Evaluation of the Motus wearable-sensor-based system to classify postures and movements in 3-14 aged children

Background: Recent advances in wearable sensor technology may enable robust systems for measurement of children’s postures and movement (e.g. sitting, walking, and running). Such systems can assist in understanding the impact of various posture and movement on children’s health and wellbeing. Motus, a wearable sensor-based system for surveillance of postures and movements, has shown high accuracy among adults. However, its accuracy among children is unknown.

Purpose: This study aimed to evaluate the accuracy of Motus to measure common postures and movements among children between 3-14 years old in a laboratory setting.

Method: Data were collected on 48 children who attended a structured laboratory-based ~1-hour data collection session. The session was video recorded and thigh acceleration was measured using a SENS accelerometer. Data from the accelerometer were processed and classified into six postures and movements using the Motus software. Human-coded video analysis provided the gold standard to calculate sensitivity, specificity, and balanced accuracy.

Results: We observed high sensitivity, specificity, and balanced accuracy for classifying lying, sitting, standing, walking, and running (balanced accuracy ranging between 72.0-98.4%). The lowest balanced accuracy was observed for classifying stair climbing (65.7%).
Conclusion: Motus showed high balanced accuracy for detecting lying, sitting, standing, and running among children. However, the system could be improved for classifying stair climbing and developed further to measure more child-specific postures and movements.

Practical implications: We found that Motus could be a suitable tool for measuring lying, sitting, standing, walking, and running among children. Future work is planned to test the feasibility of using Motus to measure postures and movements among children in free-living settings.

Funding: This study was partly funded by the Australian Research Council through the ARC Centre of Excellence for the Digital Child, grant number CE200100022 and the Curtin School of Allied Health 2022 Teaching and Research Grant.