Exercise energy expenditure estimation based on acceleration data using the linear mixed model

This paper introduces a novel algorithm for estimating energy expenditure during physical activity. The estimation is based on acceleration data measured from a wrist-worn accelerometer. Simultaneous measurements of acceleration and oxygen consumption using a biaxial accelerometer and a breath gas analyzer were made during four different activities: walking, running, Nordic walking and bicycling. A variance feature is used to compress the original acceleration signals. A linear mixed model is fitted to the data to estimate oxygen consumption based on the acceleration data. Lagged values of acceleration are used to take the delayed effect of physical activity on oxygen consumption into consideration. The algorithm also uses information on the height of the subjects. Oxygen consumption is estimated at 15-second intervals and energy expenditure is directly calculated from the oxygen consumption. Based on the experimental data gathered from 10 subjects, a new algorithm for estimating energy expenditure is suggested. It is shown that the method estimates energy expenditure very accurately. In walking, running and Nordic walking the model underestimates energy expenditure by 13, 2 and 9 percent, respectively, and in bicycling energy expenditure is overestimated by 7 percent. Thus, the new approach is a very promising method for estimating energy expenditure.