Assessment of the excitation-inhibition ratio in the Fmr1 KO2 mouse using neuronal oscillation dynamics.

In vitro and ex vivo studies have shown consistent indications of hyperexcitability in the Fragile X Messenger Ribonucleoprotein 1 (Fmr1) knockout mouse model of autism spectrum disorder. We recently introduced a method to quantify network-level functional excitation-inhibition ratio from the neuronal oscillations. Here, we used this measure to study whether the implicated synaptic excitation-inhibition disturbances translate to disturbances in network physiology in the Fragile X Messenger Ribonucleoprotein 1 (Fmr1) gene knockout model. Vigilance-state scoring was used to extract segments of inactive wakefulness as an equivalent behavioral condition to the human resting-state and, subsequently, we performed high-frequency resolution analysis of the functional excitation-inhibition biomarker, long-range temporal correlations, and spectral power. We corroborated earlier studies showing increased high-frequency power in Fragile X Messenger Ribonucleoprotein 1 (Fmr1) knockout mice. Long-range temporal correlations were higher in the gamma frequency ranges. Contrary to expectations, functional excitation-inhibition was lower in the knockout mice in high frequency ranges, suggesting more inhibition-dominated networks. Exposure to the Gamma-aminobutyric acid (GABA)-agonist clonazepam decreased the functional excitation-inhibition in both genotypes, confirming that increasing inhibitory tone results in a reduction of functional excitation-inhibition. In addition, clonazepam decreased electroencephalogram power and increased long-range temporal correlations in both genotypes. These findings show applicability of these new resting-state electroencephalogram biomarkers to animal for translational studies and allow investigation of the effects of lower-level disturbances in excitation-inhibition balance.

Long-Range Amplitude Coupling Is Optimized for Brain Networks That Function at Criticality.

Brain function depends on segregation and integration of information processing in brain networks often separated by long-range anatomic connections. Neuronal oscillations orchestrate such distributed processing through transient amplitude and phase coupling, yet surprisingly, little is known about local network properties facilitating these functional connections. Here, we test whether criticality, a dynamical state characterized by scale-free oscillations, optimizes the capacity of neuronal networks to couple through amplitude or phase, and transfer information. We coupled in silico networks which exhibit oscillations in the α band (8-16 Hz), and varied excitatory and inhibitory connectivity. We found that phase coupling of oscillations emerges at criticality, and that amplitude coupling, as well as information transfer, are maximal when networks are critical. Importantly, regulating criticality through modulation of synaptic gain showed that critical dynamics, as opposed to a static ratio of excitatory and inhibitory connections, optimize network coupling and information transfer. Our data support the idea that criticality is important for local and global information processing and may help explain why brain disorders characterized by local alterations in criticality also exhibit impaired long-range synchrony, even before degeneration of axonal connections.SIGNIFICANCE STATEMENT To perform adaptively in a changing environment, our brains dynamically coordinate activity across distant areas. Empirical evidence suggests that long-range amplitude and phase coupling of oscillations are systems-level mechanisms enabling transient formation of spatially distributed functional networks on the backbone of a relatively static structural connectome. However, surprisingly little is known about the local network properties that optimize coupling and information transfer. Here, we show that criticality, a dynamical state characterized by scale-free oscillations and a hallmark of neuronal network activity, optimizes the capacity of neuronal networks to couple through amplitude or phase and transfer information.

Catecholamines alter the intrinsic variability of cortical population activity and perception.

The ascending modulatory systems of the brain stem are powerful regulators of global brain state. Disturbances of these systems are implicated in several major neuropsychiatric disorders. Yet, how these systems interact with specific neural computations in the cerebral cortex to shape perception, cognition, and behavior remains poorly understood. Here, we probed into the effect of two such systems, the catecholaminergic (dopaminergic and noradrenergic) and cholinergic systems, on an important aspect of cortical computation: its intrinsic variability. To this end, we combined placebo-controlled pharmacological intervention in humans, recordings of cortical population activity using magnetoencephalography (MEG), and psychophysical measurements of the perception of ambiguous visual input. A low-dose catecholaminergic, but not cholinergic, manipulation altered the rate of spontaneous perceptual fluctuations as well as the temporal structure of "scale-free" population activity of large swaths of the visual and parietal cortices. Computational analyses indicate that both effects were consistent with an increase in excitatory relative to inhibitory activity in the cortical areas underlying visual perceptual inference. We propose that catecholamines regulate the variability of perception and cognition through dynamically changing the cortical excitation-inhibition ratio. The combined readout of fluctuations in perception and cortical activity we established here may prove useful as an efficient and easily accessible marker of altered cortical computation in neuropsychiatric disorders.

