Application of adaptive Savitzky–Golay filter for EEG signal processing

Abstract

A Savitzky–Golay filter typically requires pre-determined values of order and frame size for its fabrication. Generally, a random hit-and-trial method or prior experience is required to determine the suitable values of design parameters. However, the proposed adaptive Savitzky–Golay filter aims to provide a generic framework for optimal design of filter vis-à-vis the order and frame size of the filter. The algorithm uses all the possible combinations of these parameters in a certain range and the correlation coefficient is evaluated in each case to measure the filter efficiency. The parameters which provide the highest correlation coefficient are considered for filter design. In this paper the relative advantages of adaptive Savitzky–Golay filter over the standard models are also discussed. The proposed adaptive model of Savitzky–Golay filter is successfully tested for EEG signal processing.