Individual-specific muscle maximum force estimation using ultrasound for ankle joint torque prediction using an EMG-driven Hill-type model

How mechanics of individual muscle-tendon units define knee and ankle joint function in health and cerebral palsy-a narrative review

This study reviews the relationship between muscle-tendon biomechanics and joint function, with a particular focus on how cerebral palsy (CP) affects this relationship. In healthy individuals, muscle size is a critical determinant of strength, with muscle volume, cross-sectional area, and moment arm correlating with knee and ankle joint torque for different isometric/isokinetic contractions. However, in CP, impaired muscle growth contributes to joint pathophysiology even though only a limited number of studies have investigated the impact of deficits in muscle size on pathological joint function. As muscles are the primary factors determining joint torque, in this review two main approaches used for muscle force quantification are discussed. The direct quantification of individual muscle forces from their relevant tendons through intraoperative approaches holds a high potential for characterizing healthy and diseased muscles but poses challenges due to the invasive nature of the technique. On the other hand, musculoskeletal models, using an inverse dynamic approach, can predict muscle forces, but rely on several assumptions and have inherent limitations. Neither technique has become established in routine clinical practice. Nevertheless, identifying the relative contribution of each muscle to the overall joint moment would be key for diagnosis and formulating efficient treatment strategies for patients with CP. This review emphasizes the necessity of implementing the intraoperative approach into general surgical practice, particularly for joint correction operations in diverse patient groups. Obtaining in vivo data directly would enhance musculoskeletal models, providing more accurate force estimations. This integrated approach can improve the clinicians' decision-making process and advance treatment strategies by predicting changes at the muscle and joint levels before interventions, thus, holding the potential to significantly enhance clinical outcomes.

In Vivo 3D Muscle Architecture Quantification Based on 3D Freehand Ultrasound and Magnetic Resonance Imaging

Muscle architecture parameters, such as the fascicle length, pennation angle, and volume, are important muscle morphology characteristics. Accurate in vivo quantification of these parameters allows to detect changes due to pathologies, interventions, and rehabilitation trainings, which ultimately impact on muscles' force-producing capacity. In this study, we compared three-dimensional (3D) muscle architecture parameters of the tibialis anterior and gastrocnemius medialis, which were quantified by 3D freehand ultrasound (3DfUS) and a magnetic resonance imaging (MRI) technique, diffusion tensor imaging (DTI), respectively. Sixteen able-bodied subjects were recruited where seven of them received both 3DfUS and MRI measurement, while the rest underwent 3DfUS measurements twice. Good to excellent intra-rater reliability and inter-session repeatability were found in 3DfUS measurements (intra-class correlation coefficient > 0.81). Overall, the two imaging modalities yielded consistent measurements of the fascicle length, pennation angle, and volume with mean differences smaller than 2.9 mm, 1.8°, and 5.7 cm³, respectively. The only significant difference was found in the pennation angle of the tibialis anterior, although the discrepancy was small. Our study demonstrated, for the first time, that 3DfUS measurement had high reliability and repeatability for measurement of muscle architecture in vivo and could be regarded as an alternative to MRI for 3D evaluation of muscle morphology.

Influence of Hamstrings on Knee Moments During Loading Response of Gait in Individuals Following ACL Reconstruction

Biomechanical studies consistently report smaller knee extensor moments in the surgical limb during loading response (LR) of gait following ACL reconstruction (ACLr). However, this reduction in knee loading is quantified by net joint moments (NJM). As a result, in the presence of greater hamstring activity, the true contribution from the knee extensors may not be reduced. The purpose of this study is to compare hamstring activity and strength and knee joint moments between individuals post-ACLr and controls. Eighteen individuals 3 months post-ACLr and matched controls walked and net knee extensor moment peak and impulse and hamstring activity were identified during LR, as well as maximal hamstring strength. A hybrid musculoskeletal model estimated knee flexor moments from joint kinematics and hamstring electromyography. Flexor moments (SIMM) were scaled based on strength. Knee extensor moments were estimated from the sum of the net knee moment and estimated knee flexor moment; estimated extensor moment peaks and impulse were calculated during LR. Repeated measures analysis of variance compared groups and limbs. Smaller net knee extensor moment and greater hamstring activity, as well as deficits in maximal hamstring strength, were observed in the surgical limb (all p < 0.05). When accounting for the torque-producing capabilities of the knee flexors, estimated knee extensor moment peak and impulse were smaller in the surgical limb. These findings suggest that net knee moments accurately reflect smaller knee extensor loading post-ACLr. Statement of Clinical Significance: Rehabilitation programs should target increasing knee extensor loading to restore gait mechanics during early rehabilitation. © 2019 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 38:378-386, 2020.