Scale-Free Amplitude Modulation of Neuronal Oscillations Tracks Comprehension of Accelerated Speech.

Speech comprehension is preserved up to a threefold acceleration, but deteriorates rapidly at higher speeds. Current models posit that perceptual resilience to accelerated speech is limited by the brain's ability to parse speech into syllabic units using δ/θ oscillations. Here, we investigated whether the involvement of neuronal oscillations in processing accelerated speech also relates to their scale-free amplitude modulation as indexed by the strength of long-range temporal correlations (LRTC). We recorded MEG while 24 human subjects (12 females) listened to radio news uttered at different comprehensible rates, at a mostly unintelligible rate and at this same speed interleaved with silence gaps. δ, θ, and low-γ oscillations followed the nonlinear variation of comprehension, with LRTC rising only at the highest speed. In contrast, increasing the rate was associated with a monotonic increase in LRTC in high-γ activity. When intelligibility was restored with the insertion of silence gaps, LRTC in the δ, θ, and low-γ oscillations resumed the low levels observed for intelligible speech. Remarkably, the lower the individual subject scaling exponents of δ/θ oscillations, the greater the comprehension of the fastest speech rate. Moreover, the strength of LRTC of the speech envelope decreased at the maximal rate, suggesting an inverse relationship with the LRTC of brain dynamics when comprehension halts. Our findings show that scale-free amplitude modulation of cortical oscillations and speech signals are tightly coupled to speech uptake capacity.SIGNIFICANCE STATEMENT One may read this statement in 20-30 s, but reading it in less than five leaves us clueless. Our minds limit how much information we grasp in an instant. Understanding the neural constraints on our capacity for sensory uptake is a fundamental question in neuroscience. Here, MEG was used to investigate neuronal activity while subjects listened to radio news played faster and faster until becoming unintelligible. We found that speech comprehension is related to the scale-free dynamics of δ and θ bands, whereas this property in high-γ fluctuations mirrors speech rate. We propose that successful speech processing imposes constraints on the self-organization of synchronous cell assemblies and their scale-free dynamics adjusts to the temporal properties of spoken language.

Strong long-range temporal correlations of beta/gamma oscillations are associated with poor sustained visual attention performance.

Neuronal oscillations exhibit complex amplitude fluctuations with autocorrelations that persist over thousands of oscillatory cycles. Such long-range temporal correlations (LRTC) are thought to reflect neuronal systems poised near a critical state, which would render them capable of quick reorganization and responsive to changing processing demands. When we concentrate, however, the influence of internal and external sources of distraction is better reduced, suggesting that neuronal systems involved with sustained attention could benefit from a shift toward the less volatile sub-critical state. To test these ideas, we recorded electroencephalography (EEG) from healthy volunteers during eyes-closed rest and during a sustained attention task requiring a speeded response to images deviating in their presentation duration. We show that for oscillations recorded during rest, high levels of alpha-band LRTC in the sensorimotor region predicted good reaction-time performance in the attention task. During task execution, however, fast reaction times were associated with high-amplitude beta and gamma oscillations with low LRTC. Finally, we show that reduced LRTC during the attention task compared to the rest condition correlates with better performance, while increased LRTC of oscillations from rest to attention is associated with reduced performance. To our knowledge, this is the first empirical evidence that 'resting-state criticality' of neuronal networks predicts swift behavioral responses in a sensorimotor task, and that steady attentive processing of visual stimuli requires brain dynamics with suppressed temporal complexity.

Controlling the Temporal Structure of Brain Oscillations by Focused Attention Meditation.

