Prediction of oxygen uptake dynamics by machine learning analysis of wearable sensors during activities of daily living

Currently, oxygen uptake ([inline-graphic not available: see fulltext]) is the most precise means of investigating aerobic fitness and level of physical activity; however, [inline-graphic not available: see fulltext] can only be directly measured in supervised conditions. With the advancement of new wearable sensor technologies and data processing approaches, it is possible to accurately infer work rate and predict [inline-graphic not available: see fulltext] during activities of daily living (ADL). The main objective of this study was to develop and verify the methods required to predict and investigate the [inline-graphic not available: see fulltext] dynamics during ADL. The variables derived from the wearable sensors were used to create a [inline-graphic not available: see fulltext] predictor based on a random forest method. The [inline-graphic not available: see fulltext] temporal dynamics were assessed by the mean normalized gain amplitude (MNG) obtained from frequency domain analysis. The MNG provides a means to assess aerobic fitness. The predicted [inline-graphic not available: see fulltext] during ADL was strongly correlated (r = 0.87, P < 0.001) with the measured [inline-graphic not available: see fulltext] and the prediction bias was 0.2 ml·min−1·kg−1. The MNG calculated based on predicted [inline-graphic not available: see fulltext] was strongly correlated (r = 0.71, P < 0.001) with MNG calculated based on measured [inline-graphic not available: see fulltext] data. This new technology provides an important advance in ambulatory and continuous assessment of aerobic fitness with potential for future applications such as the early detection of deterioration of physical health.