Step-to-step variability in treadmill walking: influence of rhythmic auditory cueing

The Effects of Mobile Texting and Walking Speed on Gait Characteristics of Normal Weight and Obese Adults

The aim of this study was to examine how usage of mobile devices while simultaneously walking affects walking characteristics and texting performance of normal weight (NW) and obese (OB) individuals. Thirty-two OB (body mass index [BMI] = 34.4) and NW (BMI = 22.7) adults performed two 60-s walking trials at three-step frequencies along a rectangular walkway in two conditions (No Texting and Texting). Dual-task cost as well as unadjusted spatial and temporal gait characteristics were measured. Dual-task costs for the gait parameters as well as texting performance were not different between the groups, except for the lateral step variability showing a larger variability at the preferred frequency in OB individuals. For the unadjusted variables, OB exhibited longer double support, longer stance time, and lower turn velocity compared with NW. Overall, the results highlight a similar dual-task cost for the OB individuals compared with the NW individuals, in spite of underlying differences in gait mechanics.

Auditory and Visual External Cues Have Different Effects on Spatial but Similar Effects on Temporal Measures of Gait Variability

Walking synchronized to external cues is a common practice in clinical settings. Several research studies showed that this popular gait rehabilitation tool alters gait variability. There is also recent evidence which suggests that alterations in the temporal structure of the external cues could restore gait variability at healthy levels. It is unknown, however, if such alterations produce similar effects if the cueing modalities used are different; visual or auditory. The modality could affect gait variability differentially, since there is evidence that auditory cues mostly act in the temporal domain of gait, while visual cues act in the spatial domain of gait. This study investigated how synchronizing steps with visual and auditory cues that are presented with different temporal structures could affect gait variability during treadmill walking. Three different temporal structured stimuli were used, invariant, fractal and random, in both modalities. Stride times, length and speed were determined, and their fractal scaling (an indicator of complexity) and coefficient of variation (CV) were calculated. No differences were observed in the CV, regardless of the cueing modality and the temporal structure of the stimuli. In terms of the stride time's fractal scaling, we observed that the fractal stimulus induced higher values compared to random and invariant stimuli. The same was also observed in stride length, but only for the visual cueing modality. No differences were observed for stride speed. The selection of the cueing modality seems to be an important feature of gait rehabilitation. Visual cues are possibly a better choice due to the dependency on vision during walking. This is particularly evident during treadmill walking, a common practice in a clinical setting. Because of the treadmill effect on the temporal domain of gait, the use of auditory cues can be minimal, compared to visual cues.

Gait Recognition via Deep Learning of the Center-of-Pressure Trajectory

The fact that every human has a distinctive walking style has prompted a proposal to use gait recognition as an identification criterion. Using end-to-end learning, I investigated whether the center-of-pressure (COP) trajectory is sufficiently unique to identify a person with high certainty. Thirty-six adults walked for 30 min on a treadmill equipped with a force platform that continuously recorded the positions of the COP. The raw two-dimensional signals were sliced into segments of two gait cycles. A set of 20,250 segments from 30 subjects was used to configure and train convolutional neural networks (CNNs). The best CNN classified a separate set containing 2250 segments with an overall accuracy of 99.9%. A second set of 4500 segments from the six remaining subjects was then used for transfer learning. Several small subsamples of this set were selected randomly and used to fine tune the pretrained CNNs. Training with two segments per subject was sufficient to achieve 100% accuracy. The results suggest that every person produces a unique trajectory of underfoot pressures while walking and that CNNs can learn the distinctive features of these trajectories. By applying a pretrained CNN (transfer learning), a couple of strides seem enough to learn and identify new gaits. However, these promising results should be confirmed in a larger sample under realistic conditions.