Our focus of attention naturally fluctuates between different sources of information even when we desire to focus on a single object. Focused attention (FA) meditation is associated with greater control over this process, yet the neuronal mechanisms underlying this ability are not entirely understood. Here, we hypothesize that the capacity of attention to transiently focus and swiftly change relates to the critical dynamics emerging when neuronal systems balance at a point of instability between order and disorder. In FA meditation, however, the ability to stay focused is trained, which may be associated with a more homogeneous brain state. To test this hypothesis, we applied analytical tools from criticality theory to EEG in meditation practitioners and meditation-naïve participants from two independent labs. We show that in practitioners-but not in controls-FA meditation strongly suppressed long-range temporal correlations (LRTC) of neuronal oscillations relative to eyes-closed rest with remarkable consistency across frequency bands and scalp locations. The ability to reduce LRTC during meditation increased after one year of additional training and was associated with the subjective experience of fully engaging one's attentional resources, also known as absorption. Sustained practice also affected normal waking brain dynamics as reflected in increased LRTC during an eyes-closed rest state, indicating that brain dynamics are altered beyond the meditative state. Taken together, our findings suggest that the framework of critical brain dynamics is promising for understanding neuronal mechanisms of meditative states and, specifically, we have identified a clear electrophysiological correlate of the FA meditation state.

Consistency of EEG source localization and connectivity estimates.

As the EEG inverse problem does not have a unique solution, the sources reconstructed from EEG and their connectivity properties depend on forward and inverse modeling parameters such as the choice of an anatomical template and electrical model, prior assumptions on the sources, and further implementational details. In order to use source connectivity analysis as a reliable research tool, there is a need for stability across a wider range of standard estimation routines. Using resting state EEG recordings of N=65 participants acquired within two studies, we present the first comprehensive assessment of the consistency of EEG source localization and functional/effective connectivity metrics across two anatomical templates (ICBM152 and Colin27), three electrical models (BEM, FEM and spherical harmonics expansions), three inverse methods (WMNE, eLORETA and LCMV), and three software implementations (Brainstorm, Fieldtrip and our own toolbox). Source localizations were found to be more stable across reconstruction pipelines than subsequent estimations of functional connectivity, while effective connectivity estimates where the least consistent. All results were relatively unaffected by the choice of the electrical head model, while the choice of the inverse method and source imaging package induced a considerable variability. In particular, a relatively strong difference was found between LCMV beamformer solutions on one hand and eLORETA/WMNE distributed inverse solutions on the other hand. We also observed a gradual decrease of consistency when results are compared between studies, within individual participants, and between individual participants. In order to provide reliable findings in the face of the observed variability, additional simulations involving interacting brain sources are required. Meanwhile, we encourage verification of the obtained results using more than one source imaging procedure.

Association Between Resting-State Microstates and Ratings on the Amsterdam Resting-State Questionnaire.

There is a gap in understanding on how physiologically observed activity is related to the subjective, internally oriented experience during resting state. Microstate analysis is a frequent approach to evaluate resting-state EEG. But the relationship of commonly observed resting-state microstates to psychological domains of resting is not clear. The Amsterdam Resting-State Questionnaire (ARSQ) was recently introduced, offering an effective way to quantify subjective states after a period of resting and associate these quantifiers to psychological and physiological variables. In a sample of 94 healthy volunteers who participated in closed-eyes 5 min resting session with concurrent EEG recording and subsequent filling of the ARSQ we evaluated parameters of microstate Classes A, B, C, D. We showed a moderate negative association between contribution (r = -0.40) of Class C and experienced somatic awareness (SA). The negative correlation between Class C and SA seems reasonable as Class C becomes more dominant when connections to contextual information (and bodily sensations as assessed with SA) are loosened (in reduced attention states, during sleep, hypnosis, or psychosis). We suggest that the use of questionnaires such as the ARSQ is helpful in exploring the variation of resting-state EEG parameters and its relationship to variation in sensory and non-sensory experiences.

Neuronal long-range temporal correlations and avalanche dynamics are correlated with behavioral scaling laws.

