Exploring the reliability and sensitivity of the EEG power spectrum as a biomarker

Aperiodic and Hurst EEG exponents across early human brain development: A systematic review

In electroencephalographic (EEG) data, power-frequency slope exponents (1/f_β) can provide non-invasive markers of in vivo neural activity excitation-inhibition (E:I) balance. E:I balance may be altered in neurodevelopmental conditions; hence, understanding how 1/fβ evolves across infancy/childhood has implications for developing early assessments/interventions. This systematic review (PROSPERO-ID: CRD42023363294) explored the early maturation (0-26 yrs) of resting-state EEG 1/f measures (aperiodic [AE], power law [PLE] and Hurst [HE] exponents), including studies containing ≥1 1/f measures and ≥10 typically developing participants. Five databases (including Embase and Scopus) were searched during March 2023. Forty-two studies were identified (Nparticipants=3478). Risk of bias was assessed using the Quality Assessment with Diverse Studies tool. Narrative synthesis of HE data suggests non-stationary EEG activity occurs throughout development. Age-related trends were complex, with rapid decreases in AEs during infancy and heterogenous changes thereafter. Regionally, AE maxima shifted developmentally, potentially reflecting spatial trends in maturing brain connectivity. This work highlights the importance of further characterising the development of 1/f measures to better understand how E:I balance shapes brain and cognitive development.

Self-organization toward 1/f noise in deep neural networks

In biological neural networks, it has been well recognized that a healthy brain exhibits 1/f noise patterns. However, in artificial neural networks that are increasingly matching or even out-performing human cognition, this phenomenon has yet to be established. In this work, we found that similar to that of their biological counterparts, 1/f noise exists in artificial neural networks when trained on time series classification tasks. Additionally, we found that the activations of the neurons are the closest to 1/f noise when the neurons are highly utilized. Conversely, if the network is too large and many neurons are underutilized, the neuron activations deviate from 1/f noise patterns toward that of white noise.

Test-Retest Reliability of EEG Aperiodic Components in Resting and Mental Task States

Aperiodic activity is derived from the electroencephalography (EEG) power spectrum and reflects changes in the slope and shifts of the broadband spectrum. Studies have shown inconsistent test-retest reliability of the aperiodic components. This study systematically measured how the test-retest reliability of the aperiodic components was affected by data duration (1, 2, 3, 4, and 5 min), states (resting with eyes closed, resting with eyes open, performing mental arithmetic, recalling the events of the day, and mentally singing songs), and methods (the Fitting Oscillations and One-Over-F (FOOOF) and Linear Mixed-Effects Regression (LMER)) at both short (90-min) and long (one-month) intervals. The results showed that aperiodic components had fair, good, or excellent test-retest reliability (ranging from 0.53 to 0.91) at both short and long intervals. It is recommended that better reliability of the aperiodic components be obtained using data durations longer than 3 min, the resting state with eyes closed, the mental arithmetic task state, and the LMER method.

Brain health in diverse settings: How age, demographics and cognition shape brain function

Diversity in brain health is influenced by individual differences in demographics and cognition. However, most studies on brain health and diseases have typically controlled for these factors rather than explored their potential to predict brain signals. Here, we assessed the role of individual differences in demographics (age, sex, and education; n = 1298) and cognition (n = 725) as predictors of different metrics usually used in case-control studies. These included power spectrum and aperiodic (1/f slope, knee, offset) metrics, as well as complexity (fractal dimension estimation, permutation entropy, Wiener entropy, spectral structure variability) and connectivity (graph-theoretic mutual information, conditional mutual information, organizational information) from the source space resting-state EEG activity in a diverse sample from the global south and north populations. Brain-phenotype models were computed using EEG metrics reflecting local activity (power spectrum and aperiodic components) and brain dynamics and interactions (complexity and graph-theoretic measures). Electrophysiological brain dynamics were modulated by individual differences despite the varied methods of data acquisition and assessments across multiple centers, indicating that results were unlikely to be accounted for by methodological discrepancies. Variations in brain signals were mainly influenced by age and cognition, while education and sex exhibited less importance. Power spectrum activity and graph-theoretic measures were the most sensitive in capturing individual differences. Older age, poorer cognition, and being male were associated with reduced alpha power, whereas older age and less education were associated with reduced network integration and segregation. Findings suggest that basic individual differences impact core metrics of brain function that are used in standard case-control studies. Considering individual variability and diversity in global settings would contribute to a more tailored understanding of brain function.

