Broadband aperiodic components of local field potentials reflect inherent differences between cortical and subcortical activity

Abstract

Information flow in brain networks is reflected in intracerebral local field potential (LFP) measurements that have both periodic and aperiodic components. The 1/fχbroadband aperiodic component of the power spectra has been shown to track arousal level and to correlate with other physiological and pathophysiological states, with consistent patterns across cortical regions. Previous studies have focused almost exclusively on cortical neurophysiology. Here we explored the aperiodic activity of subcortical nuclei from the human thalamus and basal ganglia, in relation to simultaneously recorded cortical activity. We elaborated on the FOOOF (fitting of one over f) method by creating a new parameterization of the aperiodic component with independent and more easily interpretable parameters, which allows seamlessly fitting spectra with and without an aperiodic knee, a component of the signal that reflects the dominant timescale of aperiodic fluctuations. First, we found that the aperiodic exponent from sensorimotor cortex in Parkinson's disease (PD) patients correlated with disease severity. Second, although the aperiodic knee frequency changed across cortical regions as previously reported, no aperiodic knee was detected from subcortical regions across movement disorders patients, including the ventral thalamus (VIM), globus pallidus internus (GPi) and subthalamic nucleus (STN). All subcortical region studied exhibited a relatively low aperiodic exponent (χSTN=1.3±0.2, χVIM=1.4±0.1, χGPi=1.4±0.1) that differed markedly from cortical values (χCortex=3.2±0.4, fk,Cortex=17±5 Hz). These differences were replicated in a second dataset from epilepsy patients undergoing intracranial monitoring that included thalamic recordings. The consistently lower aperiodic exponent and lack of an aperiodic knee from all subcortical recordings may reflect cytoarchitectonic and/or functional differences between subcortical nuclei and the cortex.

Aperiodic neural activity is a biomarker for depression severity

“…Stereo-EEG electrodes placed in target regions of the affective and reward processing networks (vmPFC, dPFC, dACC, AMY, MT; see methods for regional definitions) were used to record local field potentials (LFPs) during each completion of the CAT-DI 20 across a 9-day in-patient hospital visit (Table S1) . Studies examining the aperiodic component typically constrain slope calculations to under ∼50 Hz 16,17 to avoid the impact of line noise on the slope calculation, and given previous reports of large alpha and low beta features that give rise to the ‘knee frequency’ 24,25 , we further imposed a limitation on the lower end of the range to help mitigate those known issues. Examination of the slope metrics within and across our patients confirmed greatest reliability and consistency in the 20-45 Hz sub-range.…”

Spike-phase coupling of subthalamic neurons to posterior opercular cortex predicts speech sound accuracy

“…These transient periods might represent “open windows” for effective communication between STN and cortex. The duration of these windows may be constrained by a slower subcortical neural timescale and requirements of a given motor instantiation 42,43 .…”

Empirical mode decomposition of local field potential data from optogenetic experiments

“…Most neuroscience-related studies have concentrated on narrow-band power-law scaling of power density distribution vs. frequency (Crone et al, 1998 ; Feige et al, 2005 ; Milstein et al, 2009 ; Tolkunov et al, 2010 ; Bush et al, 2023 ). Wu and Huang ( 2004 ) conducted an extensive theoretical study of the effect of noise on broadband energy vs. frequency spectrum using EMD.…”

“…Large-scale realistic models mimicking the electrophysiological LFP recordings based on particular network architectures over various model neurons show Brownean noise with an exponent of ξ = 2 (Milstein et al, 2009 ). A recent study traced the power spectral density exponents in subcortical nuclei from the human thalamus and basal ganglia while simultaneously recording cortical activity (Bush et al, 2023 ). The results show an exponent of ξ = 1.3 ± 0.2 in subcortical regions compared to ξ = 3.2 ± 0.3 in the cortex (Bush et al, 2023 ).…”

“…A recent study traced the power spectral density exponents in subcortical nuclei from the human thalamus and basal ganglia while simultaneously recording cortical activity (Bush et al, 2023 ). The results show an exponent of ξ = 1.3 ± 0.2 in subcortical regions compared to ξ = 3.2 ± 0.3 in the cortex (Bush et al, 2023 ). Although most of the EEG studies cover relatively low-frequency bands, one study, in particular, covered a wide range of frequencies 80–580 Hz and found a scaling exponent of ξ = 4.0 ± 0.1 for cortical recordings using electrocorticographic electrodes (Miller et al, 2009 ).…”