Scale-free fluctuations are ubiquitous in behavioral performance and neuronal activity. In time scales from seconds to hundreds of seconds, psychophysical dynamics and the amplitude fluctuations of neuronal oscillations are governed by power-law-form long-range temporal correlations (LRTCs). In millisecond time scales, neuronal activity comprises cascade-like neuronal avalanches that exhibit power-law size and lifetime distributions. However, it remains unknown whether these neuronal scaling laws are correlated with those characterizing behavioral performance or whether neuronal LRTCs and avalanches are related. Here, we show that the neuronal scaling laws are strongly correlated both with each other and with behavioral scaling laws. We used source reconstructed magneto- and electroencephalographic recordings to characterize the dynamics of ongoing cortical activity. We found robust power-law scaling in neuronal LRTCs and avalanches in resting-state data and during the performance of audiovisual threshold stimulus detection tasks. The LRTC scaling exponents of the behavioral performance fluctuations were correlated with those of concurrent neuronal avalanches and LRTCs in anatomically identified brain systems. The behavioral exponents also were correlated with neuronal scaling laws derived from a resting-state condition and with a similar anatomical topography. Finally, despite the difference in time scales, the scaling exponents of neuronal LRTCs and avalanches were strongly correlated during both rest and task performance. Thus, long and short time-scale neuronal dynamics are related and functionally significant at the behavioral level. These data suggest that the temporal structures of human cognitive fluctuations and behavioral variability stem from the scaling laws of individual and intrinsic brain dynamics.

Long-range temporal correlations in resting-state α oscillations predict human timing-error dynamics.

Human behavior is imperfect. This is notably clear during repetitive tasks in which sequences of errors or deviations from perfect performance result. These errors are not random, but show patterned fluctuations with long-range temporal correlations that are well described using power-law spectra P(f)∝1/f(ß), where ß is the power-law scaling exponent describing the decay in temporal correlations. The neural basis of temporal correlations in such behaviors is not known. Interestingly, long-range temporal correlations are a hallmark of amplitude fluctuations in resting-state neuronal oscillations. Here, we investigated whether the temporal dynamics in brain and behavior are related. Thirty-nine subjects' eyes-open rest EEG was measured. Next, subjects reproduced without feedback a 1 s interval by tapping with their right index finger. In line with previous reports, we found evidence for the presence of long-range temporal correlations both in the amplitude modulation of resting-state oscillations in multiple frequency bands and in the timing-error sequences. Frequency scaling exponents of finger tapping and amplitude modulation of oscillations exhibited large individual differences. Neuronal dynamics of resting-state alpha-band oscillations (9-13 Hz) recorded at precentral sites strongly predicted scaling exponents of tapping behavior. The results suggest that individual variation in resting-state brain dynamics offer a neural explanation for individual variation in the error dynamics of human behavior.

Individual differences in white matter diffusion affect sleep oscillations.

The characteristic oscillations of the sleeping brain, spindles and slow waves, show trait-like, within-subject stability and a remarkable interindividual variability that correlates with functionally relevant measures such as memory performance and intelligence. Yet, the mechanisms underlying these interindividual differences are largely unknown. Spindles and slow waves are affected by the recent history of learning and neuronal activation, indicating sensitivity to changes in synaptic strength and thus to the connectivity of the neuronal network. Because the structural backbone of this network is formed by white matter tracts, we hypothesized that individual differences in spindles and slow waves depend on the white matter microstructure across a distributed network. We recorded both diffusion-weighted magnetic resonance images and whole-night, high-density electroencephalography and investigated whether individual differences in sleep spindle and slow wave parameters were associated with diffusion tensor imaging metrics; white matter fractional anisotropy and axial diffusivity were quantified using tract-based spatial statistics. Individuals with higher spindle power had higher axial diffusivity in the forceps minor, the anterior corpus callosum, fascicles in the temporal lobe, and the tracts within and surrounding the thalamus. Individuals with a steeper rising slope of the slow wave had higher axial diffusivity in the temporal fascicle and frontally located white matter tracts (forceps minor, anterior corpus callosum). These results indicate that the profiles of sleep oscillations reflect not only the dynamics of the neuronal network at the synaptic level, but also the localized microstructural properties of its structural backbone, the white matter tracts.