Complexity of human walking: the attractor complexity index is sensitive to gait synchronization with visual and auditory cues

Background

During steady walking, gait parameters fluctuate from one stride to another with complex fractal patterns and long-range statistical persistence. When a metronome is used to pace the gait (sensorimotor synchronization), long-range persistence is replaced by stochastic oscillations (anti-persistence). Fractal patterns present in gait fluctuations are most often analyzed using detrended fluctuation analysis (DFA). This method requires the use of a discrete times series, such as intervals between consecutive heel strikes, as an input. Recently, a new nonlinear method, the attractor complexity index (ACI), has been shown to respond to complexity changes like DFA, while being computed from continuous signals without preliminary discretization. Its use would facilitate complexity analysis from a larger variety of gait measures, such as body accelerations. The aim of this study was to further compare DFA and ACI in a treadmill experiment that induced complexity changes through sensorimotor synchronization.

Methods

Thirty-six healthy adults walked 30 min on an instrumented treadmill under three conditions: no cueing, auditory cueing (metronome walking), and visual cueing (stepping stones). The center-of-pressure trajectory was discretized into time series of gait parameters, after which a complexity index (scaling exponent alpha) was computed via DFA. Continuous pressure position signals were used to compute the ACI. Correlations between ACI and DFA were then analyzed. The predictive ability of DFA and ACI to differentiate between cueing and no-cueing conditions was assessed using regularized logistic regressions and areas under the receiver operating characteristic curves (AUC).

Results

DFA and ACI were both significantly different among the cueing conditions. DFA and ACI were correlated (Pearson’s r = 0.86). Logistic regressions showed that DFA and ACI could differentiate between cueing/no cueing conditions with a high degree of confidence (AUC = 1.00 and 0.97, respectively).

Conclusion

Both DFA and ACI responded similarly to changes in cueing conditions and had comparable predictive power. This support the assumption that ACI could be used instead of DFA to assess the long-range complexity of continuous gait signals. However, future studies are needed to investigate the theoretical relationship between DFA and ACI.

Changes to gait speed and the walk ratio with rhythmic auditory cuing

BACKGROUND

Step length and cadence (i.e., step frequency or steps/minute) maintain an invariant proportion across a range of walking speeds, known as the walk ratio (WR = step length/cadence). While step length is a difficult parameter to manipulate, cadence is readily modifiable using rhythmic auditory cuing (RAC; e.g., synchronizing step timing to a metronome or music tempo).

RESEARCH QUESTION

The purpose of this study was to determine the effects of RAC-guided cadences on enacted cadence, step length, WR, and gait speed during overground walking.

METHODS

Sixteen healthy young adults repeatedly crossed a GAITRite electronic walkway while attempting to synchronize step timing to RAC-guided (metronome) tempos of 80 to 140 beats per minute. Mean absolute percent error (MAPE) was used to compare RAC tempos to enacted cadence. Repeated-measures analyses of variance were performed to test for the effects of RAC on cadence, step length, WR, and gait speed. Moreover, simple linear regressions were used to determine the precise stepwise relationship between RAC conditions and each variable.

RESULTS

Participants successfully matched their cadence to RAC beats (MAPE < 1.1%). Cadence increased proportionally to RAC (linear regression slope = 1.02), while step length also increased but at a slower rate (slope = 0.40). These dissimilar slopes resulted in a modified WR that systematically decreased with increasing cadence, although ultimately gait speed increased with increasing cadence (slope = 1.41). This relationship indicates that every 10 steps/minute incremental increase in cadence corresponded with a 14 cm/s increase in gait speed.

SIGNIFICANCE

Gait speed appears to increase in a predictable manner when cadence is guided by RAC during overground walking irrespective of apparent changes to the WR.

