Filtering voluntary motion for pathological tremor compensation

This paper describes an algorithm to estimate tremor and voluntary motion from measured motion data. Estimation is performed by means of an extended Kalman filter (EKF), which also estimates tremor parameters. Comparison of the proposed method with techniques described in the literature are conducted with two experimental data sets from tremor patients performing the same task, drawing a spiral. The presented algorithm may be directly applied in real-time pathological tremor compensation systems.