Critical-state dynamics of avalanches and oscillations jointly emerge from balanced excitation/inhibition in neuronal networks.

Criticality has gained widespread interest in neuroscience as an attractive framework for understanding the character and functional implications of variability in brain activity. The metastability of critical systems maximizes their dynamic range, storage capacity, and computational power. Power-law scaling-a hallmark of criticality-has been observed on different levels, e.g., in the distribution of neuronal avalanches in vitro and in vivo, but also in the decay of temporal correlations in behavioral performance and ongoing oscillations in humans. An unresolved issue is whether power-law scaling on different organizational levels in the brain-and possibly in other hierarchically organized systems-can be related. Here, we show that critical-state dynamics of avalanches and oscillations jointly emerge in a neuronal network model when excitation and inhibition is balanced. The oscillatory activity of the model was qualitatively similar to what is typically observed in recordings of human resting-state MEG. We propose that homeostatic plasticity mechanisms tune this balance in healthy brain networks, and that it is essential for critical behavior on multiple levels of neuronal organization with ensuing functional benefits. Based on our network model, we introduce a concept of multi-level criticality in which power-law scaling can emerge on multiple time scales in oscillating networks.

External drive to inhibitory cells induces alternating episodes of high- and low-amplitude oscillations.

Electrical oscillations in neuronal network activity are ubiquitous in the brain and have been associated with cognition and behavior. Intriguingly, the amplitude of ongoing oscillations, such as measured in EEG recordings, fluctuates irregularly, with episodes of high amplitude alternating with episodes of low amplitude. Despite the widespread occurrence of amplitude fluctuations in many frequency bands and brain regions, the mechanisms by which they are generated are poorly understood. Here, we show that irregular transitions between sub-second episodes of high- and low-amplitude oscillations in the alpha/beta frequency band occur in a generic neuronal network model consisting of interconnected inhibitory and excitatory cells that are externally driven by sustained cholinergic input and trains of action potentials that activate excitatory synapses. In the model, we identify the action potential drive onto inhibitory cells, which represents input from other brain areas and is shown to desynchronize network activity, to be crucial for the emergence of amplitude fluctuations. We show that the duration distributions of high-amplitude episodes in the model match those observed in rat prefrontal cortex for oscillations induced by the cholinergic agonist carbachol. Furthermore, the mean duration of high-amplitude episodes varies in a bell-shaped manner with carbachol concentration, just as in mouse hippocampus. Our results suggest that amplitude fluctuations are a general property of oscillatory neuronal networks that can arise through background input from areas external to the network.

Resting-state oscillations reveal disturbed excitation-inhibition ratio in Alzheimer's disease patients.

An early disruption of neuronal excitation-inhibition (E-I) balance in preclinical animal models of Alzheimer's disease (AD) has been frequently reported, but is difficult to measure directly and non-invasively in humans. Here, we examined known and novel neurophysiological measures sensitive to E-I in patients across the AD continuum. Resting-state magnetoencephalography (MEG) data of 86 amyloid-biomarker-confirmed subjects across the AD continuum (17 patients diagnosed with subjective cognitive decline, 18 with mild cognitive impairment (MCI) and 51 with dementia due to probable AD (AD dementia)), 46 healthy elderly and 20 young control subjects were reconstructed to source-space. E-I balance was investigated by detrended fluctuation analysis (DFA), a functional E/I (fE/I) algorithm, and the aperiodic exponent of the power spectrum. We found a disrupted E-I ratio in AD dementia patients specifically, by a lower DFA, and a shift towards higher excitation, by a higher fE/I and a lower aperiodic exponent. Healthy subjects showed lower fE/I ratios (< 1.0) than reported in previous literature, not explained by age or choice of an arbitrary threshold parameter, which warrants caution in interpretation of fE/I results. Correlation analyses showed that a lower DFA (E-I imbalance) and a lower aperiodic exponent (more excitation) was associated with a worse cognitive score in AD dementia patients. In contrast, a higher DFA in the hippocampi of MCI patients was associated with a worse cognitive score. This MEG-study showed E-I imbalance, likely due to increased excitation, in AD dementia, but not in early stage AD patients. To accurately determine the direction of shift in E-I balance, validations of the currently used markers and additional in vivo markers of E-I are required.