Multifractality of Unperturbed and Asymmetric Locomotion

The purpose of this study was to explore the extent of multifractality in unperturbed and constrained locomotion, and to determine if multifractality predicted gait adaptability. Young, healthy participants (n = 15) walked at preferred and slow speeds, as well as asymmetrically (one leg at half speed) on a split-belt treadmill. Stride time multifractality was assessed via local detrended fluctuation analysis, which evaluates the evolution of fluctuations both spatially and temporally. Unperturbed walking exhibited monofractal behavior. Asymmetric walking displayed greater multifractality in the faster moving limb, indicating more intermittent periods of extreme high or low variance. Multifractality was not associated with adaptation to asymmetric walking. These findings further suggest that unperturbed locomotion is monofractal and establish that perturbed walking yields multifractal behavior.

Examination of the gait pattern based on adjusting and resulting components of the stride-to-stride variability: proof of concept

Background

Stride-to-stride variability may be used as an indicator in the assessment of gait performance, but the evaluation of this parameter is not trivial. In the gait pattern, a deviation in one stride must be corrected within the next strides (elemental variables) to ensure a steady gait (performance variable). The variance in these elemental and performance variables may therefore be evaluated as adjusting and resulting components of variability. We explored this approach to gait evaluation by matching the velocity of one stride to a subsequent stride with four different time lags ranging from 0.5 to 2 strides with 0.5 stride increments. The time lag values corresponded to the following contralateral stride, the following ipsilateral stride, the second following contralateral stride and the second following ipsilateral stride.

Methods

Twenty asymptomatic young adults walked on an instrumented treadmill at their preferred gait speed. The stride velocity was calculated, and variances in the stride-to-stride differences and in the stride-to-stride sums represented the adjusting and the resulting variances, respectively. A ratio between these values of greater than one indicated a meaningful stride-to-stride interaction.

Results

For the four time lags (0.5, 1, 1.5, and 2 strides), the adjusting/resulting variance ratios (mean and CI 95%) were 1.0 (0.8–1.2), 2.9 (2.3–3.6), 1.2 (1.0–1.4) and 1.2 (0.9–1.4), respectively.

Conclusions

This new approach to the evaluation of stride-to-stride variability suggests that gait velocity adjustments occurred within one full stride cycle during treadmill walking among asymptomatic young adults. The validity of the approach needs to be tested in over-ground walking.

Error Correction and the Structure of Inter-Trial Fluctuations in a Redundant Movement Task

We study inter-trial movement fluctuations exhibited by human participants during the repeated execution of a virtual shuffleboard task. Focusing on skilled performance, theoretical analysis of a previously-developed general model of inter-trial error correction is used to predict the temporal and geometric structure of variability near a goal equivalent manifold (GEM). The theory also predicts that the goal-level error scales linearly with intrinsic body-level noise via the total body-goal sensitivity, a new derived quantity that illustrates how task performance arises from the interaction of active error correction and passive sensitivity properties along the GEM. Linear models estimated from observed fluctuations, together with a novel application of bootstrapping to the estimation of dynamical and correlation properties of the inter-trial dynamics, are used to experimentally confirm all predictions, thus validating our model. In addition, we show that, unlike “static” variability analyses, our dynamical approach yields results that are independent of the coordinates used to measure task execution and, in so doing, provides a new set of task coordinates that are intrinsic to the error-regulation process itself.

During the repeated execution of precision movement tasks, humans face two formidable challenges from the motor system itself: dimensionality and noise. Human motor performance involves biomechanical, neuromotor, and perceptual degrees of freedom far in excess of those theoretically needed to prescribe typical goal-directed tasks. At the same time, noise is present in the human body across multiple scales of observation. This high-dimensional and stochastic character of biological movement is the fundamental source of variability ubiquitously observed during task execution. However, it is becoming clear that these two challenges are not merely impediments to be overcome, but rather hold a key to understanding how humans maintain motor performance under changing circumstances, such as those caused by fatigue, injury, or aging. In this work, by studying skilled human participants as they play a virtual shuffleboard game, we demonstrate the fundamental importance of adopting a dynamical perspective when analyzing the motor variability observed over many trials. Using this dynamical approach, we can not only study the geometry of observed inter-trial variability, but can also theoretically describe and experimentally characterize how it is temporally generated and regulated. Furthermore, our theoretical framework and model-based data analysis approach helps to unify previous variability analysis approaches based on stability, correlation, control theory, or task manifolds alone. This conceptual unification supports the idea that such seemingly disparate features of motor variability arise from a single, relatively simple underlying neurophysiological process of motor regulation.

