Gait analysis using gravitational acceleration measured by wearable sensors

A novel method for measuring human gait posture using wearable sensor units is proposed. The sensor units consist of a tri-axial acceleration sensor and three gyro sensors aligned on three axes. The acceleration and angular velocity during walking were measured with seven sensor units worn on the abdomen and the lower limb segments (both thighs, shanks and feet). The three-dimensional positions of each joint are calculated from each segment length and joint angle. Joint angle can be estimated mechanically from the gravitational acceleration along the anterior axis of the segment. However, the acceleration data during walking includes three major components; translational acceleration, gravitational acceleration and external noise. Therefore, an optimization analysis was represented to separate only the gravitational acceleration from the acceleration data. Because the cyclic patterns of acceleration data can be found during constant walking, a FFT analysis was applied to obtain some characteristic frequencies in it. A pattern of gravitational acceleration was assumed using some parts of these characteristic frequencies. Every joint position was calculated from the pattern under the condition of physiological motion range of each joint. An optimized pattern of the gravitational acceleration was selected as a solution of an inverse problem. Gaits of three healthy volunteers were measured by walking for 20 s on a flat floor. As a result, the acceleration data of every segment was measured simultaneously. The characteristic three-dimensional walking could be shown by the expression using a stick figure model. In addition, the trajectories of the knee joint in the horizontal plane could be checked by visual imaging on a PC. Therefore, this method provides important quantitive information for gait diagnosis.     

Triaxial accelerometry for assessment of physical activity in young children

Objective: The purpose of the present study was to derive linear and non‐linear regression equations that estimate energy expenditure (EE) from triaxial accelerometer counts that can be used to quantitate activity in young children. We are unaware of any data regarding the validity of triaxial accelerometry for assessment of physical activity intensity in this age group. Research Methods and Procedures: EE for 27 girls and boys (6.0 ± 0.3 years) was assessed for nine activities (lying down, watching a video while sitting and standing, line drawing for coloring‐in, playing blocks, walking, stair climbing, ball toss, and running) using indirect calorimetry and was then estimated using a triaxial accelerometer (ActivTracer, GMS). Results: Significant correlations were observed between synthetic (synthesized tri‐axes as the vector), vertical, and horizontal accelerometer counts and EE for all activities (0.878 to 0.932 for EE). However, linear and non‐linear regression equations underestimated EE by >30% for stair climbing (up and down) and performing a ball toss. Therefore, linear and non‐linear regression equations were calculated for all activities except these two activities, and then evaluated for all activities. Linear and non‐linear regression equations using combined vertical and horizontal acceleration counts, synthetic counts, and horizontal counts demonstrated a better relationship between accelerometer counts and EE than did regression equations using vertical acceleration counts. Adjustment of the predicted value by the regression equations using the vertical/horizontal counts ratio improved the overestimation of EE for performing a ball toss. Discussion: The results suggest that triaxial accelerometry is a good tool for assessing daily EE in young children.     

Prey capture attempts can be detected in Steller sea lions and other marine predators using accelerometers

We attached accelerometers to the head and jaw of a Steller sea lion (Eumetopias jubatus) to determine whether feeding attempts in a controlled setting could be quantified by acceleration features characteristic of head and jaw movements. Most of the 19 experimental feeding events that occurred during the 51 dives recorded resulted in specific acceleration patterns that were clearly distinguishable from swimming accelerations. The differential acceleration between the head-mounted and jaw-mounted accelerometers detected 84% of prey captures on the vertical axis and 89% on the horizontal axis. However, the jaw-mounted accelerometer alone proved to be equally effective at detecting prey capture attempts. Acceleration along the horizontal (surge)-axis appeared to be particularly efficient in detecting prey captures, and suggests that a single accelerometer placed under the jaw of a pinniped is a promising and easily implemented means of recording prey capture attempts.     

