Activity energy expenditure assessment system based on activity classification using multi-site triaxial accelerometers

Abstract

Wireless networked multisite triaxial accelerometry system was proposed to estimate activity energy expenditure during daily life. Proposed system used different estimation algorithms according to activity classification based on measured acceleration signals. The estimation results showed higher correlation (R2 = 0.98) and smaller standard errors of estimation (SEE = 0.66 kcal) than any other system in preceding reports.