Recurring Functional Interactions Predict Network Architecture of Interictal and Ictal States in Neocortical Epilepsy

A comparative study between a power and a connectivity sEEG biomarker for seizure-onset zone identification in temporal lobe epilepsy

BACKGROUND

Ictal stereo-encephalography (sEEG) biomarkers for seizure onset zone (SOZ) localization can be classified depending on whether they target abnormalities in signal power or functional connectivity between signals, and they may depend on the frequency band and time window at which they are estimated.

NEW METHOD

This work aimed to compare and optimize the performance of a power and a connectivity-based biomarker to identify SOZ contacts from ictal sEEG recordings. To do so, we used a previously introduced power-based measure, the normalized mean activation (nMA), which quantifies the ictal average power activation. Similarly, we defined the normalized mean strength (nMS), to quantify the ictal mean functional connectivity of every contact with the rest. The optimal frequency bands and time windows were selected based on optimizing AUC and F2-score.

RESULTS

The analysis was performed on a dataset of 67 seizures from 10 patients with pharmacoresistant temporal lobe epilepsy. Our results suggest that the power-based biomarker generally performs better for the detection of SOZ than the connectivity-based one. However, an equivalent performance level can be achieved when both biomarkers are independently optimized. Optimal performance was achieved in the beta and lower-gamma range for the power biomarker and in the lower- and higher-gamma range for connectivity, both using a 20 or 30 s period after seizure onset.

CONCLUSIONS

The results of this study highlight the importance of this optimization step over frequency and time windows when comparing different SOZ discrimination biomarkers. This information should be considered when training SOZ classifiers on retrospective patients' data for clinical applications.

Chronic intracranial EEG recordings and interictal spike rate reveal multiscale temporal modulations in seizure states

Many biological processes are modulated by rhythms on circadian and multidien timescales. In focal epilepsy, various seizure features, such as spread and duration, can change from one seizure to the next within the same patient. However, the specific timescales of this variability, as well as the specific seizure characteristics that change over time, are unclear. Here, in a cross-sectional observational study, we analysed within-patient seizure variability in 10 patients with chronic intracranial EEG recordings (185-767 days of recording time, 57-452 analysed seizures/patient). We characterized the seizure evolutions as sequences of a finite number of patient-specific functional seizure network states. We then compared seizure network state occurrence and duration to (1) time since implantation and (2) patient-specific circadian and multidien cycles in interictal spike rate. In most patients, the occurrence or duration of at least one seizure network state was associated with the time since implantation. Some patients had one or more seizure network states that were associated with phases of circadian and/or multidien spike rate cycles. A given seizure network state's occurrence and duration were usually not associated with the same timescale. Our results suggest that different time-varying factors modulate within-patient seizure evolutions over multiple timescales, with separate processes modulating a seizure network state's occurrence and duration. These findings imply that the development of time-adaptive treatments in epilepsy must account for several separate properties of epileptic seizures and similar principles likely apply to other neurological conditions.

Multilayer brain networks can identify the epileptogenic zone and seizure dynamics

Seizure generation, propagation, and termination occur through spatiotemporal brain networks. In this paper, we demonstrate the significance of large-scale brain interactions in high-frequency (80-200Hz) for the identification of the epileptogenic zone (EZ) and seizure evolution. To incorporate the continuity of neural dynamics, here we have modeled brain connectivity constructed from stereoelectroencephalography (SEEG) data during seizures using multilayer networks. After introducing a new measure of brain connectivity for temporal networks, named multilayer eigenvector centrality (mlEVC), we applied a consensus hierarchical clustering on the developed model to identify the EZ as a cluster of nodes with distinctive brain connectivity in the ictal period. Our algorithm could successfully predict electrodes inside the resected volume as EZ for 88% of participants, who all were seizure-free for at least 12 months after surgery. Our findings illustrated significant and unique desynchronization between EZ and the rest of the brain in the early to mid-seizure. We showed that aging and the duration of epilepsy intensify this desynchronization, which can be the outcome of abnormal neuroplasticity. Additionally, we illustrated that seizures evolve with various network topologies, confirming the existence of different epileptogenic networks in each patient. Our findings suggest not only the importance of early intervention in epilepsy but possible factors that correlate with disease severity. Moreover, by analyzing the propagation patterns of different seizures, we demonstrate the necessity of collecting sufficient data for identifying epileptogenic networks.

Interictal Connectivity Revealed by Granger Analysis of Stereoelectroencephalography: Association With Ictal Onset Zone, Resection, and Outcome

BACKGROUND

Stereoelectroencephalography (sEEG) facilitates electrical sampling and evaluation of complex deep-seated, dispersed, and multifocal locations. Granger causality (GC), previously used to study seizure networks using interictal data from subdural grids, may help identify the seizure-onset zone from interictal sEEG recordings.

OBJECTIVE

To examine whether statistical analysis of interictal sEEG helps identify surgical target sites and whether surgical resection of highly ranked nodes correspond to favorable outcomes.