State-dependent dynamics of cuttlefish mantle activity

Cuttlefish skin is a powerful rendering device, capable of producing extraordinary changes in visual appearance over a broad range of temporal scales. This unique ability is typically associated with camouflage; however, cuttlefish often produce skin patterns that do not appear connected with the surrounding environment, such as fast large-scale fluctuations with wave-like characteristics. Little is known about the functional significance of these dynamic patterns. In this study, we developed novel tools for analyzing pattern dynamics, and demonstrate their utility for detecting changes in feeding state that occur without concomitant changes in sensory stimulation. Under these conditions, we found that the dynamic properties of specific pattern components differ for different feeding states, despite no measurable change in the overall expression of those components. Therefore, these dynamic changes are not detectable by conventional analyses focusing on pattern expression, requiring analytical tools specifically targeted to pattern dynamics.

Aperiodic component of EEG power spectrum and cognitive performance are modulated by education in aging

Recent studies have shown a growing interest in the so-called "aperiodic" component of the EEG power spectrum, which describes the overall trend of the whole spectrum with a linear or exponential function. In the field of brain aging, this aperiodic component is associated both with age-related changes and performance on cognitive tasks. This study aims to elucidate the potential role of education in moderating the relationship between resting-state EEG features (including aperiodic component) and cognitive performance in aging. N = 179 healthy participants of the "Leipzig Study for Mind-Body-Emotion Interactions" (LEMON) dataset were divided into three groups based on age and education. Older adults exhibited lower exponent, offset (i.e. measures of aperiodic component), and Individual Alpha Peak Frequency (IAPF) as compared to younger adults. Moreover, visual attention and working memory were differently associated with the aperiodic component depending on education: in older adults with high education, higher exponent predicted slower processing speed and less working memory capacity, while an opposite trend was found in those with low education. While further investigation is needed, this study shows the potential modulatory role of education in the relationship between the aperiodic component of the EEG power spectrum and aging cognition.

Associations between the resting EEG aperiodic slope and broad domains of cognitive ability

Recent studies suggest that the EEG aperiodic exponent (often represented as a slope in log-log space) is sensitive to individual differences in momentary cognitive skills such as selective attention and information processing speed. However, findings are mixed, and most of the studies have focused on just a narrow range of cognitive domains. This study used an archival dataset to help clarify associations between resting aperiodic features and broad domains of cognitive ability, which vary in their demands on momentary processing. Undergraduates (N = 166) of age 18-52 years completed a resting EEG session as well as a standardized, individually administered assessment of cognitive ability that included measures of processing speed, working memory, and higher-order visuospatial and verbal skills. A subsample (n = 110) also completed a computerized reaction time task with three difficulty levels. Data reduction analyses revealed strong correlations between the aperiodic offset and slope across electrodes, and a single component accounted for ~60% of variance in slopes across the scalp, in both eyes-closed and eyes-open conditions. Structural equation models did not support relations between the slope and specific domains tapping momentary processes. However, secondary analyses indicated that the eyes-open slope was related to higher overall performance, as represented by a single general ability factor. A latent reaction time variable was significantly inversely related to both eyes-closed and eyes-open resting exponents, such that faster reaction times were associated with steeper slopes. These findings support and help clarify the relation of the resting EEG exponent to individual differences in cognitive skills.

