Upper extremity inverse dynamics model for crutch-assisted gait assessment

Current inverse dynamics models of the upper extremity (UE) are limited for the measurement of Lofstrand crutch-assisted gait. The objective of this study is to develop, validate, and demonstrate a three-dimensional (3-D) UE motion assessment system to quantify crutch-assisted gait in children. We propose a novel 3-D dynamic model of the UEs and crutches for quantification of joint motions, forces, and moments during Lofstrand crutch-assisted gait. The model is composed of the upper body (i.e., thorax, upper arms, forearms, and hands) and Lofstrand crutches to determine joint dynamics of the thorax, shoulders, elbows, wrists, and crutches. The model was evaluated and applied to a pediatric subject with myelomeningocele (MM) to demonstrate its effectiveness in the characterization of crutch gait during multiple walking patterns. The model quantified UE dynamics during reciprocal and swing-through crutch-assisted gait patterns. Joint motions and forces were greater during swing-through gait than reciprocal gait. The model is suitable for further application to pediatric crutch-user populations. This study has potential for improving the understanding of the biomechanics of crutch-assisted gait and may impact clinical intervention strategies and therapeutic planning of ambulation.     

Effect of Trunk Sagittal Attitude on Shoulder,Thorax and Pelvis Three-Dimensional Kinematics in Able-Bodied Subjects during Gait

It has been shown that an original attitude in forward or backward inclination of the trunk is maintained at gait initiation and during locomotion, and that this affects lower limb loading patterns. However, no studies have shown the extent to which shoulder, thorax and pelvis three-dimensional kinematics are modified during gait due to this sagittal inclination attitude. Thirty young healthy volunteers were analyzed during level walking with video-based motion analysis. Reflecting markers were mounted on anatomical landmarks to form a two-marker shoulder line segment, and a four-marker thorax and pelvis segments. Absolute and relative spatial rotations were calculated, for a total of 11 degrees of freedom. The subjects were divided into two groups of 15 according to the median of mean thorax inclination angle over the gait cycle. Preliminary MANOVA analysis assessed whether gender was an independent variable. Then two-factor nested ANOVA was used to test the possible effect of thorax inclination on body segments, planes of motion and gait periods, separately. There was no significant difference in all anthropometric and spatio-temporal parameters between the two groups, except for subject mass. The three-dimensional kinematics of the thorax and pelvis were not affected by gender. Nested ANOVA revealed group effect in all segment rotations apart those at the pelvis, in the sagittal and frontal planes, and at the push-off. Attitudes in sagittal thorax inclination altered trunk segments kinematics during gait. Subjects with a backward thorax showed less thorax-to-pelvis motion, but more shoulder-to-thorax and thorax-to-laboratory motion, less motion in flexion/extension and in lateral bending, and also less motion during push-off. This contributes to the understanding of forward propulsion and sideways load transfer mechanisms, fundamental for the maintenance of balance and the risk of falling.     

Mechanical coupling between transverse plane pelvis and thorax rotations during gait is higher in people with low back pain

This study investigated whether people with low back pain (LBP) reduce variability of movement between the pelvis and thorax (trunk) in the transverse plane during gait at different speeds compared to healthy controls. Thirteen people with chronic LBP and twelve healthy controls walked on a treadmill at speeds from 0.5 to 1.72 m/s, with increments of 0.11 m/s. Step-to-step variability of the trunk, pelvis, and thorax rotations were calculated. Step-to-step deviations of pelvis and thorax rotations from the average pattern (residual rotations) were correlated to each other, and the linear regression coefficients between these deviations calculated. Spectral analysis was used to determine the frequencies of the residual rotations, to infer the relation of reduced trunk variability to trunk stiffness and/or damping. Variability of trunk motion (thorax relative to pelvis) was lower (P=0.02), covariance between the residual rotations of pelvis and thorax motions was higher (P=0.03), and the linear regression coefficients were closer to 1 (P=0.05) in the LBP group. Most power of segmental residual rotations was below stride frequency (~1 Hz). In this frequency range, trunk residual rotations had less power than pelvis or thorax residual rotations. These data show that people with LBP had lower variability of trunk rotations, as a result of the coupling of deviations of residual rotations in one segment to deviations of a similar shape (correlation) and amplitude (regression coefficient) in the other segment. These results support the argument that people with LBP adopt a protective movement strategy, possibly by increased trunk stiffness.     

A point cluster method for in vivo motion analysis: applied to a study of knee kinematics.