Postural stability of the human mandible during locomotion

Movements of the head and of the mandible relative to the head were measured in human subjects walking and running on a treadmill at various speeds and inclinations. A miniature magnet and piezo-electric accelerometer assembly was mounted on the mandibular incisors, and a Hall-effect sensor along with a second accelerometer mounted on a maxillary incisor along a common vertical axis. Signals from these sensors provided continuous records of vertical head and mandible acceleration, and relative jaw position. Landing on the heel or on the toe in different forms of locomotion was followed by rapid deceleration of the downward movement of the head and slightly less rapid deceleration of the downward movement of the mandible, i.e., the mandible moved downwards relative to the maxilla, then upwards again to near its normal posture within 200 ms. No tooth contact occurred in any forms of gait at any inclination. The movement of the mandible relative to the maxilla depended on the nature and velocity of the locomotion and their effects on head deceleration. The least deceleration and hence mandibular displacement occurred during toe-landing, for example, during "uphill" running. The maximum displacement of the mandible relative to the head was less than 1mm, even at the fastest running speed. The mechanisms that limit the vertical movements of the jaw within such a narrow range are not known, but are likely to include passive soft-tissue visco-elasticity and stretch reflexes in the jaw-closing muscles.     

Testing the assumption in ergonomics software that overall shoulder strength can be accurately calculated by treating orthopedic axes as independent

To predict shoulder strength, most current ergonomics software assume independence of the strengths about each of the orthopedic axes. Using this independent axis approach (IAA), the shoulder can be predicted to have strengths as high as the resultant of the maximum moment about any two or three axes. We propose that shoulder strength is not independent between axes, and propose an approach that calculates the weighted average (WAA) between the strengths of the axes involved in the demand.Fifteen female participants performed maximum isometric shoulder exertions with their right arm placed in a rigid adjustable brace affixed to a tri-axial load cell. Maximum exertions were performed in 24 directions, including four primary directions, horizontal flexion-extension, abduction-adduction, and at 15° increments in between those axes. Moments were computed and comparisons made between the experimentally collected strengths and those predicted by the IAA and WAA methods.The IAA over-predicted strength in 14 of 20 non-primary exertions directions, while the WAA underpredicted strength in only 2 of these directions. Therefore, it is not valid to assume that shoulder axes are independent when predicting shoulder strengths between two orthopedic axes, and the WAA is an improvement over current methods for the posture tested.     

Classification of gait patterns in the time-frequency domain

This paper describes the classification of gait patterns among descending stairs, ascending stairs and level walking activities using accelerometers arranged in antero-posterior and vertical direction on the shoulder of a garment. Gait patterns in continuous accelerometer records were classified in two steps. In the first step, direct spatial correlation of discrete dyadic wavelet coefficients was applied to separate the segments of gait patterns in the continuous accelerometer record. Compared to the reference system, averaged absolute error 0.387 s for ascending stairs and 0.404 s for descending stairs were achieved. The overall sensitivity and specificity of ascending stairs were 98.79% and 99.52%, and those of descending stairs were 97.35% and 99.62%. In the second step, powers of wavelet coefficients of 2 s time duration from separated segments of vertical and antero-posterior acceleration signals were used as features in classification. Our results proved a reliable technique of measuring gait patterns during physical activity.     

Gait posture estimation using wearable acceleration and gyro sensors

A method for gait analysis using wearable acceleration sensors and gyro sensors is proposed in this work. The volunteers wore sensor units that included a tri-axis acceleration sensor and three single axis gyro sensors. The angular velocity data measured by the gyro sensors were used to estimate the translational acceleration in the gait analysis. The translational acceleration was then subtracted from the acceleration sensor measurements to obtain the gravitational acceleration, giving the orientation of the lower limb segments. Segment orientation along with body measurements were used to obtain the positions of hip, knee, and ankle joints to create stick figure models of the volunteers. This method can measure the three-dimensional positions of joint centers of the hip, knee, and ankle during movement. Experiments were carried out on the normal gait of three healthy volunteers. As a result, the flexion–extension (F–E) and the adduction–abduction (A–A) joint angles of the hips and the flexion–extension (F–E) joint angles of the knees were calculated and compared with a camera motion capture system. The correlation coefficients were above 0.88 for the hip F–E, higher than 0.72 for the hip A–A, better than 0.92 for the knee F–E. A moving stick figure model of each volunteer was created to visually confirm the walking posture. Further, the knee and ankle joint trajectories in the horizontal plane showed that the left and right legs were bilaterally symmetric.     

