Joint kinematics estimation using a multi-body kinematics optimisation and an extended Kalman filter,and embedding a soft tissue artefact model

To reduce the impact of the soft tissue artefact (STA) on the estimate of skeletal movement using stereophotogrammetric and skin-marker data, multi-body kinematics optimisation (MKO) and extended Kalman filters (EKF) have been proposed. This paper assessed the feasibility and efficiency of these methods when they embed a mathematical model of the STA and simultaneously estimate the ankle, knee and hip joint kinematics and the model parameters. A STA model was used that provides an estimate of the STA affecting the marker-cluster located on a body segment as a function of the kinematics of the adjacent joints. The MKO and the EKF were implemented with and without the STA model. To assess these methods, intra-cortical pin and skin markers located on the thigh, shank, and foot of three subjects and tracked during the stance phase of running were used. Embedding the STA model in MKO and EKF reduced the average RMS of marker tracking from 12.6 to 1.6 mm and from 4.3 to 1.9 mm, respectively, showing that a STA model trial-specific calibration is feasible. Nevertheless, with the STA model embedded in MKO, the RMS difference between the estimated and the reference joint kinematics determined from the pin markers slightly increased (from 2.0 to 2.1 deg) On the contrary, when the STA model was embedded in the EKF, this RMS difference was slightly reduced (from 2.0 to 1.7 deg) thus showing a better potentiality of this method to attenuate STA effects and improve the accuracy of joint kinematics estimate.     

Effects of soft tissue artifacts on the calculated kinematics and kinetics of the knee during stair-ascent

Biomechanics of the knee during stair-ascent has mostly been studied using skin-marker-based motion analysis techniques, but no study has reported a complete assessment of the soft tissue artifacts (STA) and their effects on the calculated joint center translation, angles and moments at the knee in normal subjects during this activity. This study aimed to bridge the gap. Twelve young adults walked up a three-step stair while data were acquired simultaneously from a three-dimensional motion capture system, a force plate and a dynamic fluoroscopy system. The "gold standards" of poses of the knee were obtained using a 3D fluoroscopy method. The STA of the markers on the thigh and shank were then calculated, together with their effects on the calculated joint center translations, angles and moments at the knee. The STA of the thigh markers were greater than those on the shank, leading to significantly underestimated flexion and extensor moments, but overestimated joint center translations during the first half of the stance phase. The results will be useful for a better understanding of the normal biomechanics of the knee during stair-ascent, as a baseline for future clinical applications and for developing a compensation method to correct for the effects of STA.     

A constrained extended Kalman filter for the optimal estimate of kinematics and kinetics of a sagittal symmetric exercise

This paper presents a method for real-time estimation of the kinematics and kinetics of a human body performing a sagittal symmetric motor task, which would minimize the impact of the stereophotogrammetric soft tissue artefacts (STA). The method is based on a bi-dimensional mechanical model of the locomotor apparatus the state variables of which (joint angles, velocities and accelerations, and the segments lengths and inertial parameters) are estimated by a constrained extended Kalman filter (CEKF) that fuses input information made of both stereophotogrammetric and dynamometric measurement data. Filter gains are made to saturate in order to obtain plausible state variables and the measurement covariance matrix of the filter accounts for the expected STA maximal amplitudes. We hypothesised that the ensemble of constraints and input redundant information would allow the method to attenuate the STA propagation to the end results. The method was evaluated in ten human subjects performing a squat exercise. The CEKF estimated and measured skin marker trajectories exhibited a RMS difference lower than 4 mm, thus in the range of STAs. The RMS differences between the measured ground reaction force and moment and those estimated using the proposed method (9 N and 10 N m) were much lower than obtained using a classical inverse dynamics approach (22 N and 30 N m). From the latter results it may be inferred that the presented method allows for a significant improvement of the accuracy with which kinematic variables and relevant time derivatives, model parameters and, therefore, intersegmental moments are estimated.     

