Prediction of activity type in preschool children using machine learning techniques

Hagenbuchner, Markus; Cliff, Dylan P.; Trost, Stewart G.; Tuc, Nguyen Van; Peoples, Gregory E

doi:10.1016/j.jsams.2014.06.003

Cited by 54 publications

(41 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, a close relationship is likely to exist between higher vertical impacts and lower limb muscle function, which we are planning to explore further based on contemporaneous jumping mechanography data collected in COSHIBA and MAC. As well as classification of PA according to impact level, the raw accelerometry trace provides opportunities to extract other important PA characteristics, such as duration of specific activities [27]. …”

Section: Discussionmentioning

confidence: 99%

A novel accelerometer-based method to describe day-to-day exposure to potentially osteogenic vertical impacts in older adults: findings from a multi-cohort study

et al. 2016

View full text Add to dashboard Cite

SummaryThis observational study assessed vertical impacts experienced in older adults as part of their day-to-day physical activity using accelerometry and questionnaire data. Population-based older adults experienced very limited high-impact activity. The accelerometry method utilised appeared to be valid based on comparisons between different cohorts and with self-reported activity.IntroductionWe aimed to validate a novel method for evaluating day-to-day higher impact weight-bearing physical activity (PA) in older adults, thought to be important in protecting against osteoporosis, by comparing results between four cohorts varying in age and activity levels, and with self-reported PA levels.MethodsParticipants were from three population-based cohorts, MRC National Survey of Health and Development (NSHD), Hertfordshire Cohort Study (HCS) and Cohort for Skeletal Health in Bristol and Avon (COSHIBA), and the Master Athlete Cohort (MAC). Y-axis peaks (reflecting the vertical when an individual is upright) from a triaxial accelerometer (sampling frequency 50 Hz, range 0–16 g) worn at the waist for 7 days were classified as low (0.5–1.0 g), medium (1.0–1.5 g) or higher (≥1.5 g) impacts.ResultsThere were a median of 90, 41 and 39 higher impacts/week in NSHD (age 69.5), COSHIBA (age 76.8) and HCS (age 78.5) participants, respectively (total n = 1512). In contrast, MAC participants (age 68.5) had a median of 14,322 higher impacts/week. In the three population cohorts combined, based on comparison of beta coefficients, moderate-high-impact activities as assessed by PA questionnaire were suggestive of stronger association with higher impacts from accelerometers (0.25 [0.17, 0.34]), compared with medium (0.18 [0.09, 0.27]) and low impacts (0.13 [0.07,0.19]) (beta coefficient, with 95 % CI). Likewise in MAC, reported moderate-high-impact activities showed a stronger association with higher impacts (0.26 [0.14, 0.37]), compared with medium (0.14 [0.05, 0.22]) and low impacts (0.03 [−0.02, 0.08]).ConclusionsOur new accelerometer method appears to provide valid measures of higher vertical impacts in older adults. Results obtained from the three population-based cohorts indicate that older adults generally experience very limited higher impact weight-bearing PA.Electronic supplementary materialThe online version of this article (doi:10.1007/s00198-016-3810-5) contains supplementary material, which is available to authorized users.

show abstract

Section: Discussionmentioning

confidence: 99%

A novel accelerometer-based method to describe day-to-day exposure to potentially osteogenic vertical impacts in older adults: findings from a multi-cohort study

et al. 2016

View full text Add to dashboard Cite

show abstract

“…Most previous studies on activity classification did not apply LOSO cross-validation (13, 19, 23, 28, 29). A few studies did use this type of cross-validation (9, 10, 12), but they limited their testing and evaluation to a homogeneous pool of healthy adult users, with the exception of Del Rosario et al who involved two age groups: 37 elderly (average age 84 yo) and 20 young adults (average age 22 yo) (10). …”

Section: Methodsmentioning

confidence: 99%

“…An overall accuracy of 88.4 % from a hip-worn sensor was reported. The most recent paper to include 1 s epoch counts processing for activity recognition was by Hagenbuchner et al who involved 11 pre-school children (3 to 6 yo, with a total of 264 minutes of classified data) (12). Four classes were recognized ( sedentary activities , light activities , moderate to vigorous activities , walking and running ) reporting 82.6% accuracy.…”