Test-retest reliability of spectral parameterization by 1/f characterization using SpecParam

SpecParam (formally known as FOOOF) allows for the refined measurements of electroencephalography periodic and aperiodic activity, and potentially provides a non-invasive measurement of excitation: inhibition balance. However, little is known about the psychometric properties of this technique. This is integral for understanding the usefulness of SpecParam as a tool to determine differences in measurements of cognitive function, and electroencephalography activity. We used intraclass correlation coefficients to examine the test-retest reliability of parameterized activity across three sessions (90 minutes apart and 30 days later) in 49 healthy young adults at rest with eyes open, eyes closed, and during three eyes closed cognitive tasks including subtraction (Math), music recall (Music), and episodic memory (Memory). Intraclass correlation coefficients were good for the aperiodic exponent and offset (intraclass correlation coefficients > 0.70) and parameterized periodic activity (intraclass correlation coefficients > 0.66 for alpha and beta power, central frequency, and bandwidth) across conditions. Across all three sessions, SpecParam performed poorly in eyes open (40% of participants had poor fits over non-central sites) and had poor test-retest reliability for parameterized periodic activity. SpecParam mostly provides reliable metrics of individual differences in parameterized neural activity. More work is needed to understand the suitability of eyes open resting data for parameterization using SpecParam.

Associations between age-related differences in occipital alpha power and the broadband parameters of the EEG power spectrum: A cross-sectional cohort study

In adulthood, neurological structure and function are often affected by aging, with negative implications for daily life as well as laboratory-based tasks. Some of these changes include decreased efficiency modulating cortical activity and lower signal-to-noise ratios in neural processing (as inferred from surface electroencephalography). To better understand mechanisms influencing age-related changes in cortical activity, we explored the effects of aging on narrow-band alpha power (7.5-12.5 Hz) and broadband/aperiodic components that span a wider range (1.5-30.5 Hz) over the occipital region during eyes-open and eyes-closed wakeful rest in 19 healthy young adults (18-35 years) and 21 community-dwelling older adults (59+ years). Older adults exhibited a smaller change in alpha power across conditions compared to younger adults. Older adults also showed flatter aperiodic slopes in both conditions. These changes in narrow-band alpha are consistent with previous work and suggest that older adults may have a reduced ability to modulate state-specific activity. Differences in the aperiodic slope suggest age-related changes in the signal-noise-ratio in cortical oscillations. However, the relationship between narrow-band alpha modulation and the aperiodic slope was unclear, warranting further investigation into how these variables relate to each other in the aging process. In summary, aging is associated with a broadband flattening of the EEG power spectrum and reduced state-specific modulation of narrow-band alpha power, but these changes appear to be (at least partially) independent of each other. The present findings suggest that separate mechanisms may underlie age-related differences in aperiodic power and narrow-band oscillations.

Using oscillatory and aperiodic neural activity features for identifying idle state in SSVEP-based BCIs reduces false triggers

Objective.In existing studies, rhythmic (oscillatory) components were used as main features to identify brain states, such as control and idle states, while non-rhythmic (aperiodic) components were ignored. Recent studies have shown that aperiodic (1/f) activity is functionally related to cognitive processes. It is not clear if aperiodic activity can distinguish brain states in asynchronous brain-computer interfaces (BCIs) to reduce false triggers. In this paper, we propose an asynchronous method based on the fusion of oscillatory and aperiodic features for steady-state visual evoked potential-based BCIs.Approach.The proposed method first evaluates the oscillatory and aperiodic components of control and idle states using irregular-resampling auto-spectral analysis. Oscillatory features are then extracted using the spectral power of fundamental, second-harmonic, and third-harmonic frequencies of the oscillatory component, and aperiodic features are extracted using the slope and intercept of the first-order polynomial of the spectral fit of the aperiodic component under a log-logarithmic axis. The process produces two types of feature pools (oscillatory, aperiodic features). Next, feature selection (dimensionality reduction) is applied to the feature pools by Bonferroni correctedp-values from two-way analysis of variance. Last, these spatial-specific statistically significant features are used as input for classification to identify the idle state.Mainresults.On a 7-target dataset from 15 subjects, the mix of oscillatory and aperiodic features achieved an average accuracy of 88.39% compared to 83.53% when using oscillatory features alone (4.86% improvement). The results demonstrated that the proposed idle state recognition method achieved enhanced performance by incorporating aperiodic features.Significance.Our results demonstrated that (1) aperiodic features were effective in recognizing idle states and (2) fusing features of oscillatory and aperiodic components enhanced classification performance by 4.86% compared to oscillatory features alone.

