Accelerometer Placements Research Articles

Accelerometry for sleep assessment in children: Criterium validity of different algorithms in wrist- and ankle-worn devices.

Polysomnography, the gold-standard for measuring sleep, is costly, intrusive and usually limited to 1 night. Actigraphy offers a more affordable, less intrusive method over multiple nights. However, little research validates ActiGraph accelerometers against polysomnography, especially in children. This study evaluated the validity of different algorithms and compared wrist versus ankle accelerometer placements for estimating sleep in children aged 1-12 years. Twenty-nine children undergoing overnight type 1 polysomnography wore ActiGraph accelerometers. Six algorithms were evaluated against polysomnography using Pearson correlations, intraclass correlation, paired t-tests and Bland-Altman plots. Agreement was classified as poor (intraclass correlation coefficient < 0.4), fair (0.4 < intraclass correlation coefficient < 0.6), good (0.6 < intraclass correlation coefficient < 0.75) or excellent (intraclass correlation coefficient > 0.75). Total sleep time was the primary outcome. For wrist-worn devices, the Sadeh (Actilife) and Cole-Kripke (Actilife and GGIR) algorithms showed excellent agreement with polysomnography (intraclass correlation coefficient = 0.80-0.85), while vanHees showed good agreement (intraclass correlation coefficient = 0.67) and Galland showed fair agreement (intraclass correlation coefficient = 0.46). The Cole-Kripke algorithm did not significantly differ from polysomnography total sleep time, whereas others underestimated total sleep time. For ankle-worn devices, Sadeh (Actilife), Cole-Kripke (Actilife) and vanHees algorithms demonstrated excellent agreement (intraclass correlation coefficient = 0.75-0.82). No significant differences were found between wrist and ankle placements for certain algorithms. The findings support accelerometry as a valid tool for sleep assessment in children, recommending that algorithm selection be tailored to specific study requirements.

Optimizing Accelerometer Placements For Detecting Upper Body Movement During Occupational Physical Activity Using Machine Learning

Optimizing Accelerometer Placements For Detecting Upper Body Movement During Occupational Physical Activity Using Machine Learning

Validating device-based physical activity indicators with observation physical activity indicators in Swiss pre-school children

Purpose&#x0D; Accelometry is a very important tool to measure physical activity (PA) in preschoolers, as it can measure 24-hour PA and detect activity which cannot be captured by other methods. However, accelerometer validation for preschoolers is lacking (Altenburg et al. 2022). Specifically, for the ActiGraph wGT3X-BT and Move 4 (ActiGraph LLC; Move4 activity sensor, movisens GmbH) only a few, respectively no preschooler validation studies, were found. Thus, the purpose was to validate these two accelerometers in Swiss preschoolers with the System for Observing Children’s Activity and Relationships during Play (SOCARP; Ridgers et al., 2010).&#x0D; Methods and Design&#x0D; Preschoolers (2-5 years old) from two Swiss Sunday activity programs (MiniMove &amp; Ä Halle wo’s fägt) were randomly selected as part of a larger program evaluation. PA was assessed SOCARP for a duration of 12 minutes per child. During the observation, the children wore both an ActiGraph and a Move4 device taped to their right hip to record steps. Step-counts from the ActiGraph and Move4 were correlated with each other and with moderate-to-vigorous (MV)PA from SOCARP (as SOCARP does not count steps).&#x0D; Results&#x0D; Valid PA data was available for 45/58 (77.6%) children (49% girls) for SOCARP and for 47/58 (81%) children (51% girls) for accelometry. Step count correlations between the accelerometers (Actigraph and Move4) and %MVPA (SOCARP) was medium and positive (r(43) = .34, p = .03 and r(43) = .37, p = .02; respectively). There was a strong step count correlation between the two devices (r(45) = .90, p &lt; .001), although ActiGraph measured significantly more steps than Move4 (m = 557.74, SD = 255.77 versus m = 397.81, SD = 164.10); t(46)=8.47, p &lt; .001).&#x0D; Discussion&#x0D; PA measurement in preschool children can be challenging. However, the correlation between step counts and observed %MVPA indicates criterion validity for both devices. The step-counts of Actigraph and Move4 validate each other, but there is a difference in the absolute number of measured steps. Due to different outcome parameters, calculation algorithms, and inaccessibility to raw acceleration the comparison of the two devices on movement intensity was not possible. Although promising preliminary indications of validity of device-based measurement of PA in Swiss preschool children, further investigations into the methodological approaches of comparing measurements of movement intensity are warranted.&#x0D; References&#x0D; Altenburg, T. M., de Vries, L., op den Buijsch, R., Eyre, E., Dobell, A., Duncan, M., &amp; Chinapaw, M. J. M. (2022). Cross-validation of cut-points in preschool children using different accelerometer placements and data axes. Journal of Sports Sciences, 40(4), 379-385. https://doi.org/10.1080/02640414.2021.1994726&#x0D; Ridgers, N. D., Stratton, G., &amp; McKenzie, T. L. (2010). Reliability and validity of the system for observing children’s activity and relationships during play (SOCARP). Journal of Physical Activity and Health, 7(1), 17–25. https://doi.org/10.1123/jpah.7.1.17

