Validation of net joint loads calculated by inverse dynamics in case of complex movements: Application to balance recovery movements

QUANTIFICATION OF MASS AND CENTER-OF-MASS OF HEALTHY AND AMPUTATED SEGMENTS AS WELL AS FULL-BODY CENTER-OF-MASS OF AMPUTEES

Quantification of segment-inertial uniqueness can provide a relevant foundation for motion analysis, biomechanical modeling and human motor skill optimization of both normal and amputated athletes. It is known that previous studies focused on quantifying Body Segment Inertial Parameters (BSIP) of non-amputated people in order to establish regression equations for calculating BSIPs. Until now, no anthropometrical study existed on quantifying BSIPs such as mass and center of mass (COM) of both non-amputated segment (NAS) and partially-amputated segment (PAS) of amputees. This study aims to fill the gap. A quantification method derived from Damavandi approach was applied to determine the mass and COM of PAS as well as full-body COM. For validating the reliability of this method, the calculated values were compared to the values measured by balance board test. Further, two anthropometrical approaches (i.e. Zheng and Zatsiorsky) for normal subjects were tested for their validity to estimatfe the mass and COM of NASs of amputees. The results reveal that Damavandi approach can also be used for reliable quantifying of mass and COM of PAS and Zatsiorsky’s approach is more reliable to quantify NAS masses and full-body COM of amputees, therefore, Damavandi approach and Zatsiorsky’s regression model are more suitable for motion analysis, biomechanical modeling and motor skill optimization of amputees.

Quantification of multi-segment trunk kinetics during multi-directional trunk bending

BACKGROUND

Motion assessment of the body's head-arms-trunk (HAT) using linked-segment models, along with an inverse dynamics approach, can enable in vivo estimations of inter-vertebral moments. However, this mathematical approach is prone to experimental errors because of inaccuracies in (i) kinematic measurements associated with soft tissue artifacts and (ii) estimating individual-specific body segment parameters (BSPs). The inaccuracy of the BSPs is particularly challenging for the multi-segment HAT due to high inter-participant variability in the HAT's BSPs and no study currently exists that can provide a less erroneous estimation of the joint moments along the spinal column.

RESEARCH QUESTION

This study characterized three-dimensional (3D) inter-segmental moments in a multi-segment HAT model during multi-directional trunk-bending, after minimizing the experimental errors.

METHOD

Eleven healthy individuals participated in a multi-directional trunk-bending experiment in five directions with three speeds. A seven-segment HAT model was reconstructed for each participant, and its motion was recorded. After compensating for experimental errors due to soft tissue artifacts, and using optimized individual-specific BSPs, and center of pressure offsets, the inter-segmental moments were calculated via inverse dynamics.

RESULTS

Our results show a significant effect of the inter-segmental level and trunk-bending directions on the obtained moments. Compensating for soft tissue artifacts contributed significantly to reducing errors. Our results indicate complex, task-specific patterns of the 3D moments, with high inter-participant variability at different inter-segmental levels, which cannot be studied using single-segment models or without error compensation.

SIGNIFICANCE

Interpretation of inter-segmental moments after compensation of experimental errors is important for clinical evaluations and developing injury prevention and rehabilitation strategies.

Optimal Estimation of Anthropometric Parameters for Quantifying Multisegment Trunk Kinetics