EEG Phase Can Be Predicted with Similar Accuracy across Cognitive States after Accounting for Power and Signal-to-Noise Ratio

EEG phase is increasingly used in cognitive neuroscience, brain-computer interfaces, and closed-loop stimulation devices. However, it is unknown how accurate EEG phase prediction is across cognitive states. We determined the EEG phase prediction accuracy of parieto-occipital alpha waves across rest and task states in 484 participants over 11 public datasets. We were able to track EEG phase accurately across various cognitive conditions and datasets, especially during periods of high instantaneous alpha power and signal-to-noise ratio (SNR). Although resting states generally have higher accuracies than task states, absolute accuracy differences were small, with most of these differences attributable to EEG power and SNR. These results suggest that experiments and technologies using EEG phase should focus more on minimizing external noise and waiting for periods of high power rather than inducing a particular cognitive state.

Preliminary Observations of Personalized Repetitive Magnetic Stimulation (PrTMS) Guided by EEG Spectra for Concussion

There are no FDA-approved treatments for the chronic sequelae of concussion. Repetitive magnetic transcranial stimulation (rTMS) has been explored as a therapy but outcomes have been inconsistent. To address this we developed a personalized rTMS (PrTMS) protocol involving continual rTMS stimulus frequency adjustment and progressive activation of multiple cortical sites, guided by spectral electroencephalogram (EEG)-based analyses and psychological questionnaires. We acquired pilot clinical data for 185 symptomatic brain concussion patients who underwent the PrTMS protocol over an approximate 6 week period. The PrTMS protocol used a proprietary EEG spectral frequency algorithm to define an initial stimulation frequency based on an anteriorly graded projection of the measured occipital alpha center peak, which was then used to interpolate and adjust regional stimulation frequency according to weekly EEG spectral acquisitions. PrTMS improved concussion indices and normalized the cortical alpha band center frequency and peak EEG amplitude. This potentially reflected changed neurotransmitter, cognitive, and perceptual status. PrTMS may be a promising treatment choice for patients with persistent concussion symptoms. This clinical observational study was limited in that there was no control group and a number of variables were not recorded, such as time since injury and levels of depression. While the present observations are indeed preliminary and cursory, they may suggest further prospective research on PrTMS in concussion, and exploration of the spectral EEG as a concussion biomarker, with the ultimate goals of confirmation and determining optimal PrTMS treatment parameters.

Comparative analysis of multifaceted neural effects associated with varying endogenous cognitive load

Contemporary neuroscience has firmly established that mental state variation concurs with changes in neural dynamic activity in a complex way that a one-to-one mapping cannot describe. To explore the scenario of the multifaceted changes in neural dynamics associated with simple mental state variation, we took cognitive load - a common cognitive manipulation in psychology - as a venue to characterize how multiple neural dynamic features are simultaneously altered by the manipulation and how their sensitivity differs. Electroencephalogram was collected from 152 participants performing stimulus-free tasks with different demands. The results show that task demand alters wide-ranging neural dynamic features, including band-specific oscillations across broad frequency bands, scale-free dynamics, and cross-frequency phase-amplitude coupling. The scale-free dynamics outperformed others in indexing cognitive load variation. This study demonstrates a complex relationship between cognitive dynamics and neural dynamics, which points to a necessity to integrate multifaceted neural dynamic features when studying mind-brain relationship in the future.

Characteristics of changes in the functional status of the brain before and after 1,000 m all-out paddling for different levels of dragon boat athletes

Purposes

Dragon boat is a traditional sport in China, but the brain function characteristics of dragon boat athletes are still unclear. Our purpose is to explore the changing characteristics of brain function of dragon boat athletes at different levels before and after exercise by monitoring the changes of EEG power spectrum and microstate of athletes before and after rowing.

Methods

Twenty-four expert dragon boat athletes and 25 novice dragon boat athletes were selected as test subjects to perform the 1,000 m all-out paddling exercise on a dragon boat dynamometer. Their resting EEG data was collected pre- and post-exercise, and the EEG data was pre-processed and then analyzed using power spectrum and microstate based on Matlab software.