A new method for deriving limb segment motion from markers placed on the skin is described. The method provides a basis for determining the artifact associated with nonrigid body movement of points placed on the skin. The method is based on a cluster of points uniformly distributed on the limb segment. Each point is assigned an arbitrary mass. The center of mass and the inertia tensor of this cluster of points are calculated. The eigenvalues and eigenvectors of the inertia tensor are used to define a coordinate system in the cluster as well as to provide a basis for evaluating non-rigid body movement. The eigenvalues of the inertia tensor remain invariant if the segment is behaving as a rigid body, thereby providing a basis for determining variations for nonrigid body movement. The method was tested in a simulation model where systematic and random errors were introduced into a fixed cluster of points. The simulation demonstrated that the error due to nonrigid body movement could be substantially reduced. The method was also evaluated in a group of ten normal subjects during walking. The results for knee rotation and translation obtained from the point cluster method compared favorably to results previously obtained from normal subjects with intra-cortical pins placed into the femur and tibia. The resulting methodology described in this paper provides a unique approach to the measurement of in vivo motion using skin-based marker systems.     

Axial pelvis range of motion affects thorax-pelvis timing during gait

During gait, patients with pelvic girdle pain and low back pain demonstrate an altered phase relationship between axial thorax and pelvis rotations (thorax-pelvis relative phase). This could be the result of an increase in axial pelvis range of motion (ROM) which has been observed in these patients as well. To establish this relationship, we investigated if altered axial pelvis ROM during gait affects thorax-pelvis relative phase in 12 healthy subjects. These subjects walked on a treadmill and received real-time feedback on axial pelvis rotations. Subjects were asked to (1) walk normal, and walk with (2) decreased and (3) increased pelvis ROM. Gait speed and stride frequency were matched between trials. Subjects were able to increase pelvis ROM to a large extent, but the reduction in pelvis ROM was relatively small. Walking with large pelvis ROM resulted in a change in thorax-pelvis relative phase similar to that in pelvic girdle pain and low back pain. A forward dynamic model was used to predict the effect of manipulation of pelvis ROM on timing of thorax rotations independent of apparent axial trunk stiffness and arm swing amplitude (which can both affect thorax-pelvis relative phase). The model predicted a similar, even larger, effect of large axial pelvis ROM on thorax-pelvis relative phase, as observed experimentally. We conclude that walking with actively increased ROM of axial pelvis rotations in healthy subjects is associated with a shift in thorax-pelvis relative phase, similar to observations in patients with pelvic girdle pain and low back pain.     

Novel approach to ambulatory assessment of human segmental orientation on a wearable sensor system

A new method using a double-sensor difference based algorithm for analyzing human segment rotational angles in two directions for segmental orientation analysis in the three-dimensional (3D) space was presented. A wearable sensor system based only on triaxial accelerometers was developed to obtain the pitch and yaw angles of thigh segment with an accelerometer approximating translational acceleration of the hip joint and two accelerometers measuring the actual accelerations on the thigh. To evaluate the method, the system was first tested on a 2° of freedom mechanical arm assembled out of rigid segments and encoders. Then, to estimate the human segmental orientation, the wearable sensor system was tested on the thighs of eight volunteer subjects, who walked in a straight forward line in the work space of an optical motion analysis system at three self-selected speeds: slow, normal and fast. In the experiment, the subject was assumed to walk in a straight forward way with very little trunk sway, skin artifacts and no significant internal/external rotation of the leg. The root mean square (RMS) errors of the thigh segment orientation measurement were between 2.4° and 4.9° during normal gait that had a 45° flexion/extension range of motion. Measurement error was observed to increase with increasing walking speed probably because of the result of increased trunk sway, axial rotation and skin artifacts. The results show that, without integration and switching between different sensors, using only one kind of sensor, the wearable sensor system is suitable for ambulatory analysis of normal gait orientation of thigh and shank in two directions of the segment-fixed local coordinate system in 3D space. It can then be applied to assess spatio-temporal gait parameters and monitoring the gait function of patients in clinical settings.     

Hemiplegic gait: a kinematic analysis using walking speed as a basis.