Determining Subject-Specific Lower-Limb Muscle Architecture Data for Musculoskeletal Models Using Diffusion Tensor Imaging

Accurate individualized muscle architecture data are crucial for generating subject-specific musculoskeletal models to investigate movement and dynamic muscle function. Diffusion tensor imaging (DTI) magnetic resonance (MR) imaging has emerged as a promising method of gathering muscle architecture data in vivo; however, its accuracy in estimating parameters such as muscle fiber lengths for creating subject-specific musculoskeletal models has not been tested. Here, we provide a validation of the method of using anatomical magnetic resonance imaging (MRI) and DTI to gather muscle architecture data in vivo by directly comparing those data obtained from MR scans of three human cadaveric lower limbs to those from dissections. DTI was used to measure fiber lengths and pennation angles, while the anatomical images were used to estimate muscle mass, which were used to calculate physiological cross-sectional area (PCSA). The same data were then obtained through dissections, where it was found that on average muscle masses and fiber lengths matched well between the two methods (4% and 1% differences, respectively), while PCSA values had slightly larger differences (6%). Overall, these results suggest that DTI is a promising technique to gather in vivo muscle architecture data, but further refinement and complementary imaging techniques may be needed to realize these goals.

Increasing hip and knee flexion during a drop-jump task reduces tibiofemoral shear and compressive forces: implications for ACL injury prevention training

Although most ACL injury prevention programmes encourage greater hip and knee flexion during landing, it remains unknown how this technique influences tibiofemoral joint forces. We examined whether a landing strategy utilising greater hip and knee flexion decreases tibiofemoral anterior shear and compression. Twelve healthy women (25.9 ± 3.5 years) performed a drop-jump task before and after a training session (10-15 min) that emphasised greater hip and knee flexion. Peak tibiofemoral anterior shear and compressive forces were calculated using an electromyography (EMG)-driven knee model that incorporated joint kinematics, EMG and participant-specific muscle volumes and patella tendon orientation measured using magnetic resonance imaging (MRI). Participants demonstrated a decrease in peak anterior tibial shear forces (11.1 ± 3.3 vs. 9.6 ± 2.7 N · kg^-1; P = 0.008) and peak tibiofemoral compressive forces (68.4 ± 7.6 vs. 62.0 ± 5.5 N · kg^-1; P = 0.015) post-training. The decreased peak anterior tibial shear was accompanied by a decrease in the quadriceps anterior shear force, while the decreased peak compressive force was accompanied by decreased ground reaction force and hamstring forces. Our data provide justification for injury prevention programmes that encourage greater hip and knee flexion during landing to reduce tibiofemoral joint loading.

Preliminary investigation of energy comparation between gyroscope, electromyography and VO2 wearable sensors

Building on previous experiments in the domain of energy expenditure estimation using wearable sensors, the measurements of energy ratios of a runner on a treadmill were analyzed to observe any commonalities between an inertia measurement unit and an electromyograph sensor. The subjects were equipped with a VO2 gas measurement device, an Inertial Measurement Unit (IMU) measuring gyroscopic activity and an electromyography (EMG) sensor network whilst running at 5 different speeds on a calibrated treadmill. The observed results established a co-linear relationship with the gyroscope based measurements, EMG based measurements with the VO2 measurements.

Tibialis anterior analysis from functional and architectural perspective during isometric foot dorsiflexion: a cross-sectional study of repeated measures

Background

The purpose of the present study is to establish the relationship and degree of contribution between torque and sonomiography variables (pennation angle – muscle thickness), and electromyography variables (EMG_{AreaUnderCurve} – EMG_MaximalPeak) of the tibialis anterior muscle during (TA) maximal and relative isometric foot dorsiflexion (IFD). Secondary aim: To determine the measurement’s reliability.