METHODS

Ten minutes of extraoperative recordings from sequential patients who underwent sEEG evaluation were analyzed (n = 20). GC maps were compared with clinically defined surgical targets using rank order statistics. Outcomes of patients with focal resection/ablation with median follow-up of 3.6 years were classified as favorable (Engel 1, 2) or poor (Engel 3, 4) to assess their relationship with the removal of highly ranked nodes using the Wilcoxon rank-sum test.

RESULTS

In 12 of 20 cases, the rankings of contacts (based on the sum of outward connection weights) mapped to the seizure-onset zone showed higher causal node connectivity than predicted by chance ( P ≤ .02). A very low aggregate probability ( P < 10 -18 , n = 20) suggests that causal node connectivity predicts seizure networks. In 8 of 16 with outcome data, causal connectivity in the resection was significantly greater than in the remaining contacts ( P ≤ .05). We found a significant association between favorable outcome and the presence of highly ranked nodes in the resection ( P < .05).

CONCLUSION

Granger analysis can identify seizure foci from interictal sEEG and correlates highly ranked nodes with favorable outcome, potentially informing surgical decision-making without reliance on ictal recordings.

Brain network analysis of interictal epileptiform discharges from ECoG to identify epileptogenic zone in pediatric patients with epilepsy and focal cortical dysplasia type II: A retrospective study

Objective

For patients with drug-resistant focal epilepsy, intracranial monitoring remains the gold standard for surgical intervention. Focal cortical dysplasia (FCD) is the most common cause of pharmacoresistant focal epilepsy in pediatric patients who usually develop seizures in early childhood. Timely removal of the epileptogenic zone (EZ) is necessary to achieve lasting seizure freedom and favorable developmental and cognitive outcomes to improve the quality of life. We applied brain network analysis to investigate potential biomarkers for the diagnosis of EZ that will aid in the resection for pediatric focal epilepsy patients with FCD type II.

Methods

Ten pediatric patients with focal epilepsy diagnosed as FCD type II and that had a follow-up after resection surgery (Engel class I [n = 9] and Engel class II [n = 1]) were retrospectively included. Time-frequency analysis of phase transfer entropy, graph theory analysis, and power spectrum compensation were combined to calculate brain network parameters based on interictal epileptiform discharges from ECoG.

Results

Clustering coefficient, local efficiency, node out-degree, and node out-strength with higher values are the most reliable biomarkers for the delineation of EZ, and the differences between EZ and margin zone (MZ), and EZ and normal zone (NZ) were significant (p < 0.05; Mann-Whitney U-test, two-tailed). In particular, the difference between MZ and NZ was significant for patients with frontal FCD (MZ > NZ; p < 0.05) but was not significant for patients with extra-frontal FCD.

Conclusions

Brain network analysis, based on the combination of time-frequency analysis of phase transfer entropy, graph theory analysis, and power spectrum compensation, can aid in the diagnosis of EZ for pediatric focal epilepsy patients with FCD type II.

Decoding Intracranial EEG With Machine Learning: A Systematic Review

Advances in intracranial electroencephalography (iEEG) and neurophysiology have enabled the study of previously inaccessible brain regions with high fidelity temporal and spatial resolution. Studies of iEEG have revealed a rich neural code subserving healthy brain function and which fails in disease states. Machine learning (ML), a form of artificial intelligence, is a modern tool that may be able to better decode complex neural signals and enhance interpretation of these data. To date, a number of publications have applied ML to iEEG, but clinician awareness of these techniques and their relevance to neurosurgery, has been limited. The present work presents a review of existing applications of ML techniques in iEEG data, discusses the relative merits and limitations of the various approaches, and examines potential avenues for clinical translation in neurosurgery. One-hundred-seven articles examining artificial intelligence applications to iEEG were identified from 3 databases. Clinical applications of ML from these articles were categorized into 4 domains: i) seizure analysis, ii) motor tasks, iii) cognitive assessment, and iv) sleep staging. The review revealed that supervised algorithms were most commonly used across studies and often leveraged publicly available timeseries datasets. We conclude with recommendations for future work and potential clinical applications.

Dynamical Network Models From EEG and MEG for Epilepsy Surgery—A Quantitative Approach

There is an urgent need for more informative quantitative techniques that non-invasively and objectively assess strategies for epilepsy surgery. Invasive intracranial electroencephalography (iEEG) remains the clinical gold standard to investigate the nature of the epileptogenic zone (EZ) before surgical resection. However, there are major limitations of iEEG, such as the limited spatial sampling and the degree of subjectivity inherent in the analysis and clinical interpretation of iEEG data. Recent advances in network analysis and dynamical network modeling provide a novel aspect toward a more objective assessment of the EZ. The advantage of such approaches is that they are data-driven and require less or no human input. Multiple studies have demonstrated success using these approaches when applied to iEEG data in characterizing the EZ and predicting surgical outcomes. However, the limitations of iEEG recordings equally apply to these studies—limited spatial sampling and the implicit assumption that iEEG electrodes, whether strip, grid, depth or stereo EEG (sEEG) arrays, are placed in the correct location. Therefore, it is of interest to clinicians and scientists to see whether the same analysis and modeling techniques can be applied to whole-brain, non-invasive neuroimaging data (from MRI-based techniques) and neurophysiological data (from MEG and scalp EEG recordings), thus removing the limitation of spatial sampling, while safely and objectively characterizing the EZ. This review aims to summarize current state of the art non-invasive methods that inform epilepsy surgery using network analysis and dynamical network models. We also present perspectives on future directions and clinical applications of these promising approaches.

