Bio-feedback system analysis in walking using three gyro sensors

The sequential movements of the head, shoulder and pelvis cooperate and influence each other during walking. It is very important to evaluate the bio-feedback relationship of each movement in rehabilitation treatment. This is useful for diagnosing, treating and preventing functional disorders of movement. In this study, three PC assisted mini gyro sensors were placed on top of the head, the trunk at shoulder height, and the pelvis. The time series data of three angles (roll, pitch, yaw) on each part were measured while the subjects were walking on the treadmill. Subjects are hemiparesis caused by a traumatic brain injury. It is difficult to analyze the bio-feedback system of the movements of each part directly from the original data. This study presents a new approach utilizing a multivariate autoregressive (AR) model for feedback system analysis. The relative power contribution (RPC) from three variables, the main joints of the human body i.e., the head, shoulder and pelvis, shows their dynamic relationship.