Methods

Cross-sectional study. 31 participants (15 men; 16 women) performed IFD at different intensities (100, 75, 50, and 25 %) of the maximal voluntary contraction (MVC) (three times for each intensity). Outcome variables: To determine the torque, pennation angle, muscle thickness, EMG_MaximalPeak, and EMG_{AreaUnderCurve}. Statistical analysis: In order to test the measurement’s reliability, Cronbach’s alpha and standard error of the measurement were determined. An inferential analysis was carried out using Pearson correlations(r). For each contraction intensity, a multiple regression analysis was performed, where the dependent variable was torque and the independent variables were EMG_{AreaUnderCurve}, EMG_MaximalPeak, muscle thickness and pennation angle.

Results

All outcome variables show excellent reliability. The highest correlation value was 0.955 (thickness 100 % – thickness 25 %). R² values ranged from 0.713 (100 % MVC) to 0.588 (25 % MVC).

Conclusion

The outcome variables demonstrated excellent reliability in terms of measuring IFD at different intensities. The correlations between all outcome variables were moderate-to-strong. TA functional and architectural variables have a significant impact on the torque variance during IFD at different intensities.

Vastus lateralis muscle architecture to estimate knee extension moment of older individuals

Abstract The aim of this study was to compare the knee extension moment of older individuals with the muscle moment estimated through a biomechanical model. This was accomplished by using (1) the specific muscle architecture data of individuals, and (2) the generic muscle architecture available in the literature. The muscle force estimate was determined using a model with the muscle architecture from cadavers and the individual vastus lateralis muscle architecture of sixteen older volunteers. For the muscle moment comparison, all of the volunteers performed maximal voluntary isometric contractions (MVIC) in five different knee extension position angles. The architectural data was acquired using both resonance and ultrasound imaging. Both estimated muscle moments (generic and individual) were higher than the experimental. The architecture of the other vastii may be necessary to make the model more accurate for the older population. Although other factors inherent to ageing, such as co-contractions, fiber type percentage, and passive forces are not considered in the model, they could be responsible for the differences between moments in older people.

Non-invasive isometric force measurement of plantar flexors in rats

INTRODUCTION

Isometric muscle force measurement is a sensitive marker for motor function recovery in rat nerve repair models. Current methods of eliciting maximal isometric force with nerve stimulation cannot provide longitudinal data.

METHODS

We developed a novel method for measuring isometric muscle force with a device designed to allow minimally invasive nerve stimulation and measurement of plantar flexion force. This indirectly elicited muscle force was compared with muscle force elicited by direct muscle stimulation in 3 surgical models.

RESULTS

The force measured after sciatic nerve transection and repair followed a parabolic trend. There was a postinjury decrease in force that continued until postoperative day 42, after which the force increased with time, indicating muscle reinnervation.

CONCLUSIONS

This approach can track longitudinal changes in force in the most common animal model for studies of clinically relevant problems in the peripheral nerve field.

A preliminary study on the differences in male and female muscle force distribution patterns during squatting and lunging maneuvers

In the United States, 250,000 people tear their anterior cruciate ligament (ACL) annually with females at higher risk of ACL failure than males. By predicting muscle forces during low impact maneuvers we may be able to estimate possible muscle imbalances that could lead to ACL failure during highly dynamic maneuvers. The purpose of this initial study was to predict muscle forces in males and females similar in size and activity level, during squat and lunge maneuvers. We hypothesized that during basic low impact maneuvers (a) distribution of quadriceps forces are different in males and females and (b) females exhibit quadriceps dominance when compared to males. Two males and three females performed squatting and lunging maneuvers while electromyography (EMG) data, motion capture data, and ground reaction forces were collected. Nine individual muscle forces for muscles that cross the knee were estimated using an EMG-driven model. Results suggest that males activate their rectus femoris muscle more than females, who in turn activate their vastus lateralis muscle at their maximum flexion angle, and more their vastus medialis muscle when ascending from a squat. During the lunge maneuver, males used greater biceps femoris force than females, throughout the lunge, and females exhibited higher semitendinosus force. Quadriceps dominance was evident in both males and females during the prescribed tasks, and there was no statistical difference between genders. Understanding individual muscle force distributions in males and females during low impact maneuvers may provide insights regarding failure mechanisms during highly dynamic maneuvers, when ACL injuries are more prevalent.

