EEG Brain mapping based on the Duffing oscillator

Abstract

This study proposes a Duffing oscillator based measurement framework for detecting frequency components associated with postsynaptic potential related activity in EEG recordings. The Duffing oscillator is employed as a nonlinear measurement system whose high sensitivity to weak periodic inputs enables robust frequency estimation in the presence of strong noise. Because postsynaptic potentials are embedded in noisy EEG recordings, they are treated as weak signals. Based on the detected weak-signal frequency distributions across EEG channels, topographic maps of brain activity are constructed to visualize spatial variations in neural activation. For this purpose, EEG signals recorded from two musicians and two audience members during a music experiment were analysed. The frequency components of the weak signals were searched within the 4 37 Hz range across all EEG channels. Subsequently, topographic brain maps were generated according to the number of detected weak-signal components. The results indicate that Duffing oscillator based weak signal detection is an effective tool for EEG brain mapping. Compared with Fourier and wavelet-based methods, the proposed technique provides a more detailed representation of brain activation. These findings suggest that the proposed approach has potential for investigating various pathological conditions and for enhancing the understanding of cognitive and behavioural functions of the human brain.