Decomposition of motor unit firing pattern using Kalman filtering

Interdischarge interval (IDI) is one of the basic parameters to study in motor unit firing analysis. Discharge intervals of a single motor unit vary over time and they are unpredictable. IDI sequences can be considered to comprise two components, namely a long term signal, an IDI trend, and a white noise process, instantaneous firing variability (IFV). In this paper a stochastic model of the IDI signal has been developed in order to estimate the elements of an IDI sequence. IDI sequences of several patients have been recorded at a clinic and a Kalman filter has been constructed based on the developed stochastic model. The Kalman filter is utilized to decompose the recorded IDI sequences into the IDI trend and IFV components. The obtained decomposed signals, especially the IDI trend component, may provide valuable information on motor unit firing performance and help diagnose neurological diseases.