Mechanical loading prediction through accelerometry data during walking and running

ABSTRACT Currently, there is no way to assess mechanical loading variables such as peak ground reaction forces (pGRF) and peak loading rate (pLR) in clinical settings. The purpose of this study was to develop accelerometry-based equations to predict both pGRF and pLR during walking and running. One hundred and thirty one subjects (79 females; 76.9 ± 19.6 kg) walked and ran at different speeds (2–14 km·h−1) on a force plate–instrumented treadmill while wearing accelerometers at their ankle, lower back and hip. Regression equations were developed to predict pGRF and pLR from accelerometry data. Leave-one-out cross-validation was used to calculate prediction accuracy and Bland–Altman plots. Our pGRF prediction equation was compared with a reference equation previously published. Body mass and peak acceleration were included for pGRF prediction and body mass and peak acceleration rate for pLR prediction. All pGRF equation coefficients of determination were above 0.96, and a good agreement between actual and predicted pGRF was observed, with a mean absolute percent error (MAPE) below 7.3%. Accuracy indices from our equations were better than previously developed equations. All pLR prediction equations presented a lower accuracy compared to those developed to predict pGRF. Walking and running pGRF can be predicted with high accuracy by accelerometry-based equations, representing an easy way to determine mechanical loading in free-living conditions. The pLR prediction equations yielded a somewhat lower prediction accuracy compared with the pGRF equations.

Cross-validation of cut-points in preschool children using different accelerometer placements and data axes

ABSTRACT The present study cross-validated various cut-points to assess physical activity and sedentary behaviour in preschoolers, using hip- and wrist-worn accelerometers and both vertical axis and vector magnitude data. Secondly, we examined the influence of epoch length on time estimates of physical activity and sedentary behaviour. Sixty-four preschoolers (34 girls) wore two accelerometers, on their right hip and dominant wrist, during 1 hour of free play. Preschoolers’ activities were observed by two trained researchers. Area under the curve (AUC) was calculated for the receiving operating characteristic (ROC) curves as a measure of precision. AUC ranges were 0.603–0.723 for sedentary behaviour, 0.472–0.545 for light physical activity and 0.503–0.661 for moderate-to-vigorous physical activity (MVPA), indicating poor to fair precision. Percentage of time classified as sedentary behaviour, light or MVPA according to observation and accelerometer data varied largely between cut-points, accelerometer placements and axes. The influence of epoch length on time estimates was minimal across cut-points, except for one hip-based vector magnitude cut-point. Across all accelerometer placements and data axes, no set of cut-points demonstrated adequate precision for sedentary behaviour, light physical activity and MVPA. The highly variable and omnidirectional activity pattern of preschoolers may explain the lack of adequate cut-points.

Machine learning to quantify habitual physical activity in children with cerebral palsy.