Effect of various upper limb multibody models on soft tissue artefact correction: A case study

Soft tissue artefacts (STA) introduce errors in joint kinematics when using cutaneous markers, especially on the scapula. Both segmental optimisation and multibody kinematics optimisation (MKO) algorithms have been developed to improve kinematics estimates. MKO based on a chain model with joint constraints avoids apparent joint dislocation but is sensitive to the biofidelity of chosen joint constraints. Since no recommendation exists for the scapula, our objective was to determine the best models to accurately estimate its kinematics. One participant was equipped with skin markers and with an intracortical pin screwed in the scapula. Segmental optimisation and MKO for 24-chain models (including four variations of the scapulothoracic joint) were compared against the pin-derived kinematics using root mean square error (RMSE) on Cardan angles. Segmental optimisation led to an accurate scapula kinematics (1.1° ≤ RMSE ≤ 3.3°) even for high arm elevation angles. When MKO was applied, no clinically significant difference was found between the different scapulothoracic models (0.9° ≤ RMSE ≤ 4.1°) except when a free scapulothoracic joint was modelled (1.9° ≤ RMSE ≤ 9.6°). To conclude, using MKO as a STA correction method was not more accurate than segmental optimisation for estimating scapula kinematics.     

Effect of seat positions on discomfort,muscle activation,pressure distribution and pedal force during cycling

The aim of this study was to measure and analyse discomfort and biomechanics of cycling, i.e., muscle activation, centre of pressure of seat pressure profiles and pedal forces as a function of seat position. Twenty-one recreationally active individuals cycled for 10 min at 100 W on an ergometer cycle using five different seat positions. The neutral position was considered as basic seat position and was compared with upward, downward, forward and backward seat positions. The initial bout was repeated at the end of the recording session. Discomfort increased for upward and backward condition compared with neutral (P < 0.05). Normalized surface electromyography from gastrocnemius decreased in the downward and forward position but increased in the upward and backward position. The minimum force became less negative for forward position compared with neutral seat position (P < 0.05). The degree of variability of centre of pressure increased in the upward and backward position and the entropy of the centre of pressure of sitting posture for backward position decreased compared with neutral seat position (P < 0.05). The present study revealed that consecutive changes of seat position over time lead to increase in discomfort as well as alterations of the biomechanics of cycling.     

Electromyographic patterns of lower limb muscles during apprehensive gait in younger and older female adults

ObjectiveInvestigate the influence of apprehensive gait on activation and cocontraction of lower limb muscles of younger and older female adults.MethodsData of 17 younger (21.47 ± 2.06 yr) and 18 older women (65.33 ± 3.14 yr) were considered for this study. Participants walked on the treadmill at two different conditions: normal gait and apprehensive gait. The surface electromyographic signals (EMG) were recorded during both conditions on: rectus femoris (RF), vastus lateralis (VL), vastus medialis (VM), biceps femoris (BF), tibialis anterior (TA), gastrocnemius lateralis (GL), and soleus (SO).ResultsApprehensive gait promoted greater activation of thigh muscles than normal gait (F = 5.34 and p = 0.007, for significant main effect of condition; RF, p = 0.002; VM, p < 0.001; VL, p = 0.003; and BF, p = 0.001). Older adults had greater cocontraction of knee and ankle stabilizer muscles than younger women (F = 4.05 and p = 0.019, for significant main effect of groups; VM/BF, p = 0.010; TA/GL, p = 0.007; and TA/SO, p = 0.002).ConclusionApprehensive gait promoted greater activation of thigh muscles and older adults had greater cocontraction of knee and ankle stabilizer muscles. Thus, apprehensive gait may leads to increased percentage of neuromuscular capacity, which is associated with greater cocontraction and contribute to the onset of fatigue and increased risk of falling in older people.     

