NIH BREATHE mHealth pediatric activity classification dataset

The Biomedical REAl-Time Health Evaluation (BREATHE) platform is part of the Los Angeles PRISMS (Pediatric Research with Integrated Monitoring Systems) NIH-funded Center, which focuses on informatics platforms for mHealth (NIH/NIBIB U54 EB022002; PI: Alex Bui). This dataset was developed to better establish models for child activity classification based on smartwatch sensors. Specifically, a Motorola 360 smartwatch was worn by 21 subjects (age ranged 11-18), generating annotated data across different sensors (3-axis accelerometry, gyroscope, heart rate) for 6 different types of physical activity:three static postures (standing, sitting, lying) and three dynamic activities (walking, walking downstairs and walking upstairs). The study was motivated by the fact that current activity classifiers are typically geared towards adult bodies and patterns of motion; children exhibit different motion patterns, particularly as they grow. This dataset is not intended to be comprehensive, but rather a starting point for exploring differences in motion patterns across children (and relative to adults).