Results

Post-Exercise, the Heart Rate peak (HR peak), Percentage of Heart Rate max (HR max), Rating of Perceived Exertion (RPE), and Exercise duration of the novice group were significantly higher than expert group (p < 0.01). Pre-exercise, the power spectral density values in the δ, α1, α2, and β1 bands were significantly higher in the expert group compared to the novice group (p < 0.05). Post-exercise, the power spectral density values in the δ, θ, and α1 bands were significantly lower in the expert group compared to the novice group (p < 0.05), the power spectral density values of α2, β1, and β2 bands were significantly higher (p < 0.05). The results of microstate analysis showed that the duration and contribution of microstate class D were significantly higher in the pre-exercise expert group compared to the novice group (p < 0.05), the transition probabilities of A → D, C → D, and D → A were significantly higher (p < 0.05). Post-exercise, the duration, and contribution of microstate class C in the expert group decreased significantly compared to the novice group (p < 0.05), the occurrence of microstate classes A and D were significantly higher (p < 0.05), the transition probability of A → B was significantly higher (p < 0.05), and the transition probabilities of C → D and D → C were significantly lower (p < 0.05).

Conclusion

The functional brain state of dragon boat athletes was characterized by expert athletes with closer synaptic connections of brain neurons and higher activation of the dorsal attention network in the resting state pre-exercise. There still had higher activation of cortical neurons after paddling exercise. Expert athletes can better adapt to acute full-speed oar training.

Stability, change, and reliable individual differences in electroencephalography measures: a lifespan perspective on progress and opportunities

Electroencephalographic (EEG) methods have great potential to serve both basic and clinical science approaches to understand individual differences in human neural function. Importantly, the psychometric properties of EEG data, such as internal consistency and test-retest reliability, constrain their ability to differentiate individuals successfully. Rapid and recent technological and computational advancements in EEG research make it timely to revisit the topic of psychometric reliability in the context of individual difference analyses. Moreover, pediatric and clinical samples provide some of the most salient and urgent opportunities to apply individual difference approaches, but the changes these populations experience over time also provide unique challenges from a psychometric perspective. Here we take a developmental neuroscience perspective to consider progress and new opportunities for parsing the reliability and stability of individual differences in EEG measurements across the lifespan. We first conceptually map the different profiles of measurement reliability expected for different types of individual difference analyses over the lifespan. Next, we summarize and evaluate the state of the field's empirical knowledge and need for testing measurement reliability, both internal consistency and test-retest reliability, across EEG measures of power, event-related potentials, nonlinearity, and functional connectivity across ages. Finally, we highlight how standardized pre-processing software for EEG denoising and empirical metrics of individual data quality may be used to further improve EEG-based individual differences research moving forward. We also include recommendations and resources throughout that individual researchers can implement to improve the utility and reproducibility of individual differences analyses with EEG across the lifespan.

Decomposing age effects in EEG alpha power

Increasing life expectancy is prompting the need to understand how the brain changes during healthy aging. Research utilizing electroencephalography (EEG) has found that the power of alpha oscillations decrease from adulthood on. However, non-oscillatory (aperiodic) components in the data may confound results and thus require re-investigation of these findings. Thus, the present report analyzed a pilot and two additional independent samples (total N = 533) of resting-state EEG from healthy young and elderly individuals. A newly developed algorithm was utilized that allows the decomposition of the measured signal into periodic and aperiodic signal components. By using multivariate sequential Bayesian updating of the age effect in each signal component, evidence across the datasets was accumulated. It was hypothesized that previously reported age-related alpha power differences will largely diminish when total power is adjusted for the aperiodic signal component. First, the age-related decrease in total alpha power was replicated. Concurrently, decreases of the intercept and slope (i.e. exponent) of the aperiodic signal component were observed. Findings on aperiodic-adjusted alpha power indicated that this general shift of the power spectrum leads to an overestimation of the true age effects in conventional analyses of total alpha power. Thus, the importance of separating neural power spectra into periodic and aperiodic signal components is highlighted. However, also after accounting for these confounding factors, the sequential Bayesian updating analysis provided robust evidence that aging is associated with decreased aperiodic-adjusted alpha power. While the relation of the aperiodic component and aperiodic-adjusted alpha power to cognitive decline demands further investigation, the consistent findings on age effects across independent datasets and high test-retest reliabilities support that these newly emerging measures are reliable markers of the aging brain. Hence, previous interpretations of age-related decreases in alpha power are reevaluated, incorporating changes in the aperiodic signal.