The kinematics of treadmill ambulation of stroke patients (N = 9) and healthy subjects (N = 4) was studied at a wide range of different velocities (i.e. 0.25-1.5 m s-1), with a focus on the transverse rotations of the trunk. Video recordings revealed, for both stroke patients and healthy subjects, similar relations between walking speed and stride length as well as stride frequency. The phase difference between pelvic and thoracic rotations (i.e. trunk rotation) and the total range of trunk rotation were almost linearly related to the walking speed. Healthy subjects showed a marked increase in pelvic rotation from 1 to 1.5 m s-1. Using dimensional analysis in a comparison between stroke patients and healthy subjects, invariances in the coordination of gait were found for stride length, stride frequency, pelvic rotation, and trunk rotation. Constant relations were obtained between, on the one hand, dimensionless velocity and, on the other, dimensionless stride length as well as stride frequency. Transitions were found between the velocities 0.75 and 1 m s-1 for dimensionless pelvic rotation and trunk rotation, indicating that, from this velocity range onwards, pelvic swing lengthens the stride: rotations of pelvis, thorax and trunk become tightly coordinated. On the basis of the dimensionless stride length, stride frequency, pelvic rotation and trunk rotation, deficits in the gait of stroke patients could be quantified. It is concluded that walking speed is an important control parameter, which should be used as a basic variable in the evaluation of the gait of stroke patients.     

Artificial neural network model for the generation of muscle activation patterns for human locomotion.

Skilled locomotor behaviour requires information from various levels within the central nervous system (CNS). Mathematical models have permitted researchers to simulate various mechanisms in order to understand the organization of the locomotor control system. While it is difficult to adequately characterize the numerous inputs to the locomotor control system, an alternative strategy may be to use a kinematic movement plan to represent the complex inputs to the locomotor control system based on the possibility that the CNS may plan movements at a kinematic level. We propose the use of artificial neural network (ANN) models to represent the transformation of a kinematic plan into the necessary motor patterns. Essentially, kinematic representation of the actual limb movement was used as the input to an ANN model which generated the EMG activity of 8 muscles of the lower limb and trunk. Data from a wide variety of gait conditions was necessary to develop a robust model that could accommodate various environmental conditions encountered during everyday activity. A total of 120 walking strides representing normal walking and ten conditions where the normal gait was modified in terms of cadence, stride length, stance width or required foot clearance. The final network was assessed on its ability to predict the EMG activity on individual walking trials as well as its ability to represent the general activation pattern of a particular gait condition. The predicted EMG patterns closely matched those recorded experimentally, exhibiting the appropriate magnitude and temporal phasing required for each modification. Only 2 of the 96 muscle/gait conditions had RMS errors above 0.10, only 5 muscle/gait conditions exhibited correlations below 0.80 (most were above 0.90) and only 25 muscle/gait conditions deviated outside the normal range of muscle activity for more than 25% of the gait cycle. These results indicate the ability of single network ANNs to represent the transformation between a kinematic movement plan and the necessary muscle activations for normal steady state locomotion but they were also able to generate muscle activation patterns for conditions requiring changes in walking speed, foot placement and foot clearance. The abilities of this type of network have implications towards both the fundamental understanding of the control of locomotion and practical realizations of artificial control systems for use in rehabilitation medicine.     

Rotational and translational movement features of the pelvis and thorax during adult human locomotion

The kinematics of the pelvis and thorax are important in gait studies since their movement patterns are closely related to gait efficiency and 'smoothness' of locomotion. The purpose of this study was to identify features of normal gait patterns for later comparisons with pathological and developmental gait patterns. A two camera SELSPOT system interfaced with an HP1000 minicomputer was used to obtain three-dimensional kinematic/temporal data for the pelvis and thorax. Data from treadmill walking of eight adults were used for within subject (at different speeds) analyses. The analyses revealed a very complex pattern with a set of breakpoints which was consistent over all subjects. Some features were invariant over a range of walking speeds although the total range of motion changed considerably.     

Contributions of individual muscles to the sagittal- and frontal-plane angular accelerations of the trunk in walking