Body Acceleration as Indicator for Walking Economy in an Ageing Population

Background

In adults, walking economy declines with increasing age and negatively influences walking speed. This study aims at detecting determinants of walking economy from body acceleration during walking in an ageing population.

Methods

35 healthy elderly (18 males, age 51 to 83 y, BMI 25.5±2.4 kg/m2) walked on a treadmill. Energy expenditure was measured with indirect calorimetry while body acceleration was sampled at 60Hz with a tri-axial accelerometer (GT3X+, ActiGraph), positioned on the lower back. Walking economy was measured as lowest energy needed to displace one kilogram of body mass for one meter while walking (WCost_min, J/m/kg). Gait features were extracted from the acceleration signal and included in a model to predict WCost_min.

Results

On average WCost_min was 2.43±0.42 J/m/kg and correlated significantly with gait rate (r² = 0.21, p<0.01) and regularity along the frontal (anteroposterior) and lateral (mediolateral) axes (r² = 0.16, p<0.05 and r² = 0.12, p<0.05 respectively). Together, the three variables explained 46% of the inter-subject variance (p<0.001) with a standard error of estimate of 0.30 J/m/kg. WCost_min and regularity along the frontal and lateral axes were related to age (WCost_min: r² = 0.44, p<0.001; regularity: r² = 0.16, p<0.05 and r² = 0.12, p<0.05 respectively frontal and lateral).

Conclusions

The age associated decline in walking economy is induced by the adoption of an increased gait rate and by irregular body acceleration in the horizontal plane.     

Examination of Different Accelerometer Cut-Points for Assessing Sedentary Behaviors in Children

Background

Public health research on sedentary behavior (SB) in youth has heavily relied on accelerometers. However, it has been limited by the lack of consensus on the most accurate accelerometer cut-points as well as by unknown effects caused by accelerometer position (wrist vs. hip) and output (single axis vs. multiple axes). The present study systematically evaluates classification accuracy of different Actigraph cut-points for classifying SB using hip and wrist-worn monitors and establishes new cut-points to enable use of the 3-dimensional vector magnitude data (for both hip and wrist placement).

Methods

A total of 125 children ages 7–13 yrs performed 12 randomly selected activities (from a set of 24 different activities) for 5 min each while wearing tri-axial Actigraph accelerometers on both the hip and wrist. The accelerometer data were categorized as either sedentary or non-sedentary minutes using six previously studied cut-points: 100counts-per-minute (CPM), 200CPM, 300CPM, 500CPM, 800CPM and 1100CPM. Classification accuracy was evaluated with Cohen''s Kappa (κ) and new cut-points were identified from Receiver Operating Characteristic (ROC).

Results

Of the six cut-points, the 100CPM value yielded the highest classification accuracy (κ = 0.81) for hip placement. For wrist placement, all of the cut-points produced low classification accuracy (ranges of κ from 0.44 to 0.67). Optimal sedentary cut-points derived from ROC were 554.3CPM (ROC-AUC of 0.99) for vector magnitude for hip, 1756CPM (ROC-AUC of 0.94) for vertical axis for wrist, and 3958.3CPM (ROC-AUC of 0.93) for vector magnitude for wrist placement.

Conclusions

The 100CPM was supported for use with vertical axis for hip placement, but not for wrist placement. The ROC-derived cut-points can be used to classify youth SB with the wrist and with vector magnitude data.     

