instructed palmar direction in the paretic hand, but only 8.0°±7.3° in the non-paretic hand. Additionally, participants created greater palmar force with the non-paretic hand than with the paretic hand (p < 0.05). These differences in force production are likely due to differences in muscle activation pattern, as our computational models showed differences in which muscles are active and their relative activations when recreating the measured force vectors for the two hands (p < 0.01). The models predicted reduced activation in the extrinsic and greater reductions in activation in the intrinsic finger muscles, potentially due to reduced voluntary activation or muscle atrophy. As the large shear forces could lead to objects slipping from grasp, muscle activation patterns may provide an important target for therapeutic treatment in children with HCP.

UNASSIGNED: Calibrated electromyography (EMG)-driven musculoskeletal models can provide great insight into internal quantities (e.g., muscle forces) that are difficult or impossible to measure experimentally. However, the need for EMG data from all involved muscles presents a significant barrier to the widespread application of EMG-driven modeling methods. Synergy extrapolation (SynX) is a computational method that can estimate a single missing EMG signal with reasonable accuracy during the EMG-driven model calibration process, yet its performance in estimating a larger number of missing EMG signals remains unclear.
UNASSIGNED: This study assessed the accuracy with which SynX can use eight measured EMG signals to estimate muscle activations and forces associated with eight missing EMG signals in the same leg during walking while simultaneously performing EMG-driven model calibration. Experimental gait data collected from two individuals post-stroke, including 16 channels of EMG data per leg, were used to calibrate an EMG-driven musculoskeletal model, providing \"gold standard\" muscle activations and forces for evaluation purposes. SynX was then used to predict the muscle activations and forces associated with the eight missing EMG signals while simultaneously calibrating EMG-driven model parameter values. Due to its widespread use, static optimization (SO) was also utilized to estimate the same muscle activations and forces. Estimation accuracy for SynX and SO was evaluated using root mean square errors (RMSE) to quantify amplitude errors and correlation coefficient r values to quantify shape similarity, each calculated with respect to \"gold standard\" muscle activations and forces.
UNASSIGNED: On average, SynX produced significantly more accurate amplitude and shape estimates for unmeasured muscle activations (RMSE 0.08 vs. 0.15,r value 0.55 vs. 0.12) and forces (RMSE 101.3 N vs. 174.4 N,r value 0.53 vs. 0.07) compared to SO. SynX yielded calibrated Hill-type muscle-tendon model parameter values for all muscles and activation dynamics model parameter values for measured muscles that were similar to \"gold standard\" calibrated model parameter values.
UNASSIGNED: These findings suggest that SynX could make it possible to calibrate EMG-driven musculoskeletal models for all important lower-extremity muscles with as few as eight carefully chosen EMG signals and eventually contribute to the design of personalized rehabilitation and surgical interventions for mobility impairments.

Musculoskeletal (MSK) models offer great potential for predicting the muscle forces required to inform more detailed simulations of vertebral endplate loading in adolescent idiopathic scoliosis (AIS). In this work, simulations based on static optimization were compared with in vivo measurements in two AIS patients to determine whether computational approaches alone are sufficient for accurate prediction of paraspinal muscle activity during functional activities. We used biplanar radiographs and marker-based motion capture, ground reaction force, and electromyography (EMG) data from two patients with mild and moderate thoracolumbar AIS (Cobb angles: 21° and 45°, respectively) during standing while holding two weights in front (reference position), walking, running, and object lifting. Using a fully automated approach, 3D spinal shape was extracted from the radiographs. Geometrically personalized OpenSim-based MSK models were created by deforming the spine of pre-scaled full-body models of children/adolescents. Simulations were performed using an experimentally controlled backward approach. Differences between model predictions and EMG measurements of paraspinal muscle activity (both expressed as a percentage of the reference position values) at three different locations around the scoliotic main curve were quantified by root mean square error (RMSE) and cross-correlation (XCorr). Predicted and measured muscle activity correlated best for mild AIS during object lifting (XCorr\'s ≥ 0.97), with relatively low RMSE values. For moderate AIS as well as the walking and running activities, agreement was lower, with XCorr reaching values of 0.51 and comparably high RMSE values. This study demonstrates that static optimization alone seems not appropriate for predicting muscle activity in AIS patients, particularly in those with more than mild deformations as well as when performing upright activities such as walking and running.