Sources of Variation in the Spectral Slope of the Sleep EEG

The 1/f spectral slope of the electroencephalogram (EEG) estimated in the γ frequency range has been proposed as an arousal marker that differentiates wake, nonrapid eye movement (NREM) sleep, and rapid eye movement (REM) sleep. Here, we sought to replicate and extend these findings in a large sample, providing a comprehensive characterization of how slope changes with age, sex, and its test-retest reliability as well as potential confounds that could affect the slope estimation. We used 10,255 whole-night polysomnograms (PSGs) from the National Sleep Research Resource (NSRR). All preprocessing steps were performed using an open-source Luna package and the spectral slope was estimated by fitting log-log linear regression models on the absolute power from 30 to 45 Hz separately for wake, NREM, and REM stages. We confirmed that the mean spectral slope grows steeper going from wake to NREM to REM sleep. We found that the choice of mastoid referencing scheme modulated the extent to which electromyogenic, or electrocardiographic artifacts were likely to bias 30- to 45-Hz slope estimates, as well as other sources of technical, device-specific bias. Nonetheless, within individuals, slope estimates were relatively stable over time. Both cross-sectionally and longitudinal, slopes tended to become shallower with increasing age, particularly for REM sleep; males tended to show flatter slopes than females across all states. Our findings support that spectral slope can be a valuable arousal marker for both clinical and research endeavors but also underscore the importance of considering interindividual variation and multiple methodological aspects related to its estimation.

Globally elevated excitation-inhibition ratio in children with autism spectrum disorder and below-average intelligence

Background

Altered neuronal excitation–inhibition (E–I) balance is strongly implicated in ASD. However, it is not known whether the direction and degree of changes in the E–I ratio in individuals with ASD correlates with intellectual disability often associated with this developmental disorder. The spectral slope of the aperiodic 1/f activity reflects the E–I balance at the scale of large neuronal populations and may uncover its putative alternations in individuals with ASD with and without intellectual disability.

Methods

Herein, we used magnetoencephalography (MEG) to test whether the 1/f slope would differentiate ASD children with average and below–average (< 85) IQ. MEG was recorded at rest with eyes open/closed in 49 boys with ASD aged 6–15 years with IQ ranging from 54 to 128, and in 49 age-matched typically developing (TD) boys. The cortical source activity was estimated using the beamformer approach and individual brain models. We then extracted the 1/f slope by fitting a linear function to the log–log-scale power spectra in the high-frequency range.

Results

The global 1/f slope averaged over all cortical sources demonstrated high rank-order stability between the two conditions. Consistent with previous research, it was steeper in the eyes-closed than in the eyes-open condition and flattened with age. Regardless of condition, children with ASD and below-average IQ had flatter slopes than either TD or ASD children with average or above-average IQ. These group differences could not be explained by differences in signal-to-noise ratio or periodic (alpha and beta) activity.

Limitations

Further research is needed to find out whether the observed changes in E–I ratios are characteristic of children with below-average IQ of other diagnostic groups.

Conclusions

The atypically flattened spectral slope of aperiodic activity in children with ASD and below-average IQ suggests a shift of the global E–I balance toward hyper-excitation. The spectral slope can provide an accessible noninvasive biomarker of the E–I ratio for making objective judgments about treatment effectiveness in people with ASD and comorbid intellectual disability.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13229-022-00498-2.

Reliability and Validity of Power Spectrum Slope (PSS): A Metric for Measuring Resting-State Functional Magnetic Resonance Imaging Activity of Single Voxels