Comparison between treadmill training with rhythmic auditory stimulation and ground walking with rhythmic auditory stimulation on gait ability in chronic stroke patients: A pilot study

BACKGROUND

Generally, treadmill training is very effective intervention, and rhythmic auditory stimulation is designed to feedback during gait training in stroke patients.

OBJECTIVE

The purpose of this study was to compare the gait abilities in chronic stroke patients following either treadmill walking training with rhythmic auditory stimulation (TRAS) or over ground walking training with rhythmic auditory stimulation (ORAS).

METHODS

Nineteen subjects were divided into two groups: a TRAS group (9 subjects) and an ORAS group (10 subjects). Temporal and spatial gait parameters and motor recovery ability were measured before and after the training period. Gait ability was measured by the Biodex Gait trainer treadmill system, Timed up and go test (TUG), 6 meter walking distance (6MWD) and Functional gait assessment (FGA).

RESULTS

After the training periods, the TRAS group showed a significant improvement in walking speed, step cycle, step length of the unaffected limb, coefficient of variation, 6MWD, and, FGA when compared to the ORAS group (p <  0.05).

CONCLUSION

Treadmill walking training during the rhythmic auditory stimulation may be useful for rehabilitation of patients with chronic stroke.

Fractal Fluctuations in Human Walking: Comparison Between Auditory and Visually Guided Stepping

In human locomotion, sensorimotor synchronization of gait consists of the coordination of stepping with rhythmic auditory cues (auditory cueing, AC). AC changes the long-range correlations among consecutive strides (fractal dynamics) into anti-correlations. Visual cueing (VC) is the alignment of step lengths with marks on the floor. The effects of VC on the fluctuation structure of walking have not been investigated. Therefore, the objective was to compare the effects of AC and VC on the fluctuation pattern of basic spatiotemporal gait parameters. Thirty-six healthy individuals walked 3 × 500 strides on an instrumented treadmill with augmented reality capabilities. The conditions were no cueing (NC), AC, and VC. AC included an isochronous metronome. For VC, projected stepping stones were synchronized with the treadmill speed. Detrended fluctuation analysis assessed the correlation structure. The coefficient of variation (CV) was also assessed. The results showed that AC and VC similarly induced a strong anti-correlated pattern in the gait parameters. The CVs were similar between the NC and AC conditions but substantially higher in the VC condition. AC and VC probably mobilize similar motor control pathways and can be used alternatively in gait rehabilitation. However, the increased gait variability induced by VC should be considered.

Non-linear dynamics of human locomotion: effects of rhythmic auditory cueing on local dynamic stability

It has been observed that times series of gait parameters [stride length (SL), stride time (ST), and stride speed (SS)], exhibit long-term persistence and fractal-like properties. Synchronizing steps with rhythmic auditory stimuli modifies the persistent fluctuation pattern to anti-persistence. Another non-linear method estimates the degree of resilience of gait control to small perturbations, i.e., the local dynamic stability (LDS). The method makes use of the maximal Lyapunov exponent, which estimates how fast a non-linear system embedded in a reconstructed state space (attractor) diverges after an infinitesimal perturbation. We propose to use an instrumented treadmill to simultaneously measure basic gait parameters (time series of SL, ST, and SS from which the statistical persistence among consecutive strides can be assessed), and the trajectory of the center of pressure (from which the LDS can be estimated). In 20 healthy participants, the response to rhythmic auditory cueing (RAC) of LDS and of statistical persistence [assessed with detrended fluctuation analysis (DFA)] was compared. By analyzing the divergence curves, we observed that long-term LDS (computed as the reverse of the average logarithmic rate of divergence between the 4th and the 10th strides downstream from nearest neighbors in the reconstructed attractor) was strongly enhanced (relative change +73%). That is likely the indication of a more dampened dynamics. The change in short-term LDS (divergence over one step) was smaller (+3%). DFA results (scaling exponents) confirmed an anti-persistent pattern in ST, SL, and SS. Long-term LDS (but not short-term LDS) and scaling exponents exhibited a significant correlation between them (r = 0.7). Both phenomena probably result from the more conscious/voluntary gait control that is required by RAC. We suggest that LDS and statistical persistence should be used to evaluate the efficiency of cueing therapy in patients with neurological gait disorders.