Kinetics assessment of the human head-arms-trunk (HAT) complex via a multisegment model is a useful tool for objective clinical evaluation of several pathological conditions. Inaccuracies in body segment parameters (BSPs) are a major source of uncertainty in the estimation of the joint moments associated with the multisegment HAT. Given the large intersubject variability, there is currently no comprehensive database for the estimation of BSPs for the HAT. We propose a nonlinear, multistep, optimization-based, noninvasive method for estimating individual-specific BSPs and calculating joint moments in a multisegment HAT model. Eleven nondisabled individuals participated in a trunk-bending experiment and their body motion was recorded using cameras and a force plate. A seven-segment model of the HAT was reconstructed for each participant. An initial guess of the BSPs was obtained by individual-specific scaling of the BSPs calculated from the male visible human (MVH) images. The intersegmental moments were calculated using both bottom-up and top-down inverse dynamics approaches. Our proposed method adjusted the scaled BSPs and center of pressure (COP) offsets to estimate optimal individual-specific BSPs that minimize the difference between the moments obtained by top-down and bottom-up inverse dynamics approaches. Our results indicate that the proposed method reduced the error in the net joint moment estimation (defined as the difference between the net joint moment calculated via bottom-up and top-down approaches) by 79.3% (median among participants). Our proposed method enables an optimized estimation of individual-specific BSPs and, consequently, a less erroneous assessment of the three-dimensional (3D) kinetics of a multisegment HAT model.

Using a marker-less method for estimating L5/S1 moments during symmetrical lifting

The aim of this study is to analyze the validity of a computer vision-based method to estimate 3D L5/S1 joint moment during symmetrical lifting. An important criterion to identify the non-ergonomic lifting task is the value of net moment at L5/S1 joint. This is usually calculated in a laboratory environment which is not practical for on-site biomechanical analysis. The validity of the proposed method, was assessed externally by comparing the results with a lab-based reference method and internally by comparing the estimated L5/S1 joint moments from top-down model and bottom-up model. It was shown that no significant differences in peak and mean moments between the two methods and intra-class correlation coefficients revealed excellent reliability of the proposed method (>0.91). The proposed method provides a reliable tool for assessment of lower back loads during occupational lifting and can be an alternative when the use of marker-based motion tracking systems is not possible.

A simplified marker set to define the center of mass for stability analysis in dynamic situations

The extrapolated center of mass (XCoM), a valuable tool to assess balance stability, involves defining the whole body center of mass (CoMWB). However, accurate three-dimensional estimation of the CoMWB is time consuming, a severe limitation in certain applications. In this study, twenty-four subjects (young and elderly, male and female) performed three different balance tasks: quiet standing, gait and balance recovery. Three different models, based on a segmental method, were used to estimate the three-dimensional CoMWB absolute position during these movements: a reference model based on 38 markers, a simplified 13-marker model and a single marker (sacral) model. CoMWB and XCoM estimations from the proposed simplified model came closer to the reference model than estimations from the sacral marker model. It remained accurate for dynamic tasks, where the sacral marker model proved inappropriate. The simplified model proposed here yields accurate three-dimensional estimation of both the CoMWB and the XCoM with a limited number of markers. Importantly, using this model would reduce the experimental and post-processing times for future balance studies assessing dynamic stability in humans.

A 3D analysis of the joint torques developed during driver's ingress-egress motion

Providing an easy ingress-egress (I/E) movement remains a challenge for car designers. I/E has been largely studied in kinematics, but not in dynamics. This study proposes: (1) to evaluate and describe the motor torques developed in the lower limbs and lumbar joints during I/E motions and (2) to analyse the influence of the car geometry and subject anthropometry. An experiment was performed to observe 15 subjects of three anthropometrical groups getting in and out of a car mock-up simulating three different vehicle configurations. Motor torques were extracted using an inverse dynamics analysis. Both ingress and egress motions were primarily characterised by large torques. Overall, the taller a subject and the lower the seat of the vehicle were, the larger the peak torques were. Moreover, peak torques were higher for egress than ingress. These results are discussed in regard to the current knowledge on I/E ergonomics.

PRACTITIONER SUMMARY

Car ingress–egress (I/E) is an ergonomics challenge. Little is known about the physical efforts developed in this motion. Developed motor torques were experimentally assessed for three anthropometrical groups and vehicle configurations. Results obtained were discussed in regard to the current knowledge on I/E ergonomics.