Gait stability,variability and complexity on inclined surfaces

This study evaluated the gait stability, variability, and complexity of healthy young adults on inclined surfaces. A total of 49 individuals walked on a treadmill at their preferred speed for 4 min at inclinations of 6%, 8%, and 10% in upward (UP) and downward (DOWN) conditions, and in horizontal (0%) condition. Gait variability was assessed using average standard deviation trunk acceleration between strides (VAR), gait stability was assessed using margin of stability (MoS) and maximum Lyapunov exponent (λs), and gait complexity was assessed using sample entropy (SEn). Trunk variability (VAR) increased in the medial-lateral (ML), anterior-posterior, and vertical directions for all inclined conditions. The SEn values indicated that movement complexity decreased almost linearly from DOWN to UP conditions, reflecting changes in gait pattern with longer and slower steps as inclination increased. The DOWN conditions were associated with the highest variability and lowest stability in the MoS ML, but not in λs. Stability was lower in UP conditions, which exhibited the largest λs values. The overall results support the hypothesis that inclined surfaces decrease gait stability and alter gait variability, particularly in UP conditions.     

Angular acceleration,compensatory head movements and the halteres of flies (Lucilia serricata)

Summary Tethered flies were subjected to accelerations about their vertical axes while flying or walking. These accelerations were applied either suddenly to stationary animals or continuously by oscillating the animal from side to side. Head and wing movements resulting from the imposed angular accelerations were photographed with a camera and a stroboscopic flash.Analysis of the photographs shows that the wing movements act to counter the imposed angular accelerations and that during sinusoidal oscillations about the vertical axis, head turns are in antiphase with angular acceleration.Head turns do not occur when the halteres are absent or present and not oscillating. When oscillating, the halteres detect high values of angular acceleration, outside the known capabilities of the visual movement detection system.     

Classification accuracy of a single tri-axial accelerometer for training background and experience level in runners

Accelerometers are increasingly used tools for gait analysis, but there remains a lack of research on their application to running and their ability to classify running patterns. The purpose of this study was to conduct an exploratory examination into the capability of a tri-axial accelerometer to classify runners of different training backgrounds and experience levels, according to their 3-dimensional (3D) accelerometer data patterns. Training background was examined with 14 competitive soccer players and 12 experienced marathon runners, and experience level was examined with 16 first-time and the same 12 experienced marathon runners. Discrete variables were extracted from 3D accelerations during a short run using root mean square, wavelet transformation, and autocorrelation procedures. A principal component analysis (PCA) was conducted on all variables, including gait speed to account for covariance. Eight PCs were retained, explaining 88% of the variance in the data. A stepwise discriminant analysis of PCs was used to determine the binary classification accuracy for training background and experience level, with and without the PC of Speed. With Speed, the accelerometer correctly classified 96% of runners for both training background and experience level. Without Speed, the accelerometer correctly classified 85% of runners based on training background, but only 68% based on experience level. These findings suggest that the accelerometer is effective in classifying athletes of different training backgrounds, but is less effective for classifying runners of different experience levels where gait speed is the primary discriminator.     

Estimation of resistance exercise energy expenditure using triaxial accelerometry

Recently, it was demonstrated that a uniaxial accelerometer worn at the hip could estimate resistance exercise energy expenditure. As resistance exercise takes place in more than 1 plane, the use of a triaxial accelerometer may be more effective in estimating resistance exercise energy expenditure. The aims of this study were to estimate the energy cost of resistance exercise using triaxial accelerometry and to determine the optimal location for wearing triaxial accelerometers during resistance exercise. Thirty subjects (15 men and 15 women; age = 21.7 ± 1.0 years) performed a resistance exercise protocol consisting of 2 sets of 8 exercises (10RM loads). During the resistance exercise protocol, subjects wore triaxial accelerometers on the wrist, waist, and ankle; a heart rate monitor; and a portable metabolic system. Net energy expenditure was significantly correlated with vertical (r = 0.67, p < 0.001), horizontal (r = 0.43, p = 0.02), third axis (r = 0.36, p = 0.048), and sum of 3 axes (r = 0.50, p = 0.005) counts at the waist, and horizontal counts at the wrist (r = -0.40, p = 0.03). Regression analysis using fat-free mass, sex, and the sum of accelerometer counts at the waist as variables was used to develop an equation that explained 73% of the variance of resistance exercise energy expenditure. A triaxial accelerometer worn at the waist can be used to estimate resistance exercise energy expenditure but appears to offer no benefit over uniaxial accelerometry. The use of accelerometers in estimating resistance exercise energy expenditure may prove useful for individuals and athletes who participate in resistance training and are focused on maintaining a tightly regulated energy balance.     

