Inertial sensors for assessment of joint angles

In this paper, two approaches for estimating joint angles are presented: the first one uses only an accelerometer to estimate hip and knee angles on static postures, whereas the second approach fuses information from multiple sensors to dynamically calculate leg range of motion. Static joint angle results were evaluated against goniometer, which is used as a standard at clinics. Dynamic angles results were compared with Kinect joint segmentation, as well as a custom marker-based video tracking algorithm for angle estimation. For both approaches, the errors obtained when compared with reference systems were very low. Results support that inertial sensors are suitable for use in home-based rehabilitation exercises.