Influence of the knee flexion on muscle activation and transmissibility during whole body vibration

The influence of the knee flexion on muscle activation and transmissibility during whole body vibration is controversially discussed in the literature. In this study, 34 individuals had electromyography activity (EMG) of the vastus lateralis and the acceleration assessed while squatting with 60° and 90° of knee flexion either with or without whole-body vibration (WBV). The conditions were maintained for 10 s with 1 min of rest between each condition. The main findings were (1) the larger the angle of knee flexion (90° vs. 60°), the greater the EMG (p < 0.001), with no difference on acceleration transmissibility; (2) for both angles of knee flexion, the addition of WBV produced no significant difference in EMG and higher acceleration compared to without WBV (p < 0.001). These results suggest that the larger the knee flexion angle (60° vs. 90°), the greater the muscle activation without acceleration modification. However, the addition of WBV increases the transmissibility of acceleration in the lower limbs without modification in EMG of vastus lateralis.     

Sharper angle,higher risk? The effect of cutting angle on knee mechanics in invasion sport athletes

IntroductionCutting is an important skill in team-sports, but unfortunately is also related to non-contact ACL injuries. The purpose was to examine knee kinetics and kinematics at different cutting angles.Material and methods13 males and 16 females performed cuts at different angles (45°, 90°, 135° and 180°) at maximum speed. 3D kinematics and kinetics were collected. To determine differences across cutting angles (45°, 90°, 135° and 180°) and sex (female, male), a 4 × 2 repeated measures ANOVA was conducted followed by post hoc comparisons (Bonferroni) with alpha level set at α ≤ 0.05 a priori.ResultsAt all cutting angles, males showed greater knee flexion angles than females (p < 0.01). Also, where males performed all cutting angles with no differences in the amount of knee flexion −42.53° ± 8.95°, females decreased their knee flexion angle from −40.6° ± 7.2° when cutting at 45° to −36.81° ± 9.10° when cutting at 90°, 135° and 180° (p < 0.01). Knee flexion moment decreased for both sexes when cutting towards sharper angles (p < 0.05). At 90°, 135° and 180°, males showed greater knee valgus moments than females. For both sexes, knee valgus moment increased towards the sharper cutting angles and then stabilized compared to the 45° cutting angle (p < 0.01). Both females and males showed smaller vGRF when cutting to sharper angles (p < 0.01).ConclusionIt can be concluded that different cutting angles demand different knee kinematics and kinetics. Sharper cutting angles place the knee more at risk. However, females and males handle this differently, which has implications for injury prevention.     

Effect of posterior cruciate ligament creep on muscular co-activation around knee: A pilot study

The effect of posterior cruciate ligament (PCL) on muscle co-activation (MCO) is not known though MCO has been extensively studied. The purpose of the study was to investigate the effect of PCL creep on MCO and on joint moment around the knee. Twelve males and twelve females volunteered for this study. PCL creep was estimated via tibial posterior displacement which was elicited by a 20 kg dumbbell hanged on horizontal shank near patella for 10 min. Electromyography activity from both rectus femoris and biceps femoris as well as muscle strength on the right thigh was recorded synchronically during knee isokinetic flexion–extension performance in speed of 60 deg/s as well as 120 deg/s on a dynamometer before and after PCL creep. A one-way ANOVA with repeated measures was used to evaluate the effect of creep, gender and speed. The results showed that significant tibial posterior displacement was found (p = 0.01) in both male and female groups. No significant increase of joint moment was found in flexion as well as in extension phase in both female and male groups. There was a significant effect of speed (p = 0.036) on joint moment in extension phase. Co-activation index (CI) decreased significantly (p = 0.049) in extension phase with a significant effect of gender (p ⩽ 0.001). It was concluded that creep developed in PCL due to static posterior load on the proximal tibia could significantly elicit the increase of the activation of agonist muscles but with no compensation from the antagonist in flexion as well as in extension phase. The creep significantly elicited the decrease of the antagonist–agonist CI in extension phase. MCO in females was reduced significantly in extension phase. It was suggested that PCL creep might be one of risk factors to the knee injury in sports activity.     

