Experimental observation of a theoretically predicted nonlinear sleep spindle harmonic in human EEG

Physiologically based arousal state estimation and dynamics

“…Because sleep spindles are a transient phenomenon typically lasting around 1 s, calculating the power spectrum from a 30-s period of S2 sleep results in significant mixing of the sleep spindle with background sleep, because the sleep spindles are present for only a short fraction of the time. Therefore, we measure the sleep spindle power spectrum by isolating the sleep spindles to minimize the amount of background sleep activity included in the power spectrum, as implemented and validated in previous work (Abeysuriya et al, 2014b). Sleep spindles consistent with the definitions in the R&K and AASM schemes were detected in 24 of the subjects.…”

“…(b) Experimental spectra for the same sleep stages, drawn from the EEG data set used in this study. A 5 point moving average was used to smooth the spectrum at f > 20 Hz for visual comparison to the model for all experimental spectra except sleep spindles, which are already averaged over spindle events (Abeysuriya et al, 2014b). The model captures key features of the experimental power spectrum, including the alpha harmonic (the beta peak) in EC.…”

NFTsim: Theory and Simulation of Multiscale Neural Field Dynamics

“…This provides 96 a range of common applicability on scales of around 1 mm, or slightly less, where 97 complementary predictions can be made and tested -an overlap that will increase as 98 microscopic simulations increase in scale. Equally significantly, quantitative neural field 99 predictions can readily be made of quantities observable by EEG, MEG, fMRI, 100 electrocorticography (ECoG), and other imaging technologies, by adding the biophysics 101 of these signals, measurement procedures, and postprocessing [27][28][29][30]. This enables 102 predictions of a single brain model to be tested against multiple phenomena in order to 103 better determine the relevant physiological parameters.…”

“…As an illustration of the versatility of NFT approaches, we note that the particular 105 NFT on which the present NFTsim software is based has been extensively applied and 106 quantitatively tested against experiments, including EEG, evoked response potentials 107 (ERPs), ECoG, age-related changes to the physiology of the brain, sleep and arousal 108 PLOS 3/44 dynamics, seizures, Parkinson's disease, and other disorders, transcranial magnetic 109 stimulation (TMS), synaptic plasticity phenomena [1,6,[27][28][29][30][31][32][33][34][35][36][37][38][39]. Indeed, one of the major 110 strengths of this NFT is its versatility: within the same framework we can express 111 different models to study purely cortical phenomena, the corticothalamic system, basal 112 ganglia, sleep dynamics, or the visual cortex, among an essentially unlimited number of 113 other applications [1, 27-29, 31, 33, 35-38, 40-43].…”

“…detecting the zero crossings. This is the method that has been used extensively to find 532 the roots of the corticothalamic model for different parameter ranges [6,29,30,39,79]. A 533 standard root finding algorithm (e.g., Newton-Raphson) can then be used to refine the 534 roots.…”