Methods that capture the features of single voxels of resting-state fMRI (RS-fMRI) could precisely localize the abnormal spontaneous activity and hence guide precise brain stimulation. As one of these metrics, the amplitude of low-frequency fluctuation (ALFF) has been used in numerous studies, however, it is frequency-dependent and the division of frequency bands is still controversial. Based on the well-accepted power law of time series, this study proposed an approach, namely, power spectrum slope (PSS), to characterize the RS-fMRI time series of single voxels. Two metrics, i.e., linear coefficient b and power-law slope b’ were used and compared with ALFF. The reliability and validity of the PSS approach were evaluated on public RS-fMRI datasets (n = 145 in total) of eyes closed (EC) and eyes open (EO) conditions after image preprocessing, with 21 subjects scanned two times for test–retest reliability analyses. Specifically, we used the paired t-test between EC and EO conditions to assess the validity and intra-class correlation (ICC) to assess the reliability. The results included the following: (1) PSS detected similar spatial patterns of validity (i.e., EC–EO differences) and less test–retest reliability with those of ALFF; (2) PSS linear coefficient b showed better validity and reliability than power-law slope b’; (3) While the PPS showed less validity in most regions, PSS linear coefficient b showed exclusive EC–EO difference in the medial temporal lobe which did not show in ALFF. The power spectrum plot in the parahippocampus showed a “cross-over” of power magnitudes between EC and EO conditions in the higher frequency bands (>0.1 Hz). These results demonstrated that PSS (linear coefficient b) is complementary to ALFF for detecting the local spontaneous activity.

Electroencephalogram aperiodic power spectral slope can be reliably measured and predicts ADHD risk in early development

The aperiodic exponent of the electroencephalogram (EEG) power spectrum has received growing attention as a physiological marker of neurodevelopmental psychopathology, including attention-deficit/hyperactivity disorder (ADHD). However, its use as a marker of ADHD risk across development, and particularly in very young children, is limited by unknown reliability, difficulty in aligning canonical band-based measures across development periods, and unclear effects of treatment in later development. Here, we investigate the internal consistency of the aperiodic EEG power spectrum slope and its association with ADHD risk in both infants (n = 69, 1-month-old) and adolescents (n = 262, ages 11-17 years). Results confirm good to excellent internal consistency in infancy and adolescence. In infancy, a larger aperiodic exponent was associated with greater family history of ADHD. In contrast, in adolescence, ADHD diagnosis was associated with a smaller aperiodic exponent, but only in children with ADHD who had not received stimulant medication treatment. Results suggest that disruptions in cortical development associated with ADHD risk may be detectable shortly after birth via this approach. Together, findings imply a dynamic developmental shift in which the developmentally normative flattening of the EEG power spectrum is exaggerated in ADHD, potentially reflecting imbalances in cortical excitation and inhibition that could contribute to long-lasting differences in brain connectivity.

Spectral parameterization for studying neurodevelopment: How and why

A growing body of literature suggests that the explicit parameterization of neural power spectra is important for the appropriate physiological interpretation of periodic and aperiodic electroencephalogram (EEG) activity. In this paper, we discuss why parameterization is an imperative step for developmental cognitive neuroscientists interested in cognition and behavior across the lifespan, as well as how parameterization can be readily accomplished with an automated spectral parameterization ("specparam") algorithm (Donoghue et al., 2020a). We provide annotated code for power spectral parameterization, via specparam, in Jupyter Notebook and R Studio. We then apply this algorithm to EEG data in childhood (N = 60; Mage = 9.97, SD = 0.95) to illustrate its utility for developmental cognitive neuroscientists. Ultimately, the explicit parameterization of EEG power spectra may help us refine our understanding of how dynamic neural communication contributes to normative and aberrant cognition across the lifespan. Data and annotated analysis code for this manuscript are available on GitHub as a supplement to the open-access specparam toolbox.

SCC-MPGCN: Self-Attention Coherence Clustering Based on Multi-Pooling Graph Convolutional Network for EEG Emotion Recognition

The emotion recognition with electroencephalography (EEG) has been widely studied using the deep learning methods, but the topology of EEG channels is rarely exploited completely. In this paper, we propose a self-attention coherence clustering based on multi-pooling graph convolutional network (SCC-MPGCN) model for EEG emotion recognition. locking value (PLV)The adjacency matrix is constructed based on phase- to describe the intrinsic relationship between different EEG electrodes as graph signals. The Laplacian matrix of a graph is obtained from the adjacency matrix and then is fed into the graph convolutional layers to learn the generalized features. Moreover, we propose a novel graph coarsening method called self-attention coherence clustering (SCC), using the coherence to cluster the nodes. The benefits are that the global information can be achieved from the raw data and the dimensionality of input can be reduced. Meanwhile, a multi-pooling graph convolutional network (MPGCN) block is introduced to learn the generalized emotional states features and tackle the problem of imbalanced dimensionality. The fully-connected layer and a softmax layer are adopted to drive the final prediction. We carry out the extensive experiments on DEAP dataset and the experimental results show that the proposed method has better classification results than the state-of-the-art methods with the 10-fold cross-validation. And the model achieves the emotion recognition performance with a mean accuracy of 96.37%, 97.02%, 96.72% on valence, arousal, and dominance dimension, respectively.