A soft tissue artefact model driven by proximal and distal joint kinematics

When analysing human movement through stereophotogrammetry, skin-markers are used. Their movement relative to the underlying bone is known as a soft tissue artefact (STA). A mathematical model to estimate subject- and marker-specific STAs generated during a given motor task, is required for both skeletal kinematic estimators and comparative assessment using simulation. This study devises and assesses such a mathematical model using the paradigmatic case of thigh STAs. The model was based on two hypotheses: (1) that the artefact mostly depends on skin sliding, and thus on the angles of hip and knee; (2) that the relevant relationship is linear. These hypotheses were tested using data obtained from passive hip and knee movements in non-obese specimens and from running volunteers endowed with both skin- and pin-markers.     

Is myoelectric activity distributed equally within the rectus femoris muscle during loaded,squat exercises?

Recent evidence suggests different regions of the rectus femoris (RF) muscle respond differently to squat exercises. Such differential adaptation may result from neural inputs distributed locally within RF, as previously reported for isometric contractions, walking and in response to fatigue. Here we therefore investigate whether myoelectric activity distributes evenly within RF during squat. Surface electromyograms (EMGs) were sampled proximally and distally from RF with arrays of electrodes, while thirteen healthy volunteers performed 10 consecutive squats with 20% and 40% of their body weight. The root mean square (RMS) value, computed separately for thirds of the concentric and eccentric phases, was considered to assess the proximo-distal changes in EMG amplitude during squat. The channels with variations in EMG amplitude during squat associated with shifts in the muscle innervation zone were excluded from analysis. No significant differences were observed between RF regions when considering squat phases and knee joint angles individually (P > 0.16) while a significant interaction between phase and knee joint angle with detection site was observed (P < 0.005). For the two loads considered, proximal RMS values were greater during the eccentric phase and for the more flexed knee joint position (P < 0.001). Our results suggest inferences on the degree of RF activation during squat must be made cautiously from surface EMGs. Of more practical relevance, there may be a potential for the differential adaption of RF proximal and distal regions to squat exercises.     

Effects of two neuromuscular fatigue protocols on landing performance

The purpose of the study was to investigate the effects of two fatigue protocols on landing performance. A repeated measures design was used to examine the effects of fatigue and fatigue protocol on neuromuscular and biomechanical performance variables. Ten volunteers performed non-fatigued and fatigued landings on two days using different fatigue protocols. Repeated maximum isometric squats were used to induce fatigue on day one. Sub-maximum cycling was used to induce fatigue on day two. Isometric squat maximum voluntary contraction (MVC) was measured before and after fatigued landings on each day. During the landings, ground reaction force (GRF), knee kinematics, and electromyographic (EMG) data were recorded. Isometric MVC, GRF peaks, loading rates, impulse, knee flexion at contact, range of motion, max angular velocity, and EMG root mean square (RMS) values were compared pre- and post-fatiguing exercise and between fatigue protocols using repeated ANOVA. Fatigue decreased MVC strength (p ? 0.05), GRF second peak, and initial impulse (p ? 0.01), but increased quadriceps medium latency stretch reflex EMG activity (p ? 0.012). Knee flexion at contact was 5.2° greater (p ? 0.05) during fatigued landings following the squat exercise compared to cycling. Several variables exhibited non-significant but large effect sizes when comparing the effects of fatigue and fatigue protocol. In conclusion, fatigue alters landing performance and different fatigue protocols result in different performance changes.     

