Energy expenditure estimation using triaxial accelerometry and barometric pressure measurement.

Energy expenditure (EE) is a parameter of great relevance in studies involving the assessment of physical activity. However, most reliable techniques for EE estimation are impractical for use in free-living environments, and those which are practically useful often poorly track EE when the subject is working to change their altitude, for example when ascending or descending stairs or slopes. The aim of this study is to evaluate the utility of adding barometric pressure related features, as a surrogate measure for altitude, to existing accelerometry related features to estimate the subject's EE. The EE estimation system described is based on a triaxial accelerometer (triax) and a barometric pressure sensor. The device is wireless, with Bluetooth connectivity for data retrieval, and is mounted at the subject's waist. Using a number of features extracted from the triax and barometric pressure signals, a linear model is trained for EE estimation. This EE estimation model is compared to its counterpart, which solely utilizes accelerometry signals. A protocol (comprising lying, sitting, standing, walking phases) was performed by 13 healthy volunteers (8 male and 5 female; age: 23.8 ± 3.7 years; weight: 70.5 ± 14.9 kg), whose instantaneous oxygen uptake was measured by means of an indirect calorimetry system. The model incorporating barometric pressure information estimated the oxygen uptake with the lowest mean square error of 4.5 ± 1.7 (mlO(2).min(-1).kg(-1))(2), in comparison to 7.1 ± 2.3 (mlO(2).min(-1).kg(-1))(2) using only accelerometry-based features.