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.

Medial knee contact force (MCF) is related to the pathomechanics of medial knee osteoarthritis. However, MCF cannot be directly measured in the native knee, making it difficult for therapeutic gait modifications to target this metric. Static optimization, a musculoskeletal simulation technique, can estimate MCF, but there has been little work validating its ability to detect changes in MCF induced by gait modifications. In this study, we quantified the error in MCF estimates from static optimization compared to measurements from instrumented knee replacements during normal walking and seven different gait modifications. We then identified minimum magnitudes of simulated MCF changes for which static optimization correctly identified the direction of change (i.e., whether MCF increased or decreased) at least 70% of the time. A full-body musculoskeletal model with a multi-compartment knee and static optimization was used to estimate MCF. Simulations were evaluated using experimental data from three subjects with instrumented knee replacements who walked with various gait modifications for a total of 115 steps. Static optimization underpredicted the first peak (mean absolute error = 0.16 bodyweights) and overpredicted the second peak (mean absolute error = 0.31 bodyweights) of MCF. Average root mean square error in MCF over stance phase was 0.32 bodyweights. Static optimization detected the direction of change with at least 70% accuracy for early-stance reductions, late-stance reductions, and early-stance increases in peak MCF of at least 0.10 bodyweights. These results suggest that a static optimization approach accurately detects the direction of change in early-stance medial knee loading, potentially making it a valuable tool for evaluating the biomechanical efficacy of gait modifications for knee osteoarthritis.

Neuromusculoskeletal models often require three-dimensional (3D) body motions, ground reaction forces (GRF), and electromyography (EMG) as input data. Acquiring these data in real-world settings is challenging, with barriers such as the cost of instruments, setup time, and operator skills to correctly acquire and interpret data. This study investigated the consequences of limiting EMG and GRF data on modelled anterior cruciate ligament (ACL) forces during a drop-land-jump task in late-/post-pubertal females. We compared ACL forces generated by a reference model (i.e., EMG-informed neural mode combined with 3D GRF) to those generated by an EMG-informed with only vertical GRF, static optimisation with 3D GRF, and static optimisation with only vertical GRF. Results indicated ACL force magnitude during landing (when ACL injury typically occurs) was significantly overestimated if only vertical GRF were used for either EMG-informed or static optimisation neural modes. If 3D GRF were used in combination with static optimisation, ACL force was marginally overestimated compared to the reference model. None of the alternative models maintained rank order of ACL loading magnitudes generated by the reference model. Finally, we observed substantial variability across the study sample in response to limiting EMG and GRF data, indicating need for methods incorporating subject-specific measures of muscle activation patterns and external loading when modelling ACL loading during dynamic motor tasks.

Occupations or activities where donning head-supported mass (HSM) is commonplace put operators at an elevated risk of chronic neck pain. Yet, there is no consensus about what features of HSM influence the relative contributions to neck loads. Therefore, we tested four hypotheses that could increase neck loads: (i) HSM increases gravitational moments; (ii) more muscle activation is required to stabilize the head with HSM; (iii) the position of the HSM centre of mass (COM) induces gravitational moments; and (iv) the added moment of inertia (MOI) from HSM increases neck loads during head repositioning tasks. We performed a sensitivity analysis on the C5-C6 compression evaluated from a 24-degree freedom cervical spine model in OpenSim for static and dynamic movement trials. For static trials, we varied the magnitude of HSM, the position of its COM, and developed a novel stability constraint for static optimization. In dynamic trials, we varied HSM and the three principle MOIs. HSM magnitude and compression were linearly related to one another for both static and dynamic trials, with amplification factors varying between 1.9 and 3.9. Similar relationships were found for the COM position, although the relationship between C5-C6 peak compression and MOI in dynamic trials was generally nonlinear. This sensitivity analysis uncovered evidence in favour of hypotheses (i), (ii) and (iii). However, the model\'s prediction of C5-C6 compression was not overly sensitive to the magnitude of MOI. Therefore, the HSM mass properties may be more influential on neck compression than MOI properties, even during dynamic tasks.