Comparison of Resampling Techniques for Imbalanced Datasets in Machine Learning: Application to Epileptogenic Zone Localization From Interictal Intracranial EEG Recordings in Patients With Focal Epilepsy

Aim: In neuroscience research, data are quite often characterized by an imbalanced distribution between the majority and minority classes, an issue that can limit or even worsen the prediction performance of machine learning methods. Different resampling procedures have been developed to face this problem and a lot of work has been done in comparing their effectiveness in different scenarios. Notably, the robustness of such techniques has been tested among a wide variety of different datasets, without considering the performance of each specific dataset. In this study, we compare the performances of different resampling procedures for the imbalanced domain in stereo-electroencephalography (SEEG) recordings of the patients with focal epilepsies who underwent surgery.

Methods: We considered data obtained by network analysis of interictal SEEG recorded from 10 patients with drug-resistant focal epilepsies, for a supervised classification problem aimed at distinguishing between the epileptogenic and non-epileptogenic brain regions in interictal conditions. We investigated the effectiveness of five oversampling and five undersampling procedures, using 10 different machine learning classifiers. Moreover, six specific ensemble methods for the imbalanced domain were also tested. To compare the performances, Area under the ROC curve (AUC), F-measure, Geometric Mean, and Balanced Accuracy were considered.

Results: Both the resampling procedures showed improved performances with respect to the original dataset. The oversampling procedure was found to be more sensitive to the type of classification method employed, with Adaptive Synthetic Sampling (ADASYN) exhibiting the best performances. All the undersampling approaches were more robust than the oversampling among the different classifiers, with Random Undersampling (RUS) exhibiting the best performance despite being the simplest and most basic classification method.

Conclusions: The application of machine learning techniques that take into consideration the balance of features by resampling is beneficial and leads to more accurate localization of the epileptogenic zone from interictal periods. In addition, our results highlight the importance of the type of classification method that must be used together with the resampling to maximize the benefit to the outcome.

Long-term brain network reorganization predicts responsive neurostimulation outcomes for focal epilepsy

Responsive neurostimulation (RNS) devices, able to detect imminent seizures and to rapidly deliver electrical stimulation to the brain, are effective in reducing seizures in some patients with focal epilepsy. However, therapeutic response to RNS is often slow, is highly variable, and defies prognostication based on clinical factors. A prevailing view holds that RNS efficacy is primarily mediated by acute seizure termination; yet, stimulations greatly outnumber seizures and occur mostly in the interictal state, suggesting chronic modulation of brain networks that generate seizures. Here, using years-long intracranial neural recordings collected during RNS therapy, we found that patients with the greatest therapeutic benefit undergo progressive, frequency-dependent reorganization of interictal functional connectivity. The extent of this reorganization scales directly with seizure reduction and emerges within the first year of RNS treatment, enabling potential early prediction of therapeutic response. Our findings reveal a mechanism for RNS that involves network plasticity and may inform development of next-generation devices for epilepsy.

Prediction of Seizure Recurrence. A Note of Caution

Great strides have been made recently in documenting that machine-learning programs can predict seizure occurrence in people who have epilepsy. Along with this progress have come claims that appear to us to be a bit premature. We anticipate that many people will benefit from seizure prediction. We also doubt that all will benefit. Although machine learning is a useful tool for aiding discovery, we believe that the greatest progress will come from deeper understanding of seizures, epilepsy, and the EEG features that enable seizure prediction. In this essay, we lay out reasons for optimism and skepticism.

Semi-automatic Extraction of Functional Dynamic Networks Describing Patient's Epileptic Seizures

Intracranial electroencephalography (EEG) studies using stereotactic EEG (SEEG) have shown that during seizures, epileptic activity spreads across several anatomical regions from the seizure onset zone toward remote brain areas. A full and objective characterization of this patient-specific time-varying network is crucial for optimal surgical treatment. Functional connectivity (FC) analysis of SEEG signals recorded during seizures enables to describe the statistical relations between all pairs of recorded signals. However, extracting meaningful information from those large datasets is time consuming and requires high expertise. In the present study, we first propose a novel method named Brain-wide Time-varying Network Decomposition (BTND) to characterize the dynamic epileptogenic networks activated during seizures in individual patients recorded with SEEG electrodes. The method provides a number of pathological FC subgraphs with their temporal course of activation. The method can be applied to several seizures of the patient to extract reproducible subgraphs. Second, we compare the activated subgraphs obtained by the BTND method with visual interpretation of SEEG signals recorded in 27 seizures from nine different patients. As a whole, we found that activated subgraphs corresponded to brain regions involved during the course of the seizures and their time course was highly consistent with classical visual interpretation. We believe that the proposed method can complement the visual analysis of SEEG signals recorded during seizures by highlighting and characterizing the most significant parts of epileptic networks with their activation dynamics.