A Web-based semantic tagging and activity recognition system for species' accelerometry data

Increasingly, animal biologists are taking advantage of low cost micro-sensor technology, by deploying accelerometers to monitor the behavior and movement of a broad range of species. The result is an avalanche of complex tri-axial accelerometer data streams that capture observations and measurements of a wide range of animal body motion and posture parameters. Analysis of these parameters enables the identification of specific animal behaviors—however the analysis process is immature with much of the activity identification steps undertaken manually and subjectively. Consequently, there is an urgent need for the development of new tools to streamline the management, analysis, indexing, querying and visualization of such data. In this paper, we present a Semantic Annotation and Activity Recognition (SAAR) system which supports storing, visualizing, annotating and automatic recognition of tri-axial accelerometer data streams by integrating semantic annotation and visualization services with Support Vector Machine (SVM) techniques. The interactive Web interface enables biologists to visualize and correlate 3D accelerometer data streams with associated video streams. It also enables domain experts to accurately annotate or tag segments of tri-axial accelerometer data streams, with standardized terms from an activity ontology. These annotated data streams can then be used to dynamically train a hierarchical SVM activity classification model, which can be applied to new accelerometer data streams to automatically recognize specific activities. This paper describes the design, implementation and functional details of the SAAR system and the results of the evaluation experiments that assess the performance, usability and efficiency of the system. The evaluation results indicate that the SAAR system enables ecologists with little knowledge of machine learning techniques to collaboratively build classification models with high levels of accuracy, sensitivity, precision and specificity.     

Comparison of the hindfoot axes of a multi-segment foot model to the underlying bony anatomy

Musculoskeletal models used in gait analysis require coordinate systems to be identified for the body segments of interest. It is not obvious how hindfoot (or rearfoot) axes defined by skin-mounted markers relate to the anatomy of the underlying bones. The aim of this study was to compare the marker-based axes of the hindfoot in a multi-segment foot model to the orientations of the talus and calcaneus as characterized by their principal axes of inertia. Twenty adult females with no known foot deformities had radio-opaque markers placed on their feet and ankles at the foot model marker locations. CT images of the feet were acquired as the participants lay supine with their feet in a semi-weight bearing posture. The spatial coordinates of the markers were obtained from the images and used to define the foot model axes. Segmented masks of the tali and calcanei were used to create 3D bone models, from which the principal axes of the bones were obtained. The orientations of the principal axes were either within the range of typical values reported in the imaging literature or differed in ways that could be explained by variations in how the angles were defined. The model hindfoot axis orientations relative to the principal axes of the bones had little bias but were highly variable. Consideration of coronal plane hindfoot alignment as measured clinically and radiographically suggested that the model hindfoot coordinate system represents the posterior calcaneal tuberosity, rather than the calcaneus as a whole.     

