Monitoring Prosthesis User Activity and Doffing Using an Activity Monitor and Proximity Sensors

Introduction Characterizing a prosthesis user's activity in the field may provide information relevant to his or her clinical care. The purpose of this research was to incorporate doffing and donning information into activity characterization. Materials and Methods Commercial accelerometers were positioned on the thigh and lower leg, and proximity sensors on the medial and lateral brim of the socket. A novel data-processing algorithm converted collected data into a time-stamped list of postures and activities: doffed, walking (included all types of movement), standing, and sitting. Results Controlled testing on 11 people with transtibial amputation demonstrated that activity durations and transitions between activities well matched hand notes recorded by the researcher accompanying the participant during the test. Testing on 23 participants with transtibial amputation in unrestricted field testing demonstrated that detected activities well matched participants' descriptions of activities upon return to the laboratory. At least 95% of doffs were correctly detected. Conclusions The developed technology is relevant towards the design of control systems to automatically adjust the prosthesis according to the user's activity and posture.