Prediction of Behavioral Improvement Through Resting-State Electroencephalography and Clinical Severity in a Randomized Controlled Trial Testing Bumetanide in Autism Spectrum Disorder.

BACKGROUND: Mechanism-based treatments such as bumetanide are being repurposed for autism spectrum disorder. We recently reported beneficial effects on repetitive behavioral symptoms that might be related to regulating excitation-inhibition (E/I) balance in the brain. Here, we tested the neurophysiological effects of bumetanide and the relationship to clinical outcome variability and investigated the potential for machine learning-based predictions of meaningful clinical improvement. METHODS: Using modified linear mixed models applied to intention-to-treat population, we analyzed E/I-sensitive electroencephalography (EEG) measures before and after 91 days of treatment in the double-blind, randomized, placebo-controlled Bumetanide in Autism Medication and Biomarker study. Resting-state EEG of 82 subjects out of 92 participants (7-15 years) were available. Alpha frequency band absolute and relative power, central frequency, long-range temporal correlations, and functional E/I ratio treatment effects were related to the Repetitive Behavior Scale-Revised (RBS-R) and the Social Responsiveness Scale 2 as clinical outcomes. RESULTS: We observed superior bumetanide effects on EEG, reflected in increased absolute and relative alpha power and functional E/I ratio and in decreased central frequency. Associations between EEG and clinical outcome change were restricted to subgroups with medium to high RBS-R improvement. Using machine learning, medium and high RBS-R improvement could be predicted by baseline RBS-R score and EEG measures with 80% and 92% accuracy, respectively. CONCLUSIONS: Bumetanide exerts neurophysiological effects related to clinical changes in more responsive subsets, in whom prediction of improvement was feasible through EEG and clinical measures.

Scale-free modulation of resting-state neuronal oscillations reflects prolonged brain maturation in humans.

Human neuronal circuits undergo life-long functional reorganization with profound effects on cognition and behavior. Well documented prolonged development of anatomical brain structures includes white and gray matter changes that continue into the third decade of life. We investigated resting-state EEG oscillations in 1433 subjects from 5 to 71 years. Neuronal oscillations exhibit scale-free amplitude modulation as reflected in power-law decay of autocorrelations--also known as long-range temporal correlations (LRTC)--which was assessed by detrended fluctuation analysis. We observed pronounced increases in LRTC from childhood to adolescence, during adolescence, and even into early adulthood (∼25 years of age) after which the temporal structure stabilized. A principal component analysis of the spatial distribution of LRTC revealed increasingly uniform scores across the scalp. Together, these findings indicate that the scale-free modulation of resting-state oscillations reflects brain maturation, and suggests that scaling analysis may prove useful as a biomarker of pathophysiology in neurodevelopmental disorders such as attention deficit hyperactivity disorder and schizophrenia.

Altered temporal correlations in parietal alpha and prefrontal theta oscillations in early-stage Alzheimer disease.

Encoding and retention of information in memory are associated with a sustained increase in the amplitude of neuronal oscillations for up to several seconds. We reasoned that coordination of oscillatory activity over time might be important for memory and, therefore, that the amplitude modulation of oscillations may be abnormal in Alzheimer disease (AD). To test this hypothesis, we measured magnetoencephalography (MEG) during eyes-closed rest in 19 patients diagnosed with early-stage AD and 16 age-matched control subjects and characterized the autocorrelation structure of ongoing oscillations using detrended fluctuation analysis and an analysis of the life- and waiting-time statistics of oscillation bursts. We found that Alzheimer's patients had a strongly reduced incidence of alpha-band oscillation bursts with long life- or waiting-times (< 1 s) over temporo-parietal regions and markedly weaker autocorrelations on long time scales (1-25 seconds). Interestingly, the life- and waiting-times of theta oscillations over medial prefrontal regions were greatly increased. Whereas both temporo-parietal alpha and medial prefrontal theta oscillations are associated with retrieval and retention of information, metabolic and structural deficits in early-stage AD are observed primarily in temporo-parietal areas, suggesting that the enhanced oscillations in medial prefrontal cortex reflect a compensatory mechanism. Together, our results suggest that amplitude modulation of neuronal oscillations is important for cognition and that indices of amplitude dynamics of oscillations may prove useful as neuroimaging biomarkers of early-stage AD.

