Detecting and tracking tremor in spike trains using the rectangular model based extended Kalman smoother

Tremor is one of the most disabling symptoms in patients with movement disorders such as Parkinson's disease (PD) and essential tremor (ET). Spike trains extracted from microelectrode recordings are used to study the relationship of tremor exhibited by neuronal signals to physical tremor as measured with electromyograms (EMG), gyroscopes, or accelerometers. We describe a new method for continuously tracking the instantaneous tremor frequency and amplitude in spike trains based on a new state-space model and the extended Kalman smoother. This method can be used to detect periods of statistically significant tremor in recordings with intermittent tremor.