The association between working memory precision and the nonlinear dynamics of frontal and parieto-occipital EEG activity

Abstract

Electrophysiological working memory (WM) research has shown that distinct brain areas communicate through macroscopic oscillatory activities across multiple frequency bands. Such cross-frequency interactions generate nonlinear amplitude modulations (AM) in the observed signal. Traditionally, the AM of a signal is expressed as coupling strength between the signal and a pre-specified modulator at a lower frequency. Therefore, the idea of AM and coupling cannot be separately studied. This EEG study shows that the AM of parieto-occipital alpha/beta power and the coupling between frontal theta phase and parieto-occipital alpha/beta AM provide different information on WM processing. Thirty-three participants completed a color recall task with simultaneous EEG recording. The results showed that individual differences in WM precision are associated with frontal theta power enhancement and parieto-occipital alpha/beta power suppression. Furthermore, the AM of parieto-occipital alpha/beta power predicted WM precision after presenting a target-defining probe array. The phase-amplitude coupling (PAC) between frontal theta phase and parieto-occipital alpha/beta AM increased with WM load during the processing of incoming stimuli, but they did not predict the subsequent recall performance. These results indicate that the frontoparietal PAC reflects the executive control for selecting relevant WM representations, but whether the memorized information can be retrieved depends on the subsequent amplitude variation of parieto-occipital alpha/beta power. In conclusion, individuals with higher working memory precision are associated with enhanced frontal theta power and parieto-occipital alpha/beta power suppression.

Certain spatial prediction decreases the rhythm of attentional sampling

“…However, linear algorithms have limitations in accurately analyzing nonstationary and nonlinear brain signals, particularly the instantaneous phase (Aru et al 2015;Cole and Voytek 2017). Therefore, the current study utilized the nonlinear Holo-Hilbert Spectral Analysis (HHSA) method, with the first step of the multiple-layer Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) method, to analyze the preprocessed EEG segments to enhance the understanding of neural oscillations (Chang et al 2023;Huang et al 2016;Juan et al 2021;Liang et al 2021).…”

“…Studies employing nonlinear decomposition and Hilbert transform, particularly the Holo-Hilbert Spectral Analysis (HHSA), have demonstrated enhanced accuracy in analyzing phaseamplitude modulated brain signals entrained from phase-amplitude coupled external visual stimuli (Juan et al 2021;Nguyen et al 2019). Additionally, adaptive neural correlates optimizing behavioral performance in attentional tasks, such as inter-site phase coherence (ISPC) and inter-trial phase coherence (ITPC), can be quantified through precise phase extraction (Chang et al 2023;Liang et al 2021).…”