Muscle fatigue and metabolic responses following three different antagonist pre-load resistance exercises

PurposePreload of antagonist muscles can be achieved by reciprocal actions (RAs) or by opposing muscle actions. However, evidence concerning neuromuscular and fatigue responses are scarce.ObjectiveTo compare the effects of different knee flexor (KF) preload methods on knee extension (KE) vastus medialis muscle fatigue, based on EMG-spectral index (FI), load range (LR), total work (TW), blood lactate (LAC) and biceps femoris co-activation (BFc) during resistance exercise.MethodsTwenty-four healthy men (23.5 ± 3.6 yrs) performed three antagonist pre-load isokinetic exercises (4 sets, 10 repetitions, 60° s^?1, 1 min rest between sets): RA (KF contraction immediately followed by KE); Superset (SS; one KF set immediately followed by one KE set); Multiple Set (MS; four KF sets followed by four KE sets).ResultsTotal work was significantly greater in RA. There was no significant decrease in LR between sets in RA. The BFc did not differ between protocols (p = 0.063). However, RA presented greater biceps femoriscoactivation. The FI was greater during SS compared to RA and MS (p < 0.05). The SS had greater LAC when compared to MS and RA (p = 0.005 and p = 0.007, respectively).ConclusionIt is suggested that the RA protocol is more neuromuscular and metabolic efficient during the performance of knee extension resistance exercise.     

Low-frequency accelerations over-estimate impact-related shock during walking

During gait, a failure to acknowledge the low-frequency component of a segmental acceleration signal will result in an overestimation of impact-related shock and may lead to inappropriately drawn conclusions. The present study was undertaken to investigate the significance of this low-frequency component in two distinctly different modalities of gait: barefoot (BF) and shod (SHOD) walking. Twenty-seven participants performed five walking trials at self-selected speed in each condition. Peak positive accelerations (PPA) at the shank and spine were first derived from the time-domain signal. The raw acceleration signals were then resolved in the frequency-domain and the active (low-frequency) and impact-related components of the power spectrum density (PSD) were quantified. PPA was significantly higher at the shank (P < 0.0001) and spine (P = 0.0007) in the BF condition. In contrast, no significant differences were apparent between conditions for shank (P = 0.979) or spine (P = 0.178) impact-related PSD when the low-frequency component was considered. This disparity between approaches was due to a significantly higher active PSD in both signals in the BF condition (P < 0.0001; P = 0.008, respectively), due to kinematic differences between conditions (P < 0.05). These results indicate that the amplitude of the low-frequency component of an acceleration signal during gait is dependent on knee and ankle joint coordination behaviour, and highlight that impact-related shock is more accurately quantified in the frequency-domain following subtraction of this component.     

In vivo hip joint loads and pedal forces during ergometer cycling

The rising prevalence of osteoarthritis and an increase in total hip replacements calls for attention to potential therapeutic activities. Cycling is considered as a low impact exercise for the hip joint and hence recommended. However, there are limited data about hip joint loading to support this claim. The aim of this study was to measure synchronously the in vivo hip joint loads and pedal forces during cycling. The in vivo hip joint loads were measured in 5 patients with instrumented hip implants. Data were collected at several combinations of power and cadence, at two saddle heights.Joint loads and pedal forces showed strong linear correlation with power. So the relationship between the external pedal forces and internal joint forces was shown. While cycling at different cadences the minimum joint loads were acquired at 60 RPM. The lower saddle height configuration results in an approximately 15% increase compared to normal saddle height.The results offered new insights into the actual effects of cycling on the hip joint and can serve as useful tools while developing an optimum cycling regimen for individuals with coxarthrosis or following total hip arthroplasty. Due to the relatively low contact forces, cycling at a moderate power level of 90 W at a normal saddle height is suitable for patients.     