To investigate whether activity-monitors and machine learning models could provide accurate information about physical activity performed by children and adolescents with cerebral palsy (CP) who use mobility aids for ambulation. Eleven participants (mean age 11y [SD 3y]; six females, five males) classified in Gross Motor Function Classification System (GMFCS) levels III and IV, completed six physical activity trials wearing a tri-axial accelerometer on the wrist, hip, and thigh. Trials included supine rest, upper-limb task, walking, wheelchair propulsion, and cycling. Three supervised learning algorithms (decision tree, support vector machine [SVM], random forest) were trained on features in the raw-acceleration signal. Model-performance was evaluated using leave-one-subject-out cross-validation accuracy. Cross-validation accuracy for the single-placement models ranged from 59% to 79%, with the best performance achieved by the random forest wrist model (79%). Combining features from two or more accelerometer placements significantly improved classification accuracy. The random forest wrist and hip model achieved an overall accuracy of 92%, while the SVM wrist, hip, and thigh model achieved an overall accuracy of 90%. Models trained on features in the raw-acceleration signal may provide accurate recognition of clinically relevant physical activity behaviours in children and adolescents with CP who use mobility aids for ambulation in a controlled setting. Machine learning may assist clinicians in evaluating the efficacy of surgical and therapy-based interventions. Machine learning may help researchers better understand the short- and long-term benefits of physical activity for children with more severe motor impairments.

Wrist-worn triaxial accelerometry predicts the energy expenditure of non-vigorous daily physical activities.

Triaxial accelerometry is commonly used to estimate oxygen uptake (VO2) and energy expenditure in health and fitness studies. We tested the correlation of a triaxial accelerometer in terms of a summation of vector magnitudes with gravity subtracted (SVMgs) and measured VO2 for different daily physical activities. Original research, cross-sectional. Twenty volunteers wore a triaxial accelerometer on both wrists while performing 12 assigned daily physical activities for 6min for each activity. The VO2 was determined by indirect calorimetry using a portable metabolic measurement system. The last 3min of each activity was assumed to represent steady-state. The VO2 measured during these periods was averaged and converted into metabolic equivalents (METs). The range of VO2 for all activities was 0.18-3.2L/min (0.8-12.2 METs). Significant differences in SVMgs existed between accelerometer placements on the dominant (120.9±8.7gmin) versus non-dominant hand (99.9±6.8gmin; P=0.016) for the lowest levels of physical activity defined as <1.5 METs. Piecewise linear regression model using 6 METs as the transition point showed similar significant correlations for the non-dominant wrist (r2=0.85; P<0.001) and the dominant wrist (r2=0.86; P<0.001). Using the non-dominant wrist below 6 METs, the slope of the relationship between SVMgs and METs was 105.3±4.3 (95% CI 96.9 to 113.7) indicating an increase in SVMgs of approximately 100 units for every MET increase in oxygen uptake. Wrist-worn triaxial accelerometry reliably predicted energy expenditure during common physical activities <6 METs. More consistent correlations were found when the accelerometer was worn on the non-dominant wrist rather than the dominant wrist.

Comparison of linear and non-linear models for predicting energy expenditure from raw accelerometer data