Robin's Viewer: Using deep-learning predictions to assist EEG annotation.

Machine learning techniques such as deep learning have been increasingly used to assist EEG annotation, by automating artifact recognition, sleep staging, and seizure detection. In lack of automation, the annotation process is prone to bias, even for trained annotators. On the other hand, completely automated processes do not offer the users the opportunity to inspect the models' output and re-evaluate potential false predictions. As a first step toward addressing these challenges, we developed Robin's Viewer (RV), a Python-based EEG viewer for annotating time-series EEG data. The key feature distinguishing RV from existing EEG viewers is the visualization of output predictions of deep-learning models trained to recognize patterns in EEG data. RV was developed on top of the plotting library Plotly, the app-building framework Dash, and the popular M/EEG analysis toolbox MNE. It is an open-source, platform-independent, interactive web application, which supports common EEG-file formats to facilitate easy integration with other EEG toolboxes. RV includes common features of other EEG viewers, e.g., a view-slider, tools for marking bad channels and transient artifacts, and customizable preprocessing. Altogether, RV is an EEG viewer that combines the predictive power of deep-learning models and the knowledge of scientists and clinicians to optimize EEG annotation. With the training of new deep-learning models, RV could be developed to detect clinical patterns other than artifacts, for example sleep stages and EEG abnormalities.

Intellectually able adults with autism spectrum disorder show typical resting-state EEG activity.

There is broad interest in discovering quantifiable physiological biomarkers for psychiatric disorders to aid diagnostic assessment. However, finding biomarkers for autism spectrum disorder (ASD) has proven particularly difficult, partly due to high heterogeneity. Here, we recorded five minutes eyes-closed rest electroencephalography (EEG) from 186 adults (51% with ASD and 49% without ASD) and investigated the potential of EEG biomarkers to classify ASD using three conventional machine learning models with two-layer cross-validation. Comprehensive characterization of spectral, temporal and spatial dimensions of source-modelled EEG resulted in 3443 biomarkers per recording. We found no significant group-mean or group-variance differences for any of the EEG features. Interestingly, we obtained validation accuracies above 80%; however, the best machine learning model merely distinguished ASD from the non-autistic comparison group with a mean balanced test accuracy of 56% on the entirely unseen test set. The large drop in model performance between validation and testing, stress the importance of rigorous model evaluation, and further highlights the high heterogeneity in ASD. Overall, the lack of significant differences and weak classification indicates that, at the group level, intellectually able adults with ASD show remarkably typical resting-state EEG.

Improved Manual Annotation of EEG Signals through Convolutional Neural Network Guidance.

The development of validated algorithms for automated handling of artifacts is essential for reliable and fast processing of EEG signals. Recently, there have been methodological advances in designing machine-learning algorithms to improve artifact detection of trained professionals who usually meticulously inspect and manually annotate EEG signals. However, validation of these methods is hindered by the lack of a gold standard as data are mostly private and data annotation is time consuming and error prone. In the effort to circumvent these issues, we propose an iterative learning model to speed up and reduce errors of manual annotation of EEG. We use a convolutional neural network (CNN) to train on expert-annotated eyes-open and eyes-closed resting-state EEG data from typically developing children (n = 30) and children with neurodevelopmental disorders (n = 141). To overcome the circular reasoning of aiming to develop a new algorithm and benchmarking to a manually-annotated gold standard, we instead aim to improve the gold standard by revising the portion of the data that was incorrectly learned by the network. When blindly presented with the selected signals for re-assessment (23% of the data), the two independent expert-annotators changed the annotation in 25% of the cases. Subsequently, the network was trained on the expert-revised gold standard, which resulted in improved separation between artifacts and nonartifacts as well as an increase in balanced accuracy from 74% to 80% and precision from 59% to 76%. These results show that CNNs are promising to enhance manual annotation of EEG artifacts and can be improved further with better gold-standard data.