Early seizure spread and epilepsy surgery: A systematic review

OBJECTIVE

A fundamental question in epilepsy surgery is how to delineate the margins of cortex that must be resected to result in seizure freedom. Whether and which areas showing seizure activity early in ictus must be removed to avoid postoperative recurrence of seizures is an area of ongoing research. Seizure spread dynamics in the initial seconds of ictus are often correlated with postoperative outcome; there is neither a consensus definition of early spread nor a concise summary of the existing literature linking seizure spread to postsurgical seizure outcomes. The present study is intended to summarize the literature that links seizure spread to postoperative seizure outcome and to provide a framework for quantitative assessment of early seizure spread.

METHODS

A systematic review was carried out according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. A Medline search identified clinical studies reporting data on seizure spread measured by intracranial electrodes, having at least 10 subjects and reporting at least 1-year postoperative outcome in the English literature from 1990 to 2019. Studies were evaluated regarding support for a primary hypothesis: Areas of early seizure spread represent cortex with seizure-generating potential.

RESULTS

The search yielded 4562 studies: 15 studies met inclusion criteria and 7 studies supported the primary hypothesis. The methods and metrics used to describe seizure spread were heterogenous. The timeframe of seizure spread associated with seizure outcome ranged from 1-14 seconds, with large, well-designed, retrospective studies pointing to 3-10 seconds as most likely to provide meaningful correlates of postoperative seizure freedom.

SIGNIFICANCE

The complex correlation between electrophysiologic seizure spread and the potential for seizure generation needs further elucidation. Prospective cohort studies or trials are needed to evaluate epilepsy surgery targeting cortex involved in the first 3-10 seconds of ictus.

Homeostatic Plasticity in Epilepsy

In the healthy brain, neuronal excitability and synaptic strength are homeostatically regulated to keep neuronal network activity within physiological boundaries. Epilepsy is characterized by episodes of highly synchronized firing across in widespread neuronal populations, due to a failure in regulation of network activity. Here we consider epilepsy as a failure of homeostatic plasticity or as a maladaptive response to perturbations in the activity. How homeostatic compensation is involved in epileptogenic processes or in the chronic phase of epilepsy, is still debated. Although several theories have been proposed, there is relatively little experimental evidence to evaluate them. In this perspective, we will discuss recent results that shed light on the potential role of homeostatic plasticity in epilepsy. First, we will present some recent insights on how homeostatic compensations are probably active before and during epileptogenesis and how their actions are temporally regulated and closely dependent on the progression of pathology. Then, we will consider the dual role of transcriptional regulation during epileptogenesis, and finally, we will underline the importance of homeostatic plasticity in the context of therapeutic interventions for epilepsy. While classic pharmacological interventions may be counteracted by the epileptic brain to maintain its potentially dysfunctional set point, novel therapeutic approaches may provide the neuronal network with the tools necessary to restore its physiological balance.

Seizure pathways change on circadian and slower timescales in individual patients with focal epilepsy

Personalized medicine requires that treatments adapt to not only the patient but also changing factors within each individual. Although epilepsy is a dynamic disorder characterized by pathological fluctuations in brain state, surprisingly little is known about whether and how seizures vary in the same patient. We quantitatively compared within-patient seizure network evolutions using intracranial electroencephalographic (iEEG) recordings of over 500 seizures from 31 patients with focal epilepsy (mean 16.5 seizures per patient). In all patients, we found variability in seizure paths through the space of possible network dynamics. Seizures with similar pathways tended to occur closer together in time, and a simple model suggested that seizure pathways change on circadian and/or slower timescales in the majority of patients. These temporal relationships occurred independent of whether the patient underwent antiepileptic medication reduction. Our results suggest that various modulatory processes, operating at different timescales, shape within-patient seizure evolutions, leading to variable seizure pathways that may require tailored treatment approaches.

Early childhood developmental functional connectivity of autistic brains with non-negative matrix factorization

Autism spectrum disorder (ASD) is associated with altered patterns of over- and under-connectivity of neural circuits. Age-related changes in neural connectivities remain unclear for autistic children as compared with normal children. In this study, a parts-based network-decomposition technique, known as non-negative matrix factorization (NMF), was applied to identify a set of possible abnormal connectivity patterns in brains affected by ASD, using resting-state electroencephalographic (EEG) data. Age-related changes in connectivities in both inter- and intra-hemispheric areas were studied in a total of 256 children (3-6 years old), both with and without ASD. The findings included the following: (1) the brains of children affected by ASD were characterized by a general trend toward long-range under-connectivity, particularly in interhemispheric connections, combined with short-range over-connectivity; (2) long-range connections were often associated with slower rhythms (δ and θ), whereas synchronization of short-range networks tended to be associated with faster frequencies (α and β); and (3) the α-band specific patterns of interhemispheric connections in ASD could be the most prominent during early childhood neurodevelopment. Therefore, NMF would be useful for further exploring the early childhood developmental functional connectivity of children aged 3-6 with ASD as well as with typical development. Additionally, long-range interhemispheric alterations in connectivity may represent a potential biomarker for the identification of ASD.