This study had three purposes, all related to evaluating energy expenditure (EE) prediction accuracy from body-worn accelerometers: (1) compare linear regression to linear mixed models, (2) compare linear models to artificial neural network models, and (3) compare accuracy of accelerometers placed on the hip, thigh, and wrists. Forty individuals performed 13 activities in a 90 min semi-structured, laboratory-based protocol. Participants wore accelerometers on the right hip, right thigh, and both wrists and a portable metabolic analyzer (EE criterion). Four EE prediction models were developed for each accelerometer: linear regression, linear mixed, and two ANN models. EE prediction accuracy was assessed using correlations, root mean square error (RMSE), and bias and was compared across models and accelerometers using repeated-measures analysis of variance. For all accelerometer placements, there were no significant differences for correlations or RMSE between linear regression and linear mixed models (correlations: r = 0.71–0.88, RMSE: 1.11–1.61 METs; p > 0.05). For the thigh-worn accelerometer, there were no differences in correlations or RMSE between linear and ANN models (ANN—correlations: r = 0.89, RMSE: 1.07–1.08 METs. Linear models—correlations: r = 0.88, RMSE: 1.10–1.11 METs; p > 0.05). Conversely, one ANN had higher correlations and lower RMSE than both linear models for the hip (ANN—correlation: r = 0.88, RMSE: 1.12 METs. Linear models—correlations: r = 0.86, RMSE: 1.18–1.19 METs; p < 0.05), and both ANNs had higher correlations and lower RMSE than both linear models for the wrist-worn accelerometers (ANN—correlations: r = 0.82–0.84, RMSE: 1.26–1.32 METs. Linear models—correlations: r = 0.71–0.73, RMSE: 1.55–1.61 METs; p < 0.01). For studies using wrist-worn accelerometers, machine learning models offer a significant improvement in EE prediction accuracy over linear models. Conversely, linear models showed similar EE prediction accuracy to machine learning models for hip- and thigh-worn accelerometers and may be viable alternative modeling techniques for EE prediction for hip- or thigh-worn accelerometers.

A reduced modal parameter based algorithm to estimate excitation forces from optimally placed accelerometers

This paper presents a time domain technique for estimating dynamic loads acting on a structure from acceleration time response measured experimentally at a finite number of optimally placed accelerometers on the structure. The technique utilizes model reduction to obtain precise load estimates. The structure essentially acts as its own load transducer. The approach is based on the standard equilibrium equations of motion in modal coordinates. The modal parameters of a system – natural frequencies, mode shapes and damping factors – can be estimated experimentally from measured data, analytically for simple problems, or using the finite element method. For measurement of the acceleration response, there can be a large number of locations on the structure where the accelerometers can be mounted, and the precision with which the applied loads are estimated from measured acceleration response may be strongly influenced by the locations selected for accelerometer placements. A solution approach, based on the construction of D-optimal designs, is presented to determine the number and optimum locations of accelerometers that will provide the most precise load estimates. An improvement in the algorithm, based on reduced modal matrix, is further proposed to reconstruct the input forces accurately. Numerical examples that help understand the main characteristics of the proposed approach are also presented. The numerical results illustrate the effectiveness of the proposed technique in accurately recovering the loads imposed on discrete as well as continuous systems.

Posture and Movement Classification: The Comparison of Tri-Axial Accelerometer Numbers and Anatomical Placement

Patient compliance is important when assessing movement, particularly in a free-living environment when patients are asked to don their own accelerometers. Reducing the number of accelerometers could increase patient compliance. The aims of this study were (1) to determine and compare the validity of different accelerometer combinations and placements for a previously developed posture and dynamic movement identification algorithm. Custom-built activity monitors, each containing one tri-axial accelerometer, were placed on the ankles, right thigh, and waist of 12 healthy adults. Subjects performed a protocol in the laboratory including static orientations of standing, sitting, and lying down, and dynamic movements of walking, jogging, transitions between postures, and fidgeting to simulate free-living activity. When only one accelerometer was used, the thigh was found to be the optimal placement to identify both movement and static postures, with a misclassification error of 10%, and demonstrated the greatest accuracy for walking/fidgeting and jogging classification with sensitivities and positive predictive value (PPVs) greater than 93%. When two accelerometers were used, the waist-thigh accelerometers identified movement and static postures with greater accuracy than the thigh-ankle accelerometers (with a misclassification error of 11% compared to 17%). However, the thigh-ankle accelerometers demonstrated the greatest accuracy for walking/ fidgeting and jogging classification with sensitivities and PPVs greater than 93%. Movement can be accurately classified in healthy adults using tri-axial accelerometers placed on one or two of the following sites: waist, thigh, or ankle. Posture and transitions require an accelerometer placed on the waist and an accelerometer placed on the thigh.