EMGD-FE: an open source graphical user interface for estimating isometric muscle forces in the lower limb using an EMG-driven model

Background

This paper describes the “EMG Driven Force Estimator (EMGD-FE)”, a Matlab® graphical user interface (GUI) application that estimates skeletal muscle forces from electromyography (EMG) signals. Muscle forces are obtained by numerically integrating a system of ordinary differential equations (ODEs) that simulates Hill-type muscle dynamics and that utilises EMG signals as input. In the current version, the GUI can estimate the forces of lower limb muscles executing isometric contractions. Muscles from other parts of the body can be tested as well, although no default values for model parameters are provided. To achieve accurate evaluations, EMG collection is performed simultaneously with torque measurement from a dynamometer. The computer application guides the user, step-by-step, to pre-process the raw EMG signals, create inputs for the muscle model, numerically integrate the ODEs and analyse the results.

Results

An example of the application’s functions is presented using the quadriceps femoris muscle. Individual muscle force estimations for the four components as well the knee isometric torque are shown.

Conclusions

The proposed GUI can estimate individual muscle forces from EMG signals of skeletal muscles. The estimation accuracy depends on several factors, including signal collection and modelling hypothesis issues.

Prediction of maximal surface electromyographically based voluntary contractions of erector spinae muscles from sonographic measurements during isometric contractions

OBJECTIVES

Currently, there are no studies combining electromyography (EMG) and sonography to estimate the absolute and relative strength values of erector spinae (ES) muscles in healthy individuals. The purpose of this study was to establish whether the maximum voluntary contraction (MVC) of the ES during isometric contractions could be predicted from the changes in surface EMG as well as in fiber pennation and thickness as measured by sonography.

METHODS

Thirty healthy adults performed 3 isometric extensions at 45° from the vertical to calculate the MVC force. Contractions at 33% and 100% of the MVC force were then used during sonographic and EMG recordings. These measurements were used to observe the architecture and function of the muscles during contraction. Statistical analysis was performed using bivariate regression and regression equations.

RESULTS

The slope for each regression equation was statistically significant (P < .001) with R(2) values of 0.837 and 0.986 for the right and left ES, respectively. The standard error estimate between the sonographic measurements and the regression-estimated pennation angles for the right and left ES were 0.10 and 0.02, respectively.

CONCLUSIONS

Erector spinae muscle activation can be predicted from the changes in fiber pennation during isometric contractions at 33% and 100% of the MVC force. These findings could be essential for developing a regression equation that could estimate the level of muscle activation from changes in the muscle architecture.

Relationship of moderate and low isometric lumbar extension through architectural and muscular activity variables: a cross sectional study

BACKGROUND

No study relating the changes obtained in the architecture of erector spinae (ES) muscle were registered with ultrasound and different intensities of muscle contraction recorded by surface EMG (electromyography) on the ES muscle was found. The aim of this study was analyse the relationship in the response of the ES muscle during isometric moderate and light lumbar isometric extension considering architecture and functional muscle variables.

METHODS

Cross-sectional study. 46 subjects (52% men) with a group mean age of 30.4 (±7.78). The participants developed isometric lumbar extension while performing moderate and low isometric trunk and hip extension in a sitting position with hips flexed 90 degrees and the lumbar spine in neutral position. During these measurements, electromyography recordings and ultrasound images were taken bilaterally. Bilaterally pennation angle, muscle thickness, torque and muscle activation were measured. This study was developed at the human movement analysis laboratory of the Health Science Faculty of the University of Malaga (Spain).

RESULTS

Strong and moderate correlations were found at moderate and low intensities contraction between the variable of the same intensity, with correlation values ranging from 0.726 (Torque Moderate - EMG Left Moderate) to 0.923 (Angle Left Light - Angle Right Light) (p < 0.001). This correlation is observed between the variables that describe the same intensity of contraction, showing a poor correlation between variables of different intensities.

CONCLUSION

There is a strong relationship between architecture and function variables of ES muscle when describe an isometric lumbar extension at light or moderate intensity.

Ultrasound-based subject-specific parameters improve fascicle behaviour estimation in Hill-type muscle model

The estimation of muscle fascicle behaviour is decisive in a Hill-type model as they are related to muscle force by the force-length-velocity relationship and the tendon force-strain relationship. This study was aimed at investigating the influence of subject-specific tendon force-strain relationship and initial fascicle geometry (IFG) on the estimation of muscle forces and fascicle behaviour during isometric contractions. Ultrasonography was used to estimate the in vivo muscle fascicle behaviour and compare the muscle fascicle length and pennation angle estimated from the Hill-type model. The calibration-prediction process of the electromyography-driven model was performed using generic or subject-specific tendon definition with or without IFG as constraint. The combination of subject-specific tendon definition and IFG led to muscle fascicle behaviour closer to ultrasound data and significant lower forces of the ankle dorsiflexor and plantarflexor muscles compared to the other conditions. Thus, subject-specific ultrasound measurements improve the accuracy of Hill-type models on muscle fascicle behaviour.