This study was conducted to analyze the unimpaired control of the trunk during walking. Studying the unimpaired control of the trunk reveals characteristics of good control. These characteristics can be pursued in the rehabilitation of impaired control. Impaired control of the trunk during walking is associated with aging and many movement disorders. This is a concern as it is considered to increase fall risk. Muscles that contribute to the trunk control in normal walking may also contribute to it under perturbation circumstances, attempting to prevent an impending fall. Knowledge of such muscles can be used to rehabilitate impaired control of the trunk. Here, angular accelerations of the trunk induced by individual muscles, in the sagittal and frontal planes, were calculated using 3D muscle-driven simulations of seven young healthy subjects walking at free speed. Analysis of the simulations demonstrated that the abdominal and back muscles displayed large contributions throughout the gait cycle both in the sagittal and frontal planes. Proximal lower-limb muscles contributed more than distal muscles in the sagittal plane, while both proximal and distal muscles showed large contributions in the frontal plane. Along with the stance-limb muscles, the swing-limb muscles also exhibited considerable contribution. The gluteus medius was found to be an important individual frontal-plane control muscle; enhancing its function in pathologies could ameliorate gait by attenuating trunk sway. In addition, since gravity appreciably accelerated the trunk in the frontal plane, it may engender excessive trunk sway in pathologies.     

Kinematic performance of a six degree-of-freedom hand model (6DHand) for use in occupational biomechanics

Upper extremity musculoskeletal disorders represent an important health issue across all industry sectors; as such, the need exists to develop models of the hand that provide comprehensive biomechanics during occupational tasks. Previous optical motion capture studies used a single marker on the dorsal aspect of finger joints, allowing calculation of one and two degree-of-freedom (DOF) joint angles; additional algorithms were needed to define joint centers and the palmar surface of fingers. We developed a 6DOF model (6DHand) to obtain unconstrained kinematics of finger segments, modeled as frusta of right circular cones that approximate the palmar surface. To evaluate kinematic performance, twenty subjects gripped a cylindrical handle as a surrogate for a powered hand tool. We hypothesized that accessory motions (metacarpophalangeal pronation/supination; proximal and distal interphalangeal radial/ulnar deviation and pronation/supination; all joint translations) would be small (less than 5° rotations, less than 2mm translations) if segment anatomical reference frames were aligned correctly, and skin movement artifacts were negligible. For the gripping task, 93 of 112 accessory motions were small by our definition, suggesting this 6DOF approach appropriately models joints of the fingers. Metacarpophalangeal supination was larger than expected (approximately 10°), and may be adjusted through local reference frame optimization procedures previously developed for knee kinematics in gait analysis. Proximal translations at the metacarpophalangeal joints (approximately 10mm) were explained by skin movement across the metacarpals, but would not corrupt inverse dynamics calculated for the phalanges. We assessed performance in this study; a more rigorous validation would likely require medical imaging.     

Evidence for a role of antagonistic cocontraction in controlling trunk stiffness during lifting

Activity of the abdominal muscles during symmetric lifting has been a consistent finding in many studies. It has been hypothesized that this antagonistic coactivation increases trunk stiffness to provide stability to the spine. To test this, we investigated whether abdominal activity in lifting is increased in response to destabilizing conditions.

Ten healthy male subjects lifted 35 l containers containing 15 l of water (unstable condition), or ice (stable condition). 3D-kinematics, ground reaction forces, and EMG of selected trunk muscles were recorded. Euler angles of the thorax relative to the pelvis were determined. Inverse dynamics was used to calculate moments about L5S1. Averaged normalized abdominal EMG activity was calculated to express coactivation and an EMG-driven trunk muscle model was used to estimate the flexor moment produced by these muscles and to estimate the L5S1 compression force.

Abdominal coactivation was significantly higher when lifting the unstable load. This coincided with significant increases in estimated moments produced by the antagonist muscles and in estimated compression forces on the L5S1 disc, except at the instant of the peak moment about L5S1. The lifting style was not affected by load instability as evidenced by the absence of effects on moments about L5S1 and angles of the thorax relative to the pelvis. The data support the interpretation of abdominal cocontraction during lifting as subserving spinal stability. An alternative function of the increased trunk stiffness due to cocontraction might be to achieve more precise control over the trajectory of lifted weight in order to avoid sloshing of the water mass in the box and the consequent perturbations.     

Cluster-based upper body marker models for three-dimensional kinematic analysis: Comparison with an anatomical model and reliability analysis