The Relationship Between Skinfold Thicknesses and Mechanomyography At Different Locations on the Vastus Lateralis During Incremental Cycle Ergometry

Previous studies have indicated that skinfold thickness (SKF) acts as a low-pass filter to the mechanomyographic (MMG) signal. PURPOSE: The purpose of this study was to examine the effects of SKF at four different locations on the vastus lateralis (VL) muscle on MMG amplitude and mean power frequency (MPF) responses during incremental cycle ergometry. METHODS: Twenty adults (age ± SD = 23.8 ± 3.0 years) volunteered to participate in the investigation. An orientation session was performed to determine the location of the innervation zone (IZ) for the VL using a linear electrode array. The MMG signals from the VL muscle were detected during an incremental cycle ergometry test using four accelerometers placed on the right VL proximal (Prox1), over (Over IZ), and distal (Dist IZ) to the IZ. A fourth accelerometer was placed at the site recommended by the SENIAM Project (Prox2) for the placement of EMG electrodes. Prior to the test, three SKF were taken at each accelerometer placement site. The mean of the three SKF at each site were used for the analysis. The subjects were divided (at the median SKF value) into two groups of smaller (n = 10) and larger (n = 10) SKF values. Simple linear regression analyses were performed for SKF vs. MMG amplitude and MPF at each location for power outputs of 50, 75, 100, 125, 150, and 175W. Independent t-tests were also performed between the smaller and larger SKF groups for SKF measurements, MMG amplitude, and MMG MPF values at each accelerometer placement site. An alpha of p ≤ 0.05 was considered statistically significant for all analyses. RESULTS: For the 48 regression analyses, there were only 2 significant relationships (r = −0.461 and −0.519) between the SKF values and MMG MPF and 3 significant relationships between the SKF values and MMG amplitude (r = −0.535, −0.456, and −0.443). The independent t-tests showed that there were significant mean differences between the smaller and larger SKF groups for SKF measurements for all accelerometer placements sites. Of the 48 mean comparisons, only 1 was significantly different between the smaller and larger SKF groups for MMG MPF and 4 for MMG amplitude. CONCLUSIONS: The present study showed that, in general, SKF values were not correlated with the MMG signal. Only 10% of the regression analyses for SKF versus MMG MPF and MMG amplitude were statistically significant. In addition, only 10% of the mean comparisons between smaller and larger SKF groups for MMG MPF and MMG amplitude were significantly different. Thus, the results of the present study indicated that the larger SKF values did not attenuated the MMG signal. Other factors, including changes in force production and muscle length, as well as factors related to the accelerometer location, such as amount of muscle mass, muscle architecture, and/or motor unit territorial distribution may have had a greater effect on the MMG signal than SKF. PRACTICAL APPLICATIONS: These findings indicated that, in generall, smaller and larger SKF values did not correlate with the amplitude and frequency parameters of the MMG signal. Thus, there may be other anatomical, physiological, or methodological factors that have a greater effect on the MMG signal. These findings provide information about a methodological consideration necessary for the interpretation of the MMG signal.

The effects of accelerometer placement on mechanomyographic amplitude and mean power frequency during cycle ergometry

The purposes of this study were threefold: (1) to compare the power output related patterns of absolute and normalized MMG amplitude and MPF responses for proximal and distal accelerometer placements on the vastus lateralis (VL) muscle during incremental cycle ergometry; (2) to examine the influence of accelerometer placements on mean absolute MMG amplitude and MPF values; and (3) to determine the effects of normalization on mean MMG amplitude and MPF values from proximal and distal accelerometer placements. Fifteen adults (10 men and 5 women; mean ± SD age = 23.9 ± 3.1 years) performed incremental cycle ergometry tests to exhaustion. Two accelerometers were placed proximal and distal on the VL muscle. Paired t-tests indicated that absolute MMG amplitude values for the proximal accelerometer were greater ( p < 0.05) than the distal accelerometer at all power outputs. The normalized MMG amplitude also had greater values for the proximal accelerometer at all power outputs, except 50 W. There were no differences, however, between proximal and distal accelerometers for absolute MMG MPF, except at 75 W, and normalization eliminated this difference. Twenty-seven percent of the subjects exhibited different power output related patterns of responses between accelerometer placements for MMG amplitude and 47% exhibited different patterns for MPF. These findings indicated that normalization did not eliminate the influence of accelerometer placement on MMG amplitude and highlighted the importance of standardizing accelerometer placements to compare MMG values during cycle ergometry.

