Wearable computing refers to a new class of measurement devices that can be attached to the human body to measure human activity, biosignals, or environmental factors in real time. Examples of wearable devices are accelerometers, which can measure acceleartion at the hip or wrist, heart monitors, which measure heart rate or systolic blood pressure, and actigraphs, which measure the number of activities per unit of time [SMART].
My research is focused on high-density, high-throughoutput accelerometry data collected at the sub-second level in the free-living and laboratory environments. An example of such data is shown below. Data were collected by me using an IPhone camera for video and an Actigraph GT3X+ accelerometer placed on my left (non-dominant) wrist. The accelerometer collected tri-axial data (hence the three different time series in the right upper panel) at 30 Hz (30 observations per second per axis). You may notice that I am clapping my hands, which corresponds to a higher amplitude signal. This was used to synchronize the video and the accelerometry data at the sub-second level. This technique is now widely used in our studies to better synchronize various devices.
The top-right panel indicates that sustained harmonic walking (a term I invented) is characterized by periodic signals associated with the biomechanics of walking. In large studies of subjects in the free living environment we have neither the synchronized video nor another clear indication of walking periods. My primary objective is to identify these periods and characterize the micro-scale fluctuations in the signal during walking. I am also interested in validating these walking characteristics using standardized tests as well as health outcomes.
The figure displays an example of raw, tri-axial, accelerometry data collected during walking (top panel) together with the corresponding instantaneous acceleration level (middle panel) and instantaneous cadence (bottom panel). These micro-scale walking features are strongly related to physical performance, fatigability and fitness and are complementing in-lab tests. Extracting such features from the free-living environment is difficult, but could provide previously unknown insights into realized human activity. This is especially important for subjects who can perform in-lab tests, but are quite inactive as well as for subjects who under-perform, but are quite active in their daily routine.