Quantifying angular joint kinematics of the upper body is a useful method for assessing upper limb function. Joint angles are commonly obtained via motion capture, tracking markers placed on anatomical landmarks. This method is associated with limitations including administrative burden, soft tissue artifacts, and intra- and inter-tester variability. An alternative method involves the tracking of rigid marker clusters affixed to body segments, calibrated relative to anatomical landmarks or known joint angles. The accuracy and reliability of applying this cluster method to the upper body has, however, not been comprehensively explored. Our objective was to compare three different upper body cluster models with an anatomical model, with respect to joint angles and reliability. Non-disabled participants performed two standardized functional upper limb tasks with anatomical and cluster markers applied concurrently. Joint angle curves obtained via the marker clusters with three different calibration methods were compared to those from an anatomical model, and between-session reliability was assessed for all models. The cluster models produced joint angle curves which were comparable to and highly correlated with those from the anatomical model, but exhibited notable offsets and differences in sensitivity for some degrees of freedom. Between-session reliability was comparable between all models, and good for most degrees of freedom. Overall, the cluster models produced reliable joint angles that, however, cannot be used interchangeably with anatomical model outputs to calculate kinematic metrics. Cluster models appear to be an adequate, and possibly advantageous alternative to anatomical models when the objective is to assess trends in movement behavior.     

Abdominal muscle activation changes if the purpose is to control pelvis motion or thorax motion

The aim of this study was to compare trunk muscular recruitment and lumbar spine kinematics when motion was constrained to either the thorax or the pelvis. Nine healthy women performed four upright standing planar movements (rotations, anterior–posterior translations, medial–lateral translations, and horizontal circles) while constraining pelvis motion and moving the thorax or moving the pelvis while minimizing thorax motion, and four isometric trunk exercises (conventional curl-up, reverse curl-up, cross curl-up, and reverse cross curl-up). Surface EMG (upper and lower rectus abdominis, lateral and medial aspects of external oblique, internal oblique, and latissimus dorsi) and 3D lumbar displacements were recorded. Pelvis movements produced higher EMG amplitudes of the oblique abdominals than thorax motions in most trials, and larger lumbar displacements in the medial–lateral translations and horizontal circles. Conversely, thorax movements produced larger rotational lumbar displacement than pelvis motions during rotations and higher EMG amplitudes for latissimus dorsi during rotations and anterior–posterior translations and for lower rectus abdominis during the crossed curl-ups. Thus, different neuromuscular compartments appear when the objective changes from pelvis to thorax motion. This would suggest that both movement patterns should be considered when planning spine stabilization programs, to optimize exercises for the movement and muscle activations desired.     

Instantaneous helical axis estimation from 3-D video data in neck kinematics for whiplash diagnostics

To date, the diagnosis of whiplash injuries has been very difficult and largely based on subjective, clinical assessment. The work by Winters and Peles Multiple Muscle Systems—Biomechanics and Movement Organization. Springer, New York (1990) suggests that the use of finite helical axes (FHAs) in the neck may provide an objective assessment tool for neck mobility. Thus, the position of the FHA describing head-trunk motion may allow discrimination between normal and pathological cases such as decreased mobility in particular cervical joints. For noisy, unsmoothed data, the FHAs must be taken over rather large angular intervals if the FHAs are to be reconstructed with sufficient accuracy; in the Winters and Peles study, these intervals were approximately 10°.

In order to study the movements' microstructure, the present investigation uses instantaneous helical axes (IHAs) estimated from low-pass smoothed video data. Here, the small-step noise sensitivity of the FHA no longer applies, and proper low-pass filtering allows estimation of the IHA even for small rotation velocity ω of the moving neck. For marker clusters mounted on the head and trunk, technical system validation showed that the IHAs direction dispersions were on the order of one degree, while their position dispersions were on the order of 1 mm, for low-pass cut-off frequencies of a few Hz (the dispersions were calculated from ω-weighted errors, in order to account for the adverse effects of vanishing ω).

Various simple, planar models relating the instantaneous 2-D centre of rotation with the geometry and kinematics of a multi-joint neck model are derived, in order to gauge the utility of the FHA and IHa approaches.

Some preliminary results on asymptomatic and pathological subjects are provided, in terms of the ‘ruled surface’ formed by sampled IHAs and of their piercing points through the mid-sagittal plane during a prescribed flexion-extension movement of the neck.     

A surface electromyography based objective method to identify patients with nonspecific chronic low back pain,presenting a flexion related movement control impairment

Movement control impairments (MCI) are often present in patients with non-specific chronic low back pain (NS-CLBP). Therefore, movement control exercises are widely used to rehabilitate patients. However, the objective assessment remains difficult.The purpose of this study was to develop a statistical model, based on logistic regression analysis, to differentiate patients with NS-CLBP presenting a flexion-related MCI from healthy subjects. This model is based on trunk muscle activation patterns measured by surface electromyography (sEMG), during movement control exercises.Sixty-three healthy male subjects and 36 male patients with a flexion-related MCI participated in this study. Muscle activity of the internal obliques, the external obliques, the lumbar multifidus and the thoracic part of the iliocostalis was registered. Ratios of deep stabilizing to superficial torque producing muscle activity were calculated to examine trunk muscle recruitment patterns during 6 different exercises. Logistic regression analyses were performed (1) to define the ratios and exercises that were most discriminating between patients and non-patients, (2) to make a predictive model. K-Fold cross-validation was used to assess the performance of the predictive model.This study demonstrated that sEMG trunk muscle recruitment patterns during movement control tests, allows differentiating NSCLBP patients with a flexion-related MCI from healthy subjects.     

Implications of increased medio-lateral trunk sway for ambulatory mechanics

The purposes of this study was to test a mechanism to reduce the knee adduction moment by testing the hypothesis that increased medio-lateral trunk sway can reduce the knee adduction moment during ambulation in healthy subjects, and to examine the possibility that increasing medio-lateral trunk sway can produce similar potentially adverse secondary gait changes previously associated with reduced knee adduction moments in patients with knee osteoarthritis. Nineteen healthy adults performed walking trials with normal and increased medio-lateral trunk sway at a self-selected normal walking speed. Standard gait analysis was used to calculate three-dimensional lower extremity joint kinematics and kinetics. Knee and hip adduction moments were lower (-65.0% and -57.1%, respectively) for the increased medio-lateral trunk sway trials than for the normal trunk sway trials. Knee flexion angle at heel-strike was 3 degrees higher for the increased than for the normal trunk sway trials. Knee and hip abduction moments were higher for the increased medio-lateral trunk sway trials, and none of the other variables differed between the two conditions. Walking with increased medio-lateral trunk sway substantially reduces the knee adduction moment during walking in healthy subjects without some of the adverse secondary effects such as increased axial loading rates at the major joints of the lower extremity. This result supports the potential of using gait retraining for walking with increased medio-lateral trunk sway as treatment for patients with degenerative joint disease such as medial compartment knee osteoarthritis.     

The effect of attentional focus on local dynamic stability during a repetitive spine flexion task

The association between low back pain and spine movement control suggests that it is important to reliably quantify movement behavior. One method to characterize spine movement behavior is to measure the local dynamic stability (LDS) of spine movement during a repetitive flexion task in which a participant is asked to touch multiple targets repetitively. Within the literature, it has been well established that an individual’s focus of attention (FOA) can modulate their neuromuscular control and affect task performance. The goal of this project was to examine the unknown effect of FOA on LDS measurements and timing error during a repetitive spine flexion task that is commonly used to assess movement control. Fourteen healthy adults (7 male) were instructed to touch two targets (shoulder height and knee height) to the beat of a metronome (4 s/cycle) for 35 consecutive cycles. They completed this task under internal (focus on trunk movement) and external (focus on targets) FOA conditions. Motion capture data of the trunk and sacrum were collected at 120 Hz. The lumbar spine angle was defined as the orientation of the trunk relative to the pelvis. The local divergence exponent (λ_max) was calculated from the sum of squares of the 3-dimensional spine angle. Timing error was calculated as the time difference between target touches and metronome beats. Changing an individual’s FOA had no effect on λ_max calculations or timing error. Although clear task instructions are important, it is not essential to control for FOA during this movement assessment protocol.     

Changes in axial stiffness of the trunk as a function of walking speed

Research suggests that abnormal coordination patterns between the thorax and pelvis in the transverse plane observed in patients with Parkinson's disease and the elderly might be due to alteration in axial trunk stiffness. The purpose of this study was to develop a tool to estimate axial trunk stiffness during walking and to investigate its functional role. Fourteen healthy young subjects participated in this study. They were instructed to walk on the treadmill and kinematic data was collected by 3D motion analysis system. Axial trunk stiffness was estimated from the angular displacement between trunk segments and the amount of torque around vertical axis of rotation. The torque due to arm swing cancelled out the torque due to the axial trunk stiffness during walking and the thoracic rotation was of low amplitude independent of changes in walking speeds within the range used in this study (0.85-1.52 m/s). Estimated axial trunk stiffness increased with increasing walking speed. Functionally, the suppression of axial rotation of thorax may have a positive influence on head stability as well as allowing recoil between trunk segments. Furthermore, the increased stiffness at increased walking speed would facilitate the higher frequency rotation of the trunk in the transverse plane required at the higher walking speeds.