Emerging roles of network analysis for epilepsy

In recent years there has been increasing interest in applying network science tools to EEG data. At the 2018 American Epilepsy Society conference in New Orleans, LA, the yearly session of the Engineering and Neurostimulation Special Interest Group focused on emerging, translational technologies to analyze seizure networks. Each speaker demonstrated practical examples of how network tools can be utilized in clinical care and provide additional data to help care for patients with intractable epilepsy. The groups presented advances using tools from functional connectivity, control theory, and graph theory to analyze human EEG data. These tools have great potential to augment clinical interpretation of EEG signals.

Beyond the Epileptic Focus: Functional Epileptic Networks in Focal Epilepsy

Focal epilepsy can be conceptualized as a network disorder, and the functional epileptic network can be described as a complex system of multiple brain areas that interact dynamically to generate epileptic activity. However, we still do not fully understand the functional architecture of epileptic networks. We studied a cohort of 21 patients with extratemporal focal epilepsy. We used independent component analysis of functional magnetic resonance imaging (fMRI) data. In order to identify the epilepsy-related components, we examined the general linear model-derived electroencephalography-fMRI (EEG–fMRI) time courses associated with interictal epileptic activity as intrinsic hemodynamic epileptic biomarkers. Independent component analysis revealed components related to the epileptic time courses in all 21 patients. Each epilepsy-related component described a network of spatially distributed brain areas that corresponded to the specific epileptic network in each patient. We also provided evidence for the interaction between the epileptic activity generated at the epileptic network and the physiological resting state networks. Our findings suggest that independent component analysis, guided by EEG–fMRI epileptic time courses, have the potential to define the functional architecture of the epileptic network in a noninvasive way. These data could be useful in planning invasive EEG electrode placement, guiding surgical resections, and more effective therapeutic interventions.

Functional control of electrophysiological network architecture using direct neurostimulation in humans

Chronically implantable neurostimulation devices are becoming a clinically viable option for treating patients with neurological disease and psychiatric disorders. Neurostimulation offers the ability to probe and manipulate distributed networks of interacting brain areas in dysfunctional circuits. Here, we use tools from network control theory to examine the dynamic reconfiguration of functionally interacting neuronal ensembles during targeted neurostimulation of cortical and subcortical brain structures. By integrating multimodal intracranial recordings and diffusion-weighted imaging from patients with drug-resistant epilepsy, we test hypothesized structural and functional rules that predict altered patterns of synchronized local field potentials. We demonstrate the ability to predictably reconfigure functional interactions depending on stimulation strength and location. Stimulation of areas with structurally weak connections largely modulates the functional hubness of downstream areas and concurrently propels the brain towards more difficult-to-reach dynamical states. By using focal perturbations to bridge large-scale structure, function, and markers of behavior, our findings suggest that stimulation may be tuned to influence different scales of network interactions driving cognition.

Brain stimulation devices capable of perturbing the physiological state of neural systems are rapidly gaining popularity for their potential to treat neurological and psychiatric disease. A root problem is that underlying dysfunction spans a large-scale network of brain regions, requiring the ability to control the complex interactions between multiple brain areas. Here, we use tools from network control theory to examine the dynamic reconfiguration of functionally interacting neuronal ensembles during targeted neurostimulation of cortical and subcortical brain structures. We demonstrate the ability to predictably reconfigure patterns of interactions between functional brain areas by modulating the strength and location of stimulation. Our findings have high significance for designing stimulation protocols capable of modulating distributed neural circuits in the human brain.

Characterizing the role of the structural connectome in seizure dynamics

How does the human brain's structural scaffold give rise to its intricate functional dynamics? This is a central question in translational neuroscience that is particularly relevant to epilepsy, a disorder affecting over 50 million subjects worldwide. Treatment for medication-resistant focal epilepsy is often structural-through surgery or laser ablation-but structural targets, particularly in patients without clear lesions, are largely based on functional mapping via intracranial EEG. Unfortunately, the relationship between structural and functional connectivity in the seizing brain is poorly understood. In this study, we quantify structure-function coupling, specifically between white matter connections and intracranial EEG, across pre-ictal and ictal periods in 45 seizures from nine patients with unilateral drug-resistant focal epilepsy. We use high angular resolution diffusion imaging (HARDI) tractography to construct structural connectivity networks and correlate these networks with time-varying broadband and frequency-specific functional networks derived from coregistered intracranial EEG. Across all frequency bands, we find significant increases in structure-function coupling from pre-ictal to ictal periods. We demonstrate that short-range structural connections are primarily responsible for this increase in coupling. Finally, we find that spatiotemporal patterns of structure-function coupling are highly stereotyped for each patient. These results suggest that seizures harness the underlying structural connectome as they propagate. Mapping the relationship between structural and functional connectivity in epilepsy may inform new therapies to halt seizure spread, and pave the way for targeted patient-specific interventions.

