| 研究実績の概要 |
Background: The self-monitoring of physical activity is one effective strategy help people maintaining active behaviors and healthy lifestyles. Activity trackers have been the object of numerous improvements during the past decade. Still, the accurate quantification of human activity remains a challenge. The project consists in developing a floor vibration-based system to improve the continuity and accuracy of energy expenditure predictions, when people stay at home and perform housework activities.
Method: The floor vibration sensing system developed in the previous years has been improved and installed to cover the whole surface of the OchaHouse, an experimental smart-home located in central Tokyo. The system consists in 8 high sensitivity sensors for a 50m2 surface. 10 subjects performed the 4 following activities: watching TV, ironing, cooking, and cleaning the room (respectively 1.3, 1.8, 2.5, and 3.3MET). Floor vibration data have been collected. An original algorithm has been developed to compute a quantitative index of physical activity (10-sec epoch).
Results: The floor vibration-based physical activity index increases accordingly to the theoretical intensity of each experimental activity. The index is significantly different between activities (53, 183, 327 and 608 count, p<0.001). In addition, a significant correlation between the quantitative floor vibration index and the activity-count of a waist-worn Actigraph monitor (0.8, p<0.001) is noted.
Conclusion: floor vibration monitoring can produce valid quantifications of indoor physical activity.
|
