Fine Structure of Posterior Alpha Rhythm in Human EEG: Frequency Components, Their Cortical Sources, and Temporal Behavior

Abstract

Heterogeneity of the posterior alpha rhythm (AR) is a widely assumed but rarely tested phenomenon. We decomposed the posterior AR in the cortical source space with a 3-way PARAFAC technique, taking into account the spatial, frequency, and temporal aspects of mid-density EEG. We found a multicomponent AR structure in 90% of a group of 29 healthy adults. The typical resting-state structure consisted of a high-frequency occipito-parietal component of the AR (ARC1) and a low-frequency occipito-temporal component (ARC2), characterized by individual dynamics in time. In a few cases, we found a 3-component structure, with two ARC1s and one ARC2. The AR structures were stable in their frequency and spatial features over weeks to months, thus representing individual EEG alpha phenotypes. Cortical topography, individual stability, and similarity to the primate AR organization link ARC1 to the dorsal visual stream and ARC2 to the ventral one. Understanding how many and what kind of posterior AR components contribute to the EEG is essential for clinical neuroscience as an objective basis for AR segmentation and for interpreting AR dynamics under various conditions, both normal and pathological, which can selectively affect individual components.

No changes in parieto‐occipital alpha during neural phase locking to visual quasi‐periodic theta‐, alpha‐, and beta‐band stimulation

“…Also note that the EEG is likely dominated by a strong parieto‐occipital alpha generator. However, recent research suggests more than one concurrent alpha rhythmic “mode” involved in visual processing (Barzegaran, Vildavski, & Knyazeva, ; Keitel & Gross, ). It is therefore possible that our stimulation failed to influence the dominant rhythm but affected a less prominent alpha component that is more difficult to observe given the sparse spatial resolution of EEG recordings.…”

No changes in parieto-occipital alpha during neural phase locking to visual quasi-periodic theta-, alpha-, and beta-band stimulation

“…Also note that the EEG is likely dominated by a strong parieto-occipital alpha generator. However recent research suggests more than one concurrent alpha rhythmic 'mode' involved in visual processing (Keitel & Gross, 2016;Barzegaran et al, 2017). It is therefore possible that our stimulation failed to influence the dominant rhythm but affected a less prominent alpha component that is more difficult to observe given the sparse spatial resolution of EEG recordings.…”

Frequency and power of human alpha oscillations drift systematically with time-on-task

“…Though it remains unclear how essential this alignment of the external and internal oscillator is (Notbohm and Herrmann, 2016;Reato et al, 2013), systematic shifts in peak frequency of the magnitude observed here may be important to consider when attempting to interact with ongoing activity. By contrast, the heterogeneous nature of the sources of alpha activity (Barzegaran et al, 2017;Bollimunta et al, 2008;Buffalo et al, 2011;Capilla et al, 2012;Haegens et al, 2015;Hughes and Crunelli, 2005;Katayama et al, 2011;Keitel and Gross, 2016b;Sadaghiani et al, 2010;Scheeringa et al, 2016) invalidates the assumption that the peak-frequency necessarily represents a single endogenous 'oscillator' and hence raises questions regarding the rationale of targeting the peak-frequency for entrainment. Therefore, precise targeting of oscillatory networks will benefit from more refined considerations of the origin and functional roles of the frequency and power content of the EEG signal.…”

“…In order to further investigate the relative independence of the power increases and frequency decreases over time, we performed an analysis using independent component analysis (ICA) and current-dipole modelling to statistically separate sources of alpha-activity contributing to the EEG signal measured on the scalp. Scalp-level EEG alpha activity reflects a superposition of multiple alpha rhythms (Barzegaran et al, 2017;Keitel and Gross, 2016a).…”

“…However, neural oscillations as recorded with EEG or MEG (M/EEG) can reflect a large mixture of generating processes in terms of both local micro-circuitry and large-scale networks. For instance, recent studies have revealed multiple distinct alpha generators with different functional and/or laminar profiles (Barzegaran et al, 2017;Bollimunta et al, 2011;Haegens et al, 2015;Hughes and Crunelli, 2005;Keitel and Gross, 2016b;Scheeringa et al, 2016). Hence, while band-limited oscillatory activity measured at the macroscale is typically interpreted as a single process within a unitary framework (e.g.…”