Estimation of human movements from body acceleration monitoring for ubiquitous health care

We are developing a BAN (body area network) using a machine measuring human 3D acceleration and electrocardiogram communicating to a PC by wireless during 24-hours. The purpose of our study is human healthy watching to alert to himself (herself) or care staffs by network if his (her) condition is bad, and not to alert if good. Our system estimate a person's bodily movement (walking, running, standing in a house, sitting in a train or else) from his/her acceleration data, and triage the electrocardiogram according to the estimated bodily movement then classify a human health condition. We collect acceleration and bodily movement records of our staffs and voluntary people for correct estimations and correct triages.