AUTHOR=Yan Yiping , Chen Qingguo 

TITLE=Energy Expenditure Estimation of Tabata by Combining Acceleration and Heart Rate

JOURNAL=Frontiers in Public Health

VOLUME=Volume 9 - 2021

YEAR=2022

URL=https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2021.804471

DOI=10.3389/fpubh.2021.804471

ISSN=2296-2565

ABSTRACT=In order to estimating energy expenditure (EE) during Tabata exercise, we use accelerometer magnitude (ACC) and heart rate (HR) simultaneously to build Tabata exercise expenditure prediction mode. Participants are invited from June 15 to July 15, 2021 (n = 42; Mean age：28.5±11.6 years), including healthy adults who engaged in recreational activities. Each participants simultaneously wore four accelerometers (dominant hand ACC, non-dominant hand ACC, fight hip ACC, right ankle ACC), a heart rate band (HR) and a metabolic measurement system to complete Tabata exercise. BP neural network and Linear regression Tabata energy expenditure prediction model were established based on acceleration and heart rate data (ACC-HR ). Analysis of bland-Altman plots with 95% confidence intervals showed no significant difference in EE across measurements using HR-ACC linear regression and BP neural network estimates. The MAPE of ACC-HR neural network was significantly lower than ACC-HR linear regression, which were 12.6% and 14.7% respectively. Using both accelerometer outputs and HR for EE estimation during Tabata exercise is feasible, and the machine learning algorithm based on big data is better.