The effect of load on torques in point-to-point arm movements: a 3D model

A dynamic, 3-dimensional model was developed to simulate slightly restricted (pronation-supination was not allowed) point-to-point movements of the upper limb under different external loads, which were modeled using 3 objects of distinct masses held in the hand. The model considered structural and biomechanical properties of the arm and measured coordinates of joint positions. The model predicted muscle torques generated by muscles and needed to produce the measured rotations in the shoulder and elbow joints. The effect of different object masses on torque profiles, magnitudes, and directions were studied. Correlation analysis has shown that torque profiles in the shoulder and elbow joints are load invariant. The shape of the torque magnitude-time curve is load invariant but it is scaled with the mass of the load. Objects with larger masses are associated with a lower deflection of the elbow torque with respect to the sagittal plane. Torque direction-time curve is load invariant scaled with the mass of the load. The authors propose that the load invariance of the torque magnitude-time curve and torque direction-time curve holds for object transporting arm movements not restricted to a plane.

Validity of the Top-Down Approach of Inverse Dynamics Analysis in Fast and Large Rotational Trunk Movements

This study investigated the validity of the top-down approach of inverse dynamics analysis in fast and large rotational movements of the trunk about three orthogonal axes of the pelvis for nine male collegiate students. The maximum angles of the upper trunk relative to the pelvis were approximately 47°, 49°, 32°, and 55° for lateral bending, flexion, extension, and axial rotation, respectively, with maximum angular velocities of 209°/s, 201°/s, 145°/s, and 288°/s, respectively. The pelvic moments about the axes during the movements were determined using the top-down and bottom-up approaches of inverse dynamics and compared between the two approaches. Three body segment inertial parameter sets were estimated using anthropometric data sets (Ae et al., Biomechanism 11, 1992; De Leva, J Biomech, 1996; Dumas et al., J Biomech, 2007). The root-mean-square errors of the moments and the absolute errors of the peaks of the moments were generally smaller than 10 N·m. The results suggest that the pelvic moment in motions involving fast and large trunk movements can be determined with a certain level of validity using the top-down approach in which the trunk is modeled as two or three rigid-link segments.

Evaluation of force plate-less estimation of the trajectory of the centre of pressure during gait. Comparison of two anthropometric models

The estimation of the trajectory of the centre of pressure during gait is possible without using force plate by modelling the whole body as a multi-segment chain. The kinematics and inertial parameters of each segment are necessary to determine the ground reaction forces and moments. The position of the centre of pressure can then be calculated at each frame of time. The objective of the study was to evaluate the accuracy of the estimation of the position of the centre of pressure during gait obtained without force plate data. Segment inertial parameters were determined using a proportional model and a geometric model. The modelling and calculations were computed for six volunteers and the estimated centres of pressure were compared to the centre of pressure measured using force plates considered as the gold standard. The estimation was better using the geometric model with an accuracy of 33 mm (4.1% of the peak-to-peak amplitude) on the longitudinal axis and 14.2 mm (12.9% of the peak-to-peak amplitude) on the lateral axis.

Whole body inverse dynamics over a complete gait cycle based only on measured kinematics

This paper presents a three-dimensional (3D) whole body multi-segment model for inverse dynamics analysis over a complete gait cycle, based only on measured kinematic data. The sequence of inverse dynamics calculations differs significantly from the conventional application of inverse dynamics using force plate data. A new validated "Smooth Transition Assumption" was used to solve the indeterminacy problem in the double support phase. Kinematic data is required for all major body segments and, hence, a whole body gait measurement protocol is presented. Finally, sensitivity analyses were conducted to evaluate the effects of digital filtering and body segment parameters on the accuracy of the prediction results. The model gave reasonably good estimates of sagittal plane ground forces and moment; however, the estimates in the other planes were less good, which we believe is largely due to their small magnitudes in comparison to the sagittal forces and moment. The errors observed are most likely caused by errors in the kinematic data resulting from skin movement artefact and by errors in the estimated body segment parameters. A digital filtering cut-off frequency of 4.5 Hz was found to produce the best results. It was also shown that errors in the mass properties of body segments can play a crucial role, with changes in properties sometimes having a disproportionate effect on the calculated ground reactions. The implication of these results is that, even when force plate data is available, the estimated joint forces are likely to suffer from similar errors.