Adaptive State-Space Multitaper Spectral Estimation

TL;DR

The paper introduces an adaptive extension to the state-space multitaper spectral estimation method, enabling better analysis of highly nonstationary time series such as EEG data by adaptively updating parameters for improved denoising.

Contribution

It proposes an adaptive, time-varying version of the SSMT method that uses exponential smoothing to track nonstationary spectral dynamics more effectively.

Findings

ASSMT improves denoising over standard methods

ASSMT captures nonstationary spectral changes more accurately

Enhanced analysis of EEG data with ASSMT

Abstract

Short-time Fourier transform (STFT) is the most common window-based approach for analyzing the spectrotemporal dynamics of time series. To mitigate the effects of high variance on the spectral estimates due to finite-length, independent STFT windows, state-space multitaper (SSMT) method used a state-space framework to introduce dependency among the spectral estimates. However, the assumed time-invariance of the state-space parameters makes the spectral dynamics difficult to capture when the time series is highly nonstationary. We propose an adaptive SSMT (ASSMT) method as a time-varying extension of SSMT. ASSMT tracks highly nonstationary dynamics by adaptively updating the state parameters and Kalman gains using a heuristic, computationally efficient exponential smoothing technique. In analyses of simulated data and real human electroencephalogram (EEG) recordings, ASSMT showed improved denoising and smoothing properties relative to standard multitaper and SSMT approaches.