QUANTITATIVE-ANALYSIS OF TREMOR TIME-SERIES

Spectral analysis is applied to tremor time series in basic research a nd treatment monitoring. The estimation of the spectra from the data i s usually done by averaging the squared modulus of the Fourier transfo rm of segments of the data. We discuss drawbacks of this method and pr opose an alternative procedure to estimate the spectra adaptively base d on the data. Thus, the method can be applied to all types of tremor. Applying the theory of spectral estimation, we propose a method to de cide whether a spectrum exhibits multiple significant peaks and discus s different approaches to determine the amplitude of the tremor from t he spectrum.