Comparative Force Outputs From Non-Centered Accelerometers During Loaded Barbell Jump Squats

The direct acquisition of ground reaction force data during isoinertial resistance tasks typically requires a force platform. As this is not always practical, alternative devices have been used to estimate dynamic force by directly measuring other variables and then using inverse dynamics to calculate it. For example, linear position and linear velocity transducers have been tethered to bars and used for this purpose. Recently, a dynamometer incorporating an accelerometer and no tether became available for this application. However, since barbell equipment inherently allows for asymmetrical movement through multiple planes, it follows that force output may differ depending upon bar orientation, especially deviations from horizontal. Furthermore, differences related to dynamometer placement and/or variations in the lifting movement pattern could be problematic if training-induced changes are to be monitored. To determine if peak force differed during barbell jump squats using non-centered accelerometer placements. 52 subjects (27 women & 25 men) with squat training experience were recruited from a university population. Subjects performed duplicate loadspectrum countermovement jump squats (CMJ) and static jump squats (SJ) (20%, 30% & 40% of their back squat 1RM) on two separate days in a counterbalanced sequence. A triaxial accelerometer device was directly affixed to each end of the barbell midway between the lateral-most aspect of the shoulder and the thumb side of each hand. Acceleration data were downsampled from 1.5 KHz to 500Hz by averaging every 3 data points. These data were then low-pass filtered (4th order Butterworth) with a cutoff frequency of 10Hz via proprietary software. Force was determined via inverse dynamics with mass considered the sum of body and barbell weights. Peak force data from the two dynamometers were always obtained from the same trial. Paired t-tests were used for data analysis. Dynamometers provided similar peak force values for both CMJ and SJ at 20% and 40% 1RM loads (p > .05) and different values at the 30% 1RM load (p < .01). Mean differences (20, 30, & 40%, respectively) for the CMJ were 9.8N, 15.1N, 8.1N and for the SJ were -1.5N, 15.4N, 5.9N (Table 1). For relatively light countermovement and static jump squats, dynamometer placement on both ends of barbell may result in differential peak force output. It appears that accelerometer placement during jump squats should either be as close as possible to the center of the bar or a squat machine with a horizontally-fixed bar should be used. Although not tested herein, it is likely that linear position and linear velocity transducers would have the identical constraints. As contiguous accelerometer placement was not tried herein, it is unclear what, if any, contribution the dynamometers made to measurement error.

A Comparison of Feature Extraction Methods for the Classification of Dynamic Activities From Accelerometer Data

Driven by the demands on healthcare resulting from the shift toward more sedentary lifestyles, considerable effort has been devoted to the monitoring and classification of human activity. In previous studies, various classification schemes and feature extraction methods have been used to identify different activities from a range of different datasets. In this paper, we present a comparison of 14 methods to extract classification features from accelerometer signals. These are based on the wavelet transform and other well-known time- and frequency-domain signal characteristics. To allow an objective comparison between the different features, we used two datasets of activities collected from 20 subjects. The first set comprised three commonly used activities, namely, level walking, stair ascent, and stair descent, and the second a total of eight activities. Furthermore, we compared the classification accuracy for each feature set across different combinations of three different accelerometer placements. The classification analysis has been performed with robust subject-based cross-validation methods using a nearest-neighbor classifier. The findings show that, although the wavelet transform approach can be used to characterize nonstationary signals, it does not perform as accurately as frequency-based features when classifying dynamic activities performed by healthy subjects. Overall, the best feature sets achieved over 95% intersubject classification accuracy.