Increased hip and knee flexion during landing decreases tibiofemoral compressive forces in women who have undergone anterior cruciate ligament reconstruction

BACKGROUND

Those who have undergone anterior cruciate ligament reconstruction (ACLR) have been shown to exhibit increased muscle co-contraction, decreased knee flexion, and elevated tibiofemoral compressive forces. Elevated tibiofemoral compressive forces may be associated with the high risk of developing knee osteoarthritis in this population.

PURPOSE

To examine whether muscle co-contraction and tibiofemoral compressive forces in women after undergoing ACLR can be reduced through the use of a landing strategy that emphasizes greater hip and knee flexion.

STUDY DESIGN

Controlled laboratory study.

METHODS

Ten female recreational athletes who had previously undergone ACLR participated in this study. Participants performed a single-legged drop-land task before and after a training session that encouraged them to use greater hip and knee flexion during landing. Peak tibiofemoral compressive forces before and after training were estimated using an electromyography (EMG)-driven knee model that incorporated joint kinematics, EMG, and subject-specific muscle volumes and patellar tendon orientation estimated from magnetic resonance imaging. A co-contraction index (CCI) was calculated to quantify the level of co-contraction between knee flexor and extensor muscles.

RESULTS

After training, peak hip and knee flexion as well as hip and knee flexion excursions increased significantly. Additionally, participants demonstrated a significant decrease after training in the areas of muscle co-contraction (CCI [mean ± SD], 0.28 ± 0.10 vs 0.18 ± 0.05; P < .001) and peak tibiofemoral compressive force (97.3 ± 8.0 vs 91.3 ± 10.2 N·kg(-1); P = .044).

CONCLUSION

Increased muscle co-contraction as well as elevated tibiofemoral compressive loads observed in individuals following ACLR can be reduced by using a landing strategy that encourages greater hip and knee flexion.

CLINICAL RELEVANCE

The findings of the current study provide useful information for the growth of rehabilitation and/or intervention programs aimed to decrease knee joint loading to prevent or delay the development of knee osteoarthritis in those who have undergone ACLR.

An isovelocity dynamometer method to determine monoarticular and biarticular muscle parameters

This study aimed to determine whether subject-specific individual muscle models for the ankle plantar flexors could be obtained from single joint isometric and isovelocity maximum torque measurements in combination with a model of plantar flexion. Maximum plantar flexion torque measurements were taken on one subject at six knee angles spanning full flexion to full extension. A planar three-segment (foot, shank and thigh), two-muscle (soleus and gastrocnemius) model of plantar flexion was developed. Seven parameters per muscle were determined by minimizing a weighted root mean square difference (wRMSD) between the model output and the experimental torque data. Valid individual muscle models were obtained using experimental data from only two knee angles giving a wRMSD score of 16 N m, with values ranging from 11 to 17 N m for each of the six knee angles. The robustness of the methodology was confirmed through repeating the optimization with perturbed experimental torques (± 20%) and segment lengths (± 10%) resulting in wRMSD scores of between 13 and 20 N m. Hence, good representations of maximum torque can be achieved from subject-specific individual muscle models determined from single joint maximum torque measurements. The proposed methodology could be applied to muscle-driven models of human movement with the potential to improve their validity.

Importance and challenges of measuring intrinsic foot muscle strength

BACKGROUND

Intrinsic foot muscle weakness has been implicated in a range of foot deformities and disorders. However, to establish a relationship between intrinsic muscle weakness and foot pathology, an objective measure of intrinsic muscle strength is needed. The aim of this review was to provide an overview of the anatomy and role of intrinsic foot muscles, implications of intrinsic weakness and evaluate the different methods used to measure intrinsic foot muscle strength.

METHOD

Literature was sourced from database searches of MEDLINE, PubMed, SCOPUS, Cochrane Library, PEDro and CINAHL up to June 2012.