Braking characteristics during cutting and pivoting in female soccer players

Most biomechanical studies into changing direction focus on final contact (FC), whilst limited research has examined penultimate contact (PEN). The aim of this study was to explore the kinematic and kinetic differences between PEN and FC of cutting and pivoting in 22 female soccer players (mean ± SD; age: 21 ± 3.1 years, height: 1.68 ± 0.07 m, mass: 58.9 ± 7.3 kg). Furthermore, the study investigated whether horizontal force–time characteristics during PEN were related to peak knee abduction moments during FC. Three dimensional motion analyses of cutting and pivoting on the right leg were performed using Qualysis ‘Proreflex’ infrared cameras (240 Hz). Ground reaction forces (GRF) were collected from two AMTI force platforms (1200 Hz) to examine PEN and FC. Both manoeuvres involved significantly (P < 0.05) greater knee joint flexion angles, peak horizontal GRF, but lower average horizontal GRF during PEN compared to FC. Average horizontal GRF during PEN (R = −0.569, R² = 32%, P = 0.006) and average horizontal GRF ratio (R = 0.466, R² = 22%, P = 0.029) were significantly related to peak knee abduction moments during the FC of cutting and pivoting, respectively. The results indicate PEN during pre-planned changing direction helps reduce loading on the turning leg where there is greater risk of injuries to knee ligaments.     

Alternative methods of normalising EMG during cycling

We evaluated possible methods of normalisation for EMG measured during cycling. The MVC method, Sprint method and 70% Peak Power Output Method were investigated and their repeatability, reliability and sensitivity to change in workload were compared.Thirteen cyclists performed the same experimental protocol on three separate occasions. Each day, subjects firstly performed MVCs, followed by a 10 s maximal sprint on a cycle ergometer. Subjects then performed a Peak Power Output (PPO) test until exhaustion. After which they cycled at 70% of PPO for 5 min at 90 rpm. Results indicated that normalising EMG data to 70% PPO is more repeatable, the intra-class correlation (ICC) of 70% PPO (0.87) was significantly higher than for MVC (0.66) (p = 0.03) and 10 s sprint (0.65) (p = 0.04). The 70% PPO method also demonstrated the least intra-subject variability for five out of the six muscles. The Sprint and 70% PPO method highlighted greater sensitivity to changes in muscle activity than the MVC method. The MVC method showed the highest intra-subject variability for most muscles except VM.The data suggests that normalising EMG to dynamic methods is the most appropriate for examining muscle activity during cycling over different days and for once-off measurements.     

Soft tissue artifact compensation by linear 3D interpolation and approximation methods

Several compensation methods estimate bone pose from a cluster of skin-mounted makers, each influenced by soft tissue artifact (STA). In this study, linear 3D interpolation and approximation methods (affine mapping, Kriging and radial basis function (RBF)) and the conventional singular value decomposition (SVD) method were examined to determine their suitability for STA compensation. The ability of these four methods to estimate knee angles and displacements was compared using simulated gait data with and without added STA. The knee angle and the displacement estimates of all four methods were similar with root-mean-square errors (RMSEs) near 1.5° and 4 mm, respectively. The 3D interpolation and approximation methods were more complicated to implement than the conventional SVD method. However, these non-standard methods provided additional geometric (homothety, stretch) and time functions that model the deformation of the cluster of markers. This additional information may be useful to model and compensate the STA.     

Assessment of the lower limb soft tissue artefact at marker-cluster level with a high-density marker set during walking

The estimation of joint kinematics from skin markers is hindered by the soft tissue artefact (STA), a well-known phenomenon although not fully characterized. While most assessments of the STA have been performed based on the individual skin markers displacements, recent assessments were based on the marker-cluster geometrical transformations using, e.g., principal component or modal analysis. However, these marker-clusters were generally made of 4–6 markers and the current findings on the STA could have been biased by the limited number of skin makers analysed. The objective of the present study was therefore to confirm them with a high-density marker set, i.e. 40 markers placed on the segments.A larger number of modes than found in the literature was required to describe the STA. Nevertheless, translations and rotations of the marker-cluster remained the main STA modes, archetypally the translation along the proximal-distal and anterior-posterior axes for the shank and the translation along the proximal-distal axis and the rotation about the medial-lateral axis for the thigh. High correlations were also found between the knee flexion angle and the amplitude of these modes for the thigh whereas moderate ones were found for the shank.These findings support the current re-orientation of the STA compensation methods, from bone pose estimators which typically address the non-rigid components of the marker-cluster to kinematic-driven rigid-component STA models.