Participant-specific musculoskeletal models are needed to accurately estimate lower back internal kinetic demands and injury risk. In this study we developed the framework for incorporating an electromyography optimization (EMGopt) approach within OpenSim (https://simtk.org/projects/emg_opt_tool) and evaluated lower back demands estimated from the model during gait. Kinematic, external kinetic, and EMG data were recorded from six participants as they performed walking and carrying tasks on a treadmill. For evaluation, predicted lumbar vertebral joint forces were compared to those from a generic static optimization approach (SOpt) and to previous studies. Further, model-estimated muscle activations were compared to recorded EMG, and model sensitivity to day-to-day EMG variability was evaluated. Results showed the vertebral joint forces from the model were qualitatively similar in pattern and magnitude to literature reports. Compared to SOpt, the EMGopt approach predicted larger joint loads (p<.01) with muscle activations better matching individual participant EMG patterns. L5/S1 vertebral joint forces from EMGopt were sensitive to the expected variability of recorded EMG, but the magnitude of these differences (±4%) did not impact between-task comparisons. Despite limitations inherent to such models, the proposed musculoskeletal model and EMGopt approach appears well-suited for evaluating internal lower back demands during gait tasks.

InverseMuscleNET, a machine learning model, is proposed as an alternative to static optimization for resolving the redundancy issue in inverse muscle models. A recurrent neural network (RNN) was optimally configured, trained, and tested to estimate the pattern of muscle activation signals. Five biomechanical variables (joint angle, joint velocity, joint acceleration, joint torque, and activation torque) were used as inputs to the RNN. A set of surface electromyography (EMG) signals, experimentally measured around the shoulder joint for flexion/extension, were used to train and validate the RNN model. The obtained machine learning model yields a normalized regression in the range of 88-91% between experimental data and estimated muscle activation. A sequential backward selection algorithm was used as a sensitivity analysis to discover the less dominant inputs. The order of most essential signals to least dominant ones was as follows: joint angle, activation torque, joint torque, joint velocity, and joint acceleration. The RNN model required 0.06 s of the previous biomechanical input signals and 0.01 s of the predicted feedback EMG signals, demonstrating the dynamic temporal relationships of the muscle activation profiles. The proposed approach permits a fast and direct estimation ability instead of iterative solutions for the inverse muscle model. It raises the possibility of integrating such a model in a real-time device for functional rehabilitation and sports evaluation devices with real-time estimation and tracking. This method provides clinicians with a means of estimating EMG activity without an invasive electrode setup.

Stable label movement and smooth label trajectory are critical for effective information understanding. Sudden label changes cannot be avoided by whatever forced directed methods due to the unreliability of resultant force or global optimization methods due to the complex trade-off on the different aspects. To solve this problem, we proposed a hybrid optimization method by taking advantages of the merits of both approaches. We first detect the spatial-temporal intersection regions from whole trajectories of the features, and initialize the layout by optimization in decreasing order by the number of the involved features. The label movements between the spatial-temporal intersection regions are determined by force directed methods. To cope with some features with high speed relative to neighbors, we introduced a force from future, called temporal force, so that the labels of related features can elude ahead of time and retain smooth movements. We also proposed a strategy by optimizing the label layout to predict the trajectories of features so that such global optimization method can be applied to streaming data.
UNASSIGNED: Supplementary material is available in the online version of this article at 10.1007/s41095-021-0231-y.