RESULTS

There is no widely accepted method of measuring intrinsic foot muscle strength. Methods to estimate toe flexor muscle strength include the paper grip test, plantar pressure, toe dynamometry, and the intrinsic positive test. Hand-held dynamometry has excellent interrater and intrarater reliability and limits toe curling, which is an action hypothesised to activate extrinsic toe flexor muscles. However, it is unclear whether any method can actually isolate intrinsic muscle strength. Also most methods measure only toe flexor strength and other actions such as toe extension and abduction have not been adequately assessed. Indirect methods to investigate intrinsic muscle structure and performance include CT, ultrasonography, MRI, EMG, and muscle biopsy. Indirect methods often discriminate between intrinsic and extrinsic muscles, but lack the ability to measure muscle force.

CONCLUSIONS

There are many challenges to accurately measure intrinsic muscle strength in isolation. Most studies have measured toe flexor strength as a surrogate measure of intrinsic muscle strength. Hand-held dynamometry appears to be a promising method of estimating intrinsic muscle strength. However, the contribution of extrinsic muscles cannot be excluded from toe flexor strength measurement. Future research should clarify the relative contribution of intrinsic and extrinsic muscles during intrinsic foot muscle strength testing.

Quantification of tibiofemoral shear and compressive loads using a MRI-based EMG-driven knee model

The purpose of this study is to describe an MRI-based EMG-driven knee model to quantify tibiofemoral compressive and shear forces. Twelve healthy females participated. Subjects underwent 2 phases of data collection: (1) MRI assessment of the lower extremity to quantify muscle volumes and patella tendon orientation and (2) biomechanical evaluation of a drop-jump task. A subject-specific EMG-driven knee model that incorporated lower extremity kinematics, EMG, and muscle volumes and patella tendon orientation estimated from MRI was developed to quantify tibiofemoral shear and compressive forces. A resultant anterior tibial shear force generated from the ground reaction force (GRF) and muscle forces was observed during the first 30% of the stance phase of the drop-jump task. All of the muscle forces and GRF resulted in tibiofemoral compression, with the quadriceps force being the primary contributor. Acquiring subject-specific muscle volumes and patella tendon orientation for use in an EMG-driven knee model may be useful to quantify tibiofemoral forces in persons with altered patella position or muscle atrophy following knee injury or pathology.

Effect of innervation zones in estimating biceps brachii force-EMG relationship during isometric contraction

Measuring muscle forces in vivo is invasive and consequently indirect methods e.g., electromyography (EMG) are used in estimating muscular force production. The aim of the present paper was to examine what kind of effect the disruption of the physiological signal caused by the innervation zone has in predicting the force/torque output from surface EMG. Twelve men (age 26 (SD ±3)years; height 179 (±6)cm; body mass 73 (±6)kg) volunteered as subjects. They were asked to perform maximal voluntary isometric contraction (MVC) in elbow flexion, and submaximal contractions at 10%, 20%, 30%, 40%, 50% and 75% of the recorded MVC. EMG was measured from biceps brachii muscle with an electrode grid of 5 columns×13 rows. Force-EMG relationships were determined from individual channels and as the global mean value. The relationship was deemed inconsistent if EMG value did not increase in successive force levels. Root mean squared errors were calculated for 3rd order polynomial fits. All subjects had at least one (4-52) inconsistent channel. Two subjects had inconsistent relationship calculated from the global mean. The mean root mean squared error calculated using leave one out method for the fits of the individual channels (0.33±0.17) was higher (P<0.001) than the error for the global mean fit (0.16±0.08). It seems that the disruption of the physiological signal caused by the innervation zone affects the consistency of the force-EMG relationship on single bipolar channel level. Multichannel EMG recordings used for predicting force overcame this disruption.

Sensitivity analysis of an energetic muscle model applied at whole body level in recumbent pedalling

Musculoskeletal models are used in order to describe and analyse the mechanics of human movement. In order to get a complete evaluation of the human movement, energetic muscle models were developed and were shown to be promising. The aim of this work is to determine the sensitivity of muscle mechanical and energetic model estimates to changes in parameters during recumbent pedalling. Inputs of the model were electromyography and joint angles, collected experimentally on one participant. The sensitivity analysis was performed on muscle-specific tension, physiological cross-sectional area, muscle maximal force, tendon rest length and percentage of fast-twitch fibres using an integrated sensitivity ratio. Soleus, gastrocnemius, vasti, gluteus and medial hamstrings were selected for the analyses. The energetic model was found to be always less sensitive to parameter changes than the mechanical model. Tendon slack length was found to be the most critical parameter for both energetic and mechanical models even if the effect on the energetic output was smaller than on muscle force and joint moments.