Section: Introductionmentioning

confidence: 99%

Activity Recognition in Youth Using Single Accelerometer Placed at Wrist or Ankle

Mannini

Rosenberger

Haskell

et al. 2017

Medicine &Amp; Science in Sports &Amp; Exercise

View full text Add to dashboard Cite

Purpose State-of-the-art methods for recognizing human activity using raw data from body worn accelerometers have primarily been validated with data collected from adults. This study applies a previously available method for activity classification using wrist or ankle accelerometer to work on datasets collected from both adults and youth. Methods An algorithm for detecting activity from wrist-worn accelerometers, originally developed using data from 33 adults, is tested on a dataset of 20 youth (age 13±1.3). The algorithm is also extended by adding new features required to improve performance on the youth dataset. Subsequent tests on both the adult and youth data were performed using crossed tests (training on one group and testing on the other) and leave-one-subject-out cross-validation. Results The new feature set improved overall recognition using wrist data by 2.3% for adults and 5.1% for youth. Leave-one-subject-out cross-validation accuracy performance was 87.0% (wrist) and 94.8% (ankle) for adults, and 91.0% (wrist) and 92.4% (ankle) for youth. Merging the two datasets, overall accuracy was 88.5% (wrist) and 91.6% (ankle). Conclusions Previously available methodological approaches for activity classification in adults can be extended to youth data. Including youth data in the training phase and using features designed to capture information on the peculiar activity fragmentation of young participants allows a better fit of the methodological framework to the characteristics of activity in youth, improving its overall performance. The proposed algorithm differentiates ambulation from sedentary activities that involve gesturing in wrist data, such as that being collected in large surveillance studies.

show abstract

“…7 Eleven children (5 girls, 6 boys) aged 3 to 6y (mean age = 4.8 ± 0.9y; mean BMI = 15.9 ± 1.0 kg/m 2 ; see Table 1) were recruited from the Illawarra region of New South Wales in Australia from April to November 2013. Children were ineligible if they had a disease known to influence their energy balance (for example McArdle's disease), had a physical disability, or asthma, or were claustrophobic.…”

Section: Participantsmentioning

confidence: 99%

Energy Cost of Physical Activities and Sedentary Behaviors in Young Children

Groβek

Loo²,

Peoples³

et al. 2016

Journal of Physical Activity and Health

Self Cite

View full text Add to dashboard Cite

Background: This study reports energy expenditure (EE) data for lifestyle and ambulatory activities in young children. Methods: Eleven children aged 3 to 6 years (mean age = 4.8 ± 0.9; 55% boys) completed 12 semistructured activities including sedentary behaviors (SB), light (LPA), and moderate-to-vigorous physical activities (MVPA) over 2 laboratory visits while wearing a portable metabolic system to measure EE. Results: Mean EE values for SB (TV, reading, tablet and toy play) were between 0.9 to 1.1 kcal/min. Standing art had an energy cost that was 1.5 times that of SB (mean = 1.4 kcal/min), whereas bike riding (mean = 2.5 kcal/min) was similar to LPA (cleaning-up, treasure hunt and walking) (mean = 2.3 to 2.5 kcal/min), which had EE that were 2.5 times SB. EE for MVPA (running, active games and obstacle course) was 4.2 times SB (mean = 3.8 to 3.9 kcal/ min). Conclusion: EE values reported in this study can contribute to the limited available data on the energy cost of lifestyle and ambulatory activities in young children.

show abstract

Prediction of activity type in preschool children using machine learning techniques

Cited by 54 publications

References 14 publications

A novel accelerometer-based method to describe day-to-day exposure to potentially osteogenic vertical impacts in older adults: findings from a multi-cohort study

A novel accelerometer-based method to describe day-to-day exposure to potentially osteogenic vertical impacts in older adults: findings from a multi-cohort study

Activity Recognition in Youth Using Single Accelerometer Placed at Wrist or Ankle

Energy Cost of Physical Activities and Sedentary Behaviors in Young Children

Contact Info

Product

Resources

About