Resting EEG spectral slopes are associated with age-related differences in information processing speed

BACKGROUND

Previous research has shown the slope of the EEG power spectrum differentiates between older and younger adults in various experimental cognitive tasks. We extend that work, assessing the relation between the EEG power spectrum and performance on the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS).

METHODS

Twenty-one younger and twenty-three older adults completed the RBANS with EEG data collected at rest. Using spectral parameterization, we tested the mediating effect of the spectral slope on differences in subsequent cognitive task performance.

RESULTS

Older adults performed reliably worse on the RBANS overall, and on the Attention and Delayed Memory domains specifically. However, evidence of mediation was only found for the Coding subtest.

CONCLUSIONS

The slope of the resting EEG power spectrum mediated age-related differences in cognition, but only in a task requiring speeded processing. Mediation was not statistically significant for delayed memory, even though age-related differences were present.

Time-resolved parameterization of aperiodic and periodic brain activity

Macroscopic neural dynamics comprise both aperiodic and periodic signal components. Recent advances in parameterizing neural power spectra offer practical tools for evaluating these features separately. Although neural signals vary dynamically and express non-stationarity in relation to ongoing behaviour and perception, current methods yield static spectral decompositions. Here, we introduce Spectral Parameterization Resolved in Time (SPRiNT) as a novel method for decomposing complex neural dynamics into periodic and aperiodic spectral elements in a time-resolved manner. First, we demonstrate, with naturalistic synthetic data, SPRiNT's capacity to reliably recover time-varying spectral features. We emphasize SPRiNT's specific strengths compared to other time-frequency parameterization approaches based on wavelets. Second, we use SPRiNT to illustrate how aperiodic spectral features fluctuate across time in empirical resting-state EEG data (n=178) and relate the observed changes in aperiodic parameters over time to participants' demographics and behaviour. Lastly, we use SPRiNT to demonstrate how aperiodic dynamics relate to movement behaviour in intracranial recordings in rodents. We foresee SPRiNT responding to growing neuroscientific interests in the parameterization of time-varying neural power spectra and advancing the quantitation of complex neural dynamics at the natural time scales of behaviour.

Stability of spectral estimates in resting-state magnetoencephalography: Recommendations for minimal data duration with neuroanatomical specificity

The principle of resting-state paradigms is appealing and practical for collecting data from impaired patients and special populations, especially if data collection times can be minimized. To achieve this goal, researchers need to ensure estimated signal features of interest are robust. In electro- and magnetoencephalography (EEG, MEG) we are not aware of any studies of the minimal length of data required to yield a robust one-session snapshot of the frequency-spectrum derivatives that are typically used to characterize the complex dynamics of the brain’s resting-state. We aimed to fill this knowledge gap by studying the stability of common spectra measures of resting-state MEG source time series obtained from large samples of single-session recordings from shared data repositories featuring different recording conditions and instrument technologies (OMEGA: N = 107; Cam-CAN: N = 50). We discovered that the rhythmic and arrhythmic spectral properties of intrinsic brain activity can be robustly estimated in most cortical regions when derived from relatively short segments of 30-s to 120-s of resting-state data, regardless of instrument technology and resting-state paradigm. Using an adapted leave-one-out approach and Bayesian analysis, we also provide evidence that the stability of spectral features over time is unaffected by age, sex, handedness, and general cognitive function. In summary, short MEG sessions are sufficient to yield robust estimates of frequency-defined brain activity during resting-state. This study may help guide future empirical designs in the field, particularly when recording times need to be minimized, such as with patient or special populations.