Background EEG Connectivity Captures the Time-Course of Epileptogenesis in a Mouse Model of Epilepsy

Involvement of the contralateral hippocampus in ictal-like but not interictal epileptic activities in the kainate mouse model of temporal lobe epilepsy

OBJECTIVE

Animal and human studies have shown that the seizure-generating region is vastly dependent on distant neuronal hubs that can decrease duration and propagation of ongoing seizures. However, we still lack a comprehensive understanding of the impact of distant brain areas on specific interictal and ictal epileptic activities (e.g., isolated spikes, spike trains, seizures). Such knowledge is critically needed, because all kinds of epileptic activities are not equivalent in terms of clinical expression and impact on the progression of the disease.

METHODS

We used surface high-density electroencephalography and multisite intracortical recordings, combined with pharmacological silencing of specific brain regions in the well-known kainate mouse model of temporal lobe epilepsy. We tested the impact of selective regional silencing on the generation of epileptic activities within a continuum ranging from very transient to more sustained and long-lasting discharges reminiscent of seizures.

RESULTS

Silencing the contralateral hippocampus completely suppresses sustained ictal activities in the focus, as efficiently as silencing the focus itself, but whereas focus silencing abolishes all focus activities, contralateral silencing fails to control transient spikes. In parallel, we observed that sustained focus epileptiform discharges in the focus are preceded by contralateral firing and more strongly phase-locked to bihippocampal delta/theta oscillations than transient spiking activities, reinforcing the presumed dominant role of the contralateral hippocampus in promoting long-lasting, but not transient, epileptic activities.

SIGNIFICANCE

Altogether, our work provides suggestive evidence that the contralateral hippocampus is necessary for the interictal to ictal state transition and proposes that crosstalk between contralateral neuronal activity and ipsilateral delta/theta oscillation could be a candidate mechanism underlying the progression from short- to long-lasting epileptic activities.

Visually or auditorily induced seizures involve the activation of nonhippocampal brain areas and hippocampal removal does not alleviate seizures in a mouse model of temporal lobe epilepsy

OBJECTIVE

Several studies have attributed epileptic activities in temporal lobe epilepsy (TLE) to the hippocampus; however, the participation of nonhippocampal neuronal networks in the development of TLE is often neglected. Here, we sought to understand how these nonhippocampal networks are involved in the pathology that is associated with TLE disease.

METHODS

A kainic acid (KA) model of temporal lobe epilepsy was induced by injecting KA into dorsal hippocampus of C57BL/6J mice. Network activation after spontaneous seizure was assessed using c-Fos expression. Protocols to induce seizure using visual or auditory stimulation were developed, and seizure onset zone (SOZ) and frequency of epileptic spikes were evaluated using electrophysiology. The hippocampus was removed to assess seizure recurrence in the absence of hippocampus.

RESULTS

Our results showed that cortical and hippocampal epileptic networks are activated during spontaneous seizures. Perturbation of these networks using visual or auditory stimulation readily precipitates seizures in TLE mice; the frequency of the light-induced or noise-induced seizures depends on the induction modality adopted during the induction period. Localization of SOZ revealed the existence of cortical and hippocampal SOZ in light-induced and noise-induced seizures, and the development of local and remote epileptic spikes in TLE occurs during the early stage of the disease. Importantly, we further discovered that removal of the hippocampi does not stop seizure activities in TLE mice, revealing that seