Distinguishing low frequency oscillations within the 1/f spectral behaviour of electromagnetic brain signals

Abstract

Background: It has been acknowledged that the frequency spectrum of measured electromagnetic (EM) brain signals shows a decrease in power with increasing frequency. This spectral behaviour may lead to difficulty in distinguishing event-related peaks from ongoing brain activity in the electro-and magnetoencephalographic (EEG and MEG) signal spectra. This can become an issue especially in the analysis of low frequency oscillations (LFOs) -below 0.5 Hzwhich are currently being observed in signal recordings linked with specific pathologies such as epileptic seizures or attention deficit hyperactivity disorder (ADHD), in sleep studies, etc.

Complexity Analysis of Spontaneous Brain Activity in Attention-Deficit/Hyperactivity Disorder: Diagnostic Implications

“…The clinical group comprised 14 male (mean age, 9.64 ± 1.04 years; range [8][9][10][11][12] children with ADHD recruited from the community. Inclusion criteria included a full DSM-IV diagnosis of ADHD combined type with associated impairment in at least 2 settings and a Conners' Parents Hyperactivity rating greater than 2 SD above age-and sex-specific means.…”

“…Moreover, MEG is a complementary signal to EEG and represents an entirely non-invasive procedure for brain analysis in children. MEG has been scarcely utilized in ADHD investigation (9)(10)(11)(12) and, as far as we know, the diagnostic utility of this technique has never been tested in ADHD. It is important to notice that there are some major differences between EEG and MEG.…”

Advancing the detection of steady-state visual evoked potentials in brain–computer interfaces

“…These plots show that the difference between the distributions of the target and nontarget canonical correlations ρ i is most pronounced for ρ 1 . Most importantly, we could see that the typical means (for all canonical correlations) per stimulation frequency differ significantly, in a way that reflects the general power density of EEG data which, similar to pink noise, exhibits a characteristic 1/f profile [32,33]. The mean and standard deviation of all target and nontarget canonical correlations are shown in figure 11 as a function of stimulation frequency f .…”

Neuronal oscillations in the EEG under varying cognitive load: A comparative study between slow waves and faster oscillations

“…Hence, it is important to outline the main reasons why we opted for this particular experimental set-up and data analysis methods, and the potential benefits of such an approach. First, VLF brain activity is notoriously difficult to study; one must ensure that the observations made arise from neuronal activity and are not due to other physiological (respiratory, cardiac and vasomotion) processes Voipio et al, 2003, Auer, 2008, and/or arise from inherent properties in the data (such as the 1/f spectral trend of the EEG; Demanuele et al, 2007), or from the data analysis procedure itself. Hence both the analysis and interpretation of slow signals related to cognitive tasks needs to be done with caution.…”

“…It is worth noting here that direct comparison of the slow waves with higher frequency bands is not very straightforward; for example, the 1/f intrinsic nature of the EEG power spectrum implies that the slow wave band will always contain higher power in comparison to other frequency bands (see Demanuele et al 2007 for a method that tackles this). Furthermore, since evidence in the literature suggests that this spontaneous VLF activity might act as a 'baseline' that underlies brain activity elicited by the task Balduzzi et al, 2008;Monto et al, 2008;Demanuele et al, 2010), it may not be revealed as 'oscillations' in the classical sense of having peaks in the power spectra of the EEG channels per se, or as separate independent components following the demixing procedure.…”