Effect of aging on inter-joint synergies during machine-paced assembly tasks

In recent years, uncontrolled manifold (UCM) analysis has emerged as an important method to study variability of human movements. The current study investigated the upper extremity movements during typical assembly tasks using the framework of the UCM analysis. Younger and older participants performed machine-paced assembly tasks, while the kinematics of upper extremities were recorded using a motion tracking system. The upper extremity was modeled as a 7 degrees-of-freedom system. The variance of joint angles within the UCM space (V UCM) and the variance perpendicular to the UCM space (V ORT) were analyzed. The results indicated that V UCM were not significantly different for the older and younger groups. For the older group, V ORT was significantly less than the younger group and resulted in less total variance (V TOT) and a better synergy level (Z ΔV ). Therefore, the synergies of upper extremity movement may not be impaired for machine-paced tasks as people age. While current results showed a different effect of aging on the synergies of body movement compared with one previous study, they were in line with a recently proposed theory that for natural tasks, aging people did not have impairment in the ability to organize upper extremity movement into synergies.

Reliability of spatiotemporal and kinetic gait parameters determined by a new instrumented treadmill system

Background

Despite the emerging use of treadmills integrated with pressure platforms as outcome tools in both clinical and research settings, published evidence regarding the measurement properties of these new systems is limited. This study evaluated the within– and between–day repeatability of spatial, temporal and vertical ground reaction force parameters measured by a treadmill system instrumented with a capacitance–based pressure platform.

Methods

Thirty three healthy adults (mean age, 21.5 ± 2.8 years; height, 168.4 ± 9.9 cm; and mass, 67.8 ± 18.6 kg), walked barefoot on a treadmill system (FDM–THM–S, Zebris Medical GmbH) on three separate occasions. For each testing session, participants set their preferred pace but were blinded to treadmill speed. Spatial (foot rotation, step width, stride and step length), temporal (stride and step times, duration of stance, swing and single and double support) and peak vertical ground reaction force variables were collected over a 30–second capture period, equating to an average of 52 ± 5 steps of steady–state walking. Testing was repeated one week following the initial trial and again, for a third time, 20 minutes later. Repeated measures ANOVAs within a generalized linear modelling framework were used to assess between–session differences in gait parameters. Agreement between gait parameters measured within the same day (session 2 and 3) and between days (session 1 and 2; 1 and 3) were evaluated using the 95% repeatability coefficient.

Results

There were statistically significant differences in the majority (14/16) of temporal, spatial and kinetic gait parameters over the three test sessions (P < .01). The minimum change that could be detected with 95% confidence ranged between 3% and 17% for temporal parameters, 14% and 33% for spatial parameters, and 4% and 20% for kinetic parameters between days. Within–day repeatability was similar to that observed between days. Temporal and kinetic gait parameters were typically more consistent than spatial parameters. The 95% repeatability coefficient for vertical force peaks ranged between ± 53 and ± 63 N.

Conclusions

The limits of agreement in spatial parameters and ground reaction forces for the treadmill system encompass previously reported changes with neuromuscular pathology and footwear interventions. These findings provide clinicians and researchers with an indication of the repeatability and sensitivity of the Zebris treadmill system to detect changes in common spatiotemporal gait parameters and vertical ground reaction forces.