Sensitivity of predicted muscle forces during gait to anatomical variability in musculotendon geometry

Abstract

Scaled generic musculoskeletal models are commonly used to drive dynamic simulations of motions. It is however, acknowledged that not accounting for variability in musculoskeletal geometry and musculotendon parameters may confound the simulation results, even when analysing control subjects. This study documents the three-dimensional anatomical variability of musculotendon origins and insertions of 33 lower limb muscles determined based on magnetic resonance imaging in six subjects. This anatomical variability was compared to the musculotendon point location in a generic musculoskeletal model. Furthermore, the sensitivity of muscle forces during gait, calculated using static optimization, to perturbations of the musculotendon point location was analyzed with a generic model. More specific, a probabilistic approach was used: for each analyzed musculotendon point, the three-dimensional location was re-sampled with a uniform Latin hypercube method within the anatomical variability and the static optimization problem was then re-solved for all perturbations. We found that musculotendon point locations in the generic model showed only variable correspondences with the anatomical variability. The anatomical variability of musculotendon point location did affect the calculated muscle forces: muscles most sensitive to perturbations within the anatomical variability are iliacus and psoas. Perturbation of the gluteus medius anterior, iliacus and psoas induces the largest concomitant changes in muscle forces of the unperturbed muscles. Therefore, when creating subject-specific musculoskeletal models, these attachment points should be defined accurately. In addition, the size of the anatomical variability of the musculotendon point location was not related to the sensitivity of the calculated muscle forces.

A screening method to analyse the sensitivity of a lower limb multibody kinematic model

“…Whatever the approach used to identify the influential parameters, it seems important to analyse the evolution of the sensitivity measures or indices over the gait cycle as they may be different between the stance and swing phases or may vary with the joint amplitude, as it appears in the present study, depending on the knee flexion-extension. Some previous sensitivity analyses of a lower limb models have reported time averages [4,12,15,35,41], time integrals [9,29,43] as well as correlation coefficients [17,25,26,38] or root mean square differences [1] computed on the whole gait cycle. One study has reported partial derivatives of the output with respect to the parameters (calculated using a finite-difference approximations) at each instant of time of the gait cycle [36] and demonstrated varying influence of some parameters on the lower limb joint moments.…”

“…The sensitivity analyses 1 Computer Methods in Biomechanics and Biomedical Engineering, 22:10, 925-935, 2019 were generally based on a Monte Carlo propagation of uncertainties and different ad hoc sensitivity indices. As far as gait analysis is concerned, the sensitivity is to be assessed at different instants of time and for numerous outputs (e.g., angles, moments) and that is why many indices involved time averages [4,12,15,35,41] or time integrals [9,29,43]. As a matter of fact, these sensitivity analyses have a high numerical cost which may also explain why only first order sensitivity indices, which give the expected reduction in the response variance by fixing one factor [32], have been used although non-linear effects and interactions between parameters can be of high importance [15].…”

I.S.Mu.L.T. Achilles tendon ruptures guidelines

“…First, the assumption that IC can be uniquely addressed to the plantar-flexors muscles (hence, excluding co-contraction of antagonist muscles 362 and neglecting the contribution of passive forces exerted by ligaments 366 ). Second, the accuracy of the estimated AT force strongly depends on the reliability of the collected experimental data (anatomical landmarks identification and skin artefact in the first place [367][368][369] ) and on the chosen musculoskeletal model (inertial parameters and musculoskeletal geometries are based on generic models scaled on the subject's proportions) 370 . For this latter reason, the scientific community has been recently focusing on the availability of imaging techniques to assess subject-specific musculoskeletal geometries simultaneously to motion data collection to estimate ankle dynamics [371][372][373] .…”

Subject-specific lower extremity modeling: personalization of musculoskeletal models using medical imaging and functional measurements

“…However, this simplification had hardly an effect on the model outcomes, since these points have shown really low sensitivity . Similarly, low sensitivity of segment inertial parameters (Bosmans et al, 2015) ensure negligible effects of the linear scaling on the model predictions. The main limitation of the subject-specific models used in this study lies in the resolution and overall quality of the available medical images.…”

“…For these complex applications, an accurate description of the musculoskeletal geometry is essential to obtain reliable model predictions. It is well known that inaccuracies in muscle-tendon (MT) moment arms (Hoy et al, 1990;Out et al, 1996;Maganaris, 2004), MT attachment sites and lines-of-action (Pal et al, 2007;Valente et al, 2014;Bosmans et al, 2015), and joint geometry (Martelli et al, 2015;Valente et al, 2015) affect model outcomes. Unfortunately, application of linear scaling laws to generic models (Delp et al, 1990;Klein Horsman et al, 2007;Arnold et al, 2010) cannot account for the inter-individual anatomical variability due to differences in age, height, weight and gender (White et al, 1989;Duda et al, 1996;Kepple et al, 1998).…”