Wireless Tri-Axial Trunk Accelerometry Detects Deviations in Dynamic Center of Mass Motion Due to Running-Induced Fatigue

Small wireless trunk accelerometers have become a popular approach to unobtrusively quantify human locomotion and provide insights into both gait rehabilitation and sports performance. However, limited evidence exists as to which trunk accelerometry measures are suitable for the purpose of detecting movement compensations while running, and specifically in response to fatigue. The aim of this study was therefore to detect deviations in the dynamic center of mass (CoM) motion due to running-induced fatigue using tri-axial trunk accelerometry. Twenty runners aged 18–25 years completed an indoor treadmill running protocol to volitional exhaustion at speeds equivalent to their 3.2 km time trial performance. The following dependent measures were extracted from tri-axial trunk accelerations of 20 running steps before and after the treadmill fatigue protocol: the tri-axial ratio of acceleration root mean square (RMS) to the resultant vector RMS, step and stride regularity (autocorrelation procedure), and sample entropy. Running-induced fatigue increased mediolateral and anteroposterior ratios of acceleration RMS (p < .05), decreased the anteroposterior step regularity (p < .05), and increased the anteroposterior sample entropy (p < .05) of trunk accelerometry patterns. Our findings indicate that treadmill running-induced fatigue might reveal itself in a greater contribution of variability in horizontal plane trunk accelerations, with anteroposterior trunk accelerations that are less regular from step-to-step and are less predictable. It appears that trunk accelerometry parameters can be used to detect deviations in dynamic CoM motion induced by treadmill running fatigue, yet it is unknown how robust or generalizable these parameters are to outdoor running environments.     

Mechanics of locomotion of dogs (Canis familiaris) and sheep (Ovis aries)

Records have been made of the forces exerted on the ground by dogs and a sheep, in walking, trotting, cantering and slow galloping. Film has been taken simultaneously. The difference between walking and trotting was much less marked for the sheep than for the dogs.
Step length and stride length increase as speed increases. They are expressed as functions of the Froude number.
The vertical component of the force exerted by a foot on the ground shows two main maxima in walking, except in the case of the fore feet of sheep. In this case and in other gaits there is only one main maximum. The vertical movements of the fore and hind quarters which occurred in examples of each gait have been calculated from the force records.
The force exerted by a foot on the ground changes direction in the course of a step so as to remain more or less in line with a point fixed relative to the animal, but dorsal to its back.
The force records show impact disturbances in the first 003 sec of contact of each foot with the ground.
The point of application of the force on the sole of a foot tends to move posteriorly as the force increases.
The results are discussed in relation to a theoretical account of the mechanics of locomotion on legs.     

The biomechanics of skipping gaits: a third locomotion paradigm?

Skipping, a gait children display when they are about four- to five-years-old, is revealed to be more than a behavioural peculiarity. By means of metabolic and biomechanical measurements at several speeds, the relevance of skipping is shown to extend from links between bipedal and quadrupedal locomotion (namely galloping) to understanding why it could be a gait of choice in low-gravity conditions, and to some aspects of locomotion evolution (ground reaction forces of skipping seem to originate from pushing the walking gait to unnaturally high speeds). When the time-courses of mechanical energy and the horizontal ground reaction force are considered, a different locomotion paradigm emerges, enabling us to separate, among the bouncing gaits, the trot from the gallop (quadrupeds) and running from skipping (bipeds). The simultaneous use of pendulum-like and elastic mechanisms in skipping gaits, as shown by the energy curve analysis, helps us to understand the low cost of transport of galloping quadrupeds.     

Heel-off perturbation during gait initiation: biomechanical analysis using triaxial accelerometry and a force plate.

This study analyzes the movements of the hips, shoulders and of the body center of gravity before and at heel-off, when step execution begins to initiate gait from an upright posture. The heel-off movement was considered as a dynamic perturbation induced by the stepping movement. The experimental paradigm used for studying this perturbation was the single-step movement, in which the initial posture and voluntary movements are identical to those of gait initiation. Data were collected from accelerometer recordings of the triaxial accelerations at the joints of the upper part of the body, and by calculating the triaxial accelerations of the center of gravity using force plate measurements. The resultant vectors were used to establish and compare the magnitude and direction of the accelerations at different joints, and from them, the roles of the pelvis and the scapular girdles with respect to the objectives of the gait movement.