Using the RT3 accelerometer to measure everyday activity in functionally impaired older people

Triaxial accelerometry may provide a simple measure of physical activity in older people, but the effect of different walking aids and accelerometer placements on measurement is not known. This study aimed to examine the effect of accelerometer placement, use of walking aids, and different types of physical activity on Stay- Healthy RT3 triaxial accelerometer readings in older people. Twenty subjects aged over 65 years and five younger volunteers were recruited from Medicine for the Elderly services. Subjects performed six minutes each of standardized standing activity, sitting activity, sitting at rest, walking, and stair climbing. Counts generated from RT3 accelerometers worn anteriorly over both hips were recorded in subjects using different walking aids during these standardized activities. There were significant differences between counts generated by the left and right hip positions. The intraclass correlation coefficient of RT3 counts between left and right hip positions was 0.48, 0.39 and 0.99 for sedentary tasks (standing, sitting and rest), stair and walking tasks respectively. Counts ranged between 250-3000 min-1 during the walking task. Counts were proportional to the distance walked. Resting, sitting or standing all generated counts below 250 min-1, but there was no clear demarcation between these activities. The use of different walking aids did not affect the counts generated for any activity. Walking can be distinguished from other activities by upper and lower cutoffs. The RT3 accelerometer should be used on the same side of the body. Different walking aids do not appear to affect RT3 counts in older people.

Validity And Inter-instrument Reliability Of The Actigraph Accelerometer In 3–5 Year Old Children

Accelerometers have been used to objectively measure physical activity in several populations. However, the validity and reliability of these instruments has not been well-established in young children. In order to accurately assess activity in this age group, the validity and reliability of accelerometers needs to be examined. PURPOSE The purpose of this study was to examine the correlation between VO2 and Actigraph accelerometer counts for two different accelerometer placements and to examine the reliability between two monitors during structured and unstructured activities. METHODS Thirty children (age = 4.4 ± 0.75, 53% female, 57% African American, BMI = 16.5 ± 2.1) wore a portable metabolic analyzer (Cosmed K4b2; Rome, Italy) and two Actigraph accelerometers (MTI; Fort Walton Beach, FL); one placed at the hip and one placed on the back. The children wore the equipment during rest, three structured activities, and during inside and outside freeplay times. Pearson correlations were used to determine the relationship between VO2 and accelerometer counts, and to determine the inter-instrument reliability. RESULTS The correlation coefficient for the Actigraph at the hip placement was significantly higher (r=0.78; p=0.018) than the back placement (r=0.68) during the rest and structured activities. During freeplay, the correlation between VO2 and counts was r=0.35 for the hip and r=0.45 for the back. During both structured and unstructured activities, the monitors correlated with each other (r=0.73 and 0.78, respectively). CONCLUSIONS Structured and unstructured activities yielded different correlations between VO2 and counts. Most research in this area has been conducted in laboratory settings (structured). It is therefore necessary to further investigate instrument reliability in field settings (unstructured) in this population and others. Supported by an ILSI Research Foundation grant and NIH grant 5R01HD043125.

Vibration monitoring for damage in offshore platforms

A number of vibration‐based techniques have been considered for the detection of underwater structural damage in fixed offshore platforms for oil and gas production. Most platforms are primarily composed of welded steel pipes, and are pile anchored to the sea floor in depths of up to 320 m (1050 ft). The standard inspection technique employs divers to visually survey for severance or cracking of members. This is costly, can be hazardous, and is subject to environment constraints. Various vibration instrumentation approaches are described. All employ accelerometers to measure ambient excited or forced vibration. The crucial issues are (1) the ability to reliably detect damage before there is a significant loss of structural integrity, (2) the avoidance of false alarms due to nondamage changes, and (3) the masking of the desired information by machinery and other noise. A highly promising new technique, soon to be field tested, employs accelerometer placements down special chutes to reach various underwater positions for detection of overall shear stiffness of the individual platform bays. Results are described for a USGS/ONR sponsored blind laboratory test program in which several techniques were evaluated.