The PA levels were assessed with the MOX activity monitor (MOX; Maastricht Instruments B.V., the Netherlands (Figure 1)). The device contained a tri-axial accelerometer sensor (ADXL362; Analog Devices, Norwood, MA, USA) in a small waterproof housing (35 × 35 × 10 mm, 11 g). Raw acceleration data (±8 g) were measured in three orthogonal sensor axes (X, Y and Z) at a 25 Hz sampling rate. The accelerometer was factory calibrated against gravity for each axis. The raw acceleration data were converted to a PA classification using a previously described embedded algorithm [34 (link)]. After sensor noise reduction, the data were segmented in to one-second long windows using a fixed non-overlapping sliding window. Based on the amount of activity, each window was classified as dynamic or static. For the static windows, the sensor orientation was assessed. Based on a cut-off value of 0.8 g the static windows were classified as standing or sedentary. Each minute the classified results were sent to the Hospital Fit smartphone application via a Bluetooth protocol.
Free full text: Click here