We aimed to improve standard algorithms to use processed accelerometer data collected in one minute epochs from a wrist actigraph to measures sleep and wake.
We reanalyzed a subset of data collected by the Study of Osteoporotic Fractures Research Group.
Participants wore a wrist actigraph while undergoing PSG recording during one night of sleep in a lab. Our goal is to build a sleep algorithm using the actigraph data that accurately detects sleep and wake better than standard algorithms. The results of our algorithm were then compared to the PSG recording of sleep which is the gold standard for sleep recording.
Information about the data: https://www.ncbi.nlm.nih.gov/pubmed/18274276
Please examine the Sherman sleep algorithm.pdf for details and preliminary results