High interictal connectivity within the resection zone is associated with favorable post-surgical outcomes in focal epilepsy patients

Patients with drug-resistant focal epilepsy are often candidates for invasive surgical therapies. In these patients, it is necessary to accurately localize seizure generators to ensure seizure freedom following intervention. While intracranial electroencephalography (iEEG) is the gold standard for mapping networks for surgery, this approach requires inducing and recording seizures, which may cause patient morbidity. The goal of this study is to evaluate the utility of mapping interictal (non-seizure) iEEG networks to identify targets for surgical treatment. We analyze interictal iEEG recordings and neuroimaging from 27 focal epilepsy patients treated via surgical resection. We generate interictal functional networks by calculating pairwise correlation of iEEG signals across different frequency bands. Using image coregistration and segmentation, we identify electrodes falling within surgically resected tissue (i.e. the resection zone), and compute node-level and edge-level synchrony in relation to the resection zone. We further associate these metrics with post-surgical outcomes. Greater overlap between resected electrodes and highly synchronous electrodes is associated with favorable post-surgical outcomes. Additionally, good-outcome patients have significantly higher connectivity localized within the resection zone compared to those with poorer postoperative seizure control. This finding persists following normalization by a spatially-constrained null model. This study suggests that spatially-informed interictal network synchrony measures can distinguish between good and poor post-surgical outcomes. By capturing clinically-relevant information during interictal periods, our method may ultimately reduce the need for prolonged invasive implants and provide insights into the pathophysiology of an epileptic brain. We discuss next steps for translating these findings into a prospectively useful clinical tool.

•

We analyze interictal iEEG recordings and neuroimaging from epilepsy patients

•

We find that high interictal strength selectivity is associated with better outcomes

•

This effect appears to be driven largely by connectivity within the resection zone

•

Interictal recordings can guide identification of seizure-generating networks

Dynamics of large-scale electrophysiological networks: A technical review

For several years it has been argued that neural synchronisation is crucial for cognition. The idea that synchronised temporal patterns between different neural groups carries information above and beyond the isolated activity of these groups has inspired a shift in focus in the field of functional neuroimaging. Specifically, investigation into the activation elicited within certain regions by some stimulus or task has, in part, given way to analysis of patterns of co-activation or functional connectivity between distal regions. Recently, the functional connectivity community has been looking beyond the assumptions of stationarity that earlier work was based on, and has introduced methods to incorporate temporal dynamics into the analysis of connectivity. In particular, non-invasive electrophysiological data (magnetoencephalography/electroencephalography (MEG/EEG)), which provides direct measurement of whole-brain activity and rich temporal information, offers an exceptional window into such (potentially fast) brain dynamics. In this review, we discuss challenges, solutions, and a collection of analysis tools that have been developed in recent years to facilitate the investigation of dynamic functional connectivity using these imaging modalities. Further, we discuss the applications of these approaches in the study of cognition and neuropsychiatric disorders. Finally, we review some existing developments that, by using realistic computational models, pursue a deeper understanding of the underlying causes of non-stationary connectivity.

Modeling and interpreting mesoscale network dynamics

Recent advances in brain imaging techniques, measurement approaches, and storage capacities have provided an unprecedented supply of high temporal resolution neural data. These data present a remarkable opportunity to gain a mechanistic understanding not just of circuit structure, but also of circuit dynamics, and its role in cognition and disease. Such understanding necessitates a description of the raw observations, and a delineation of computational models and mathematical theories that accurately capture fundamental principles behind the observations. Here we review recent advances in a range of modeling approaches that embrace the temporally-evolving interconnected structure of the brain and summarize that structure in a dynamic graph. We describe recent efforts to model dynamic patterns of connectivity, dynamic patterns of activity, and patterns of activity atop connectivity. In the context of these models, we review important considerations in statistical testing, including parametric and non-parametric approaches. Finally, we offer thoughts on careful and accurate interpretation of dynamic graph architecture, and outline important future directions for method development.

Localization of the Epileptogenic Zone Using Interictal MEG and Machine Learning in a Large Cohort of Drug-Resistant Epilepsy Patients

Objective: Epilepsy surgery results in seizure freedom in the majority of drug-resistant patients. To improve surgery outcome we studied whether MEG metrics combined with machine learning can improve localization of the epileptogenic zone, thereby enhancing the chance of seizure freedom.

Methods: Presurgical interictal MEG recordings of 94 patients (64 seizure-free >1y post-surgery) were analyzed to extract four metrics in source space: delta power, low-to-high-frequency power ratio, functional connectivity (phase lag index), and minimum spanning tree betweenness centrality. At the group level, we estimated the overlap of the resection area with the five highest values for each metric and determined whether this overlap differed between surgery outcomes. At the individual level, those metrics were used in machine learning classifiers (linear support vector machine (SVM) and random forest) to distinguish between resection and non-resection areas and between surgery outcome groups.

Results: The highest values, for all metrics, overlapped with the resection area in more than half of the patients, but the overlap did not differ between surgery outcome groups. The classifiers distinguished the resection areas from non-resection areas with 59.94% accuracy (95% confidence interval: 59.67–60.22%) for SVM and 60.34% (59.98–60.71%) for random forest, but could not differentiate seizure-free from not seizure-free patients [43.77% accuracy (42.08–45.45%) for SVM and 49.03% (47.25–50.82%) for random forest].

Significance: All four metrics localized the resection area but did not distinguish between surgery outcome groups, demonstrating that metrics derived from interictal MEG correspond to expert consensus based on several presurgical evaluation modalities, but do not yet localize the epileptogenic zone. Metrics should be improved such that they correspond to the resection area in seizure-free patients but not in patients with persistent seizures. It is important to test such localization strategies at an individual level, for example by using machine learning or individualized models, since surgery is individually tailored.

Subgraphs of functional brain networks identify dynamical constraints of cognitive control

Brain anatomy and physiology support the human ability to navigate a complex space of perceptions and actions. To maneuver across an ever-changing landscape of mental states, the brain invokes cognitive control-a set of dynamic processes that engage and disengage different groups of brain regions to modulate attention, switch between tasks, and inhibit prepotent responses. Current theory posits that correlated and anticorrelated brain activity may signify cooperative and competitive interactions between brain areas that subserve adaptive behavior. In this study, we use a quantitative approach to identify distinct topological motifs of functional interactions and examine how their expression relates to cognitive control processes and behavior. In particular, we acquire fMRI BOLD signal in twenty-eight healthy subjects as they perform two cognitive control tasks-a Stroop interference task and a local-global perception switching task using Navon figures-each with low and high cognitive control demand conditions. Based on these data, we construct dynamic functional brain networks and use a parts-based, network decomposition technique called non-negative matrix factorization to identify putative cognitive control subgraphs whose temporal expression captures distributed network structures involved in different phases of cooperative and competitive control processes. Our results demonstrate that temporal expression of the subgraphs fluctuate alongside changes in cognitive demand and are associated with individual differences in task performance. These findings offer insight into how coordinated changes in the cooperative and competitive roles of cognitive systems map trajectories between cognitively demanding brain states.

Applications of community detection techniques to brain graphs: Algorithmic considerations and implications for neural function

The human brain can be represented as a graph in which neural units such as cells or small volumes of tissue are heterogeneously connected to one another through structural or functional links. Brain graphs are parsimonious representations of neural systems that have begun to offer fundamental insights into healthy human cognition, as well as its alteration in disease. A critical open question in network neuroscience lies in how neural units cluster into densely interconnected groups that can provide the coordinated activity that is characteristic of perception, action, and adaptive behaviors. Tools that have proven particularly useful for addressing this question are community detection approaches, which can identify communities or modules: groups of neural units that are densely interconnected with other units in their own group but sparsely interconnected with units in other groups. In this paper, we describe a common community detection algorithm known as modularity maximization, and we detail its applications to brain graphs constructed from neuroimaging data. We pay particular attention to important algorithmic considerations, especially in recent extensions of these techniques to graphs that evolve in time. After recounting a few fundamental insights that these techniques have provided into brain function, we highlight potential avenues of methodological advancements for future studies seeking to better characterize the patterns of coordinated activity in the brain that accompany human behavior. This tutorial provides a naive reader with an introduction to theoretical considerations pertinent to the generation of brain graphs, an understanding of modularity maximization for community detection, a resource of statistical measures that can be used to characterize community structure, and an appreciation of the usefulness of these approaches in uncovering behaviorally-relevant network dynamics in neuroimaging data.

Degenerate time-dependent network dynamics anticipate seizures in human epileptic brain

Epileptic seizures are known to follow specific changes in brain dynamics. While some algorithms can nowadays robustly detect these changes, a clear understanding of the mechanism by which these alterations occur and generate seizures is still lacking. Here, we provide crossvalidated evidence that such changes are initiated by an alteration of physiological network state dynamics. Specifically, our analysis of long intracranial electroencephalography (iEEG) recordings from a group of 10 patients identifies a critical phase of a few hours in which time-dependent network states become less variable ("degenerate"), and this phase is followed by a global functional connectivity reduction before seizure onset. This critical phase is characterized by an abnormal occurrence of highly correlated network instances and is shown to be particularly associated with the activity of the resected regions in patients with validated postsurgical outcome. Our approach characterizes preseizure network dynamics as a cascade of 2 sequential events providing new insights into seizure prediction and control.

Understanding and predicting the generation of seizures in epileptic patients is fundamental to improving the quality of life of the more than 1% of the world population who suffer from this illness. Although seizure prediction has made important advances over the last decade, there is a lack of understanding of the common principles explaining the transitions that brain activity undergoes before a seizure. In this study, we characterized this transition from a novel perspective grounded on the mathematical analysis of continuous recordings inside the brains of epileptic patients over several days using depth electrodes. We show that the critical period preceding a seizure unfolds in a two-stage process. It begins with a phase of several hours when the highly correlated activity in the preceding days is altered, and it proceeds with a second, shorter phase of decrease in global connectivity before the seizure onset. Furthermore, our analysis reveals that these global alterations are more locally manifested in areas that are selected for surgical treatment. Our study suggests that preseizure activity might follow global stereotyped dynamics that could be targeted and modulated to prevent epileptic seizures.

Beyond modularity: Fine-scale mechanisms and rules for brain network reconfiguration

The human brain is in constant flux, as distinct areas engage in transient communication to support basic behaviors as well as complex cognition. The collection of interactions between cortical and subcortical areas forms a functional brain network whose topology evolves with time. Despite the nontrivial dynamics that are germane to this networked system, experimental evidence demonstrates that functional interactions organize into putative brain systems that facilitate different facets of cognitive computation. We hypothesize that such dynamic functional networks are organized around a set of rules that constrain their spatial architecture - which brain regions may functionally interact - and their temporal architecture - how these interactions fluctuate over time. To objectively uncover these organizing principles, we apply an unsupervised machine learning approach called non-negative matrix factorization to time-evolving, resting state functional networks in 20 healthy subjects. This machine learning approach automatically groups temporally co-varying functional interactions into subgraphs that represent putative topological modes of dynamic functional architecture. We find that subgraphs are stratified based on both the underlying modular organization and the topographical distance of their strongest interactions: while many subgraphs are largely contained within modules, others span between modules and are expressed differently over time. The relationship between dynamic subgraphs and modular architecture is further highlighted by the ability of time-varying subgraph expression to explain inter-individual differences in module reorganization. Collectively, these results point to the critical role that subgraphs play in constraining the topography and topology of functional brain networks. More broadly, this machine learning approach opens a new door for understanding the architecture of dynamic functional networks during both task and rest states, and for probing alterations of that architecture in disease.

Just a thought: How mind-wandering is represented in dynamic brain connectivity

The neuroscience of mind-wandering has begun to flourish, with roles of brain regions and networks being defined for various components of spontaneous thought. However, most of brain activity does not represent immediately occurring thoughts. Instead, spontaneous, organized network activity largely reflects "intrinsic" functions that are unrelated to the current experience. There remains no consensus on how brain networks represent mind-wandering in parallel to functioning in other ongoing, predominantly unconscious processes. Commonly, in network analysis of functional neuroimaging data, functional connectivity (FC; correlated time series) between remote brain regions is considered over several minutes or longer. In contrast, dynamic functional connectivity (dFC) is a new, promising approach to characterizing spontaneous changes in neural network communication on the faster time-scale at which intra-individual fluctuations in thought contents may occur. Here I describe how a potential relationship between mind-wandering and FC has traditionally been considered in the literature, and I review methods and results pertaining to the study of the dFC-mind-wandering relationship. While acknowledging challenges to the dFC approach and to behaviorally capturing fluctuations in inner experiences, I describe a framework for describing spontaneous thoughts in terms of brain-network activity patterns that are comprised of connections weighted by time-varying relevance to conscious and unconscious processing. This perspective suggests preferential roles of certain anatomical communication avenues (e.g., via the default mode network) in mind-wandering, while also implying that a region's connectivity fluctuates over time in its immediate degree of relevance to conscious contents, ultimately allowing novelty and diversity of thought.

The energy landscape underpinning module dynamics in the human brain connectome

Human brain dynamics can be viewed through the lens of statistical mechanics, where neurophysiological activity evolves around and between local attractors representing mental states. Many physically-inspired models of these dynamics define brain states based on instantaneous measurements of regional activity. Yet, recent work in network neuroscience has provided evidence that the brain might also be well-characterized by time-varying states composed of locally coherent activity or functional modules. We study this network-based notion of brain state to understand how functional modules dynamically interact with one another to perform cognitive functions. We estimate the functional relationships between regions of interest (ROIs) by fitting a pair-wise maximum entropy model to each ROI's pattern of allegiance to functional modules. This process uses an information theoretic notion of energy (as opposed to a metabolic one) to produce an energy landscape in which local minima represent attractor states characterized by specific patterns of modular structure. The clustering of local minima highlights three classes of ROIs with similar patterns of allegiance to community states. Visual, attention, sensorimotor, and subcortical ROIs are well-characterized by a single functional community. The remaining ROIs affiliate with a putative executive control community or a putative default mode and salience community. We simulate the brain's dynamic transitions between these community states using a random walk process. We observe that simulated transition probabilities between basins are statistically consistent with empirically observed transitions in resting state fMRI data. These results offer a view of the brain as a dynamical system that transitions between basins of attraction characterized by coherent activity in groups of brain regions, and that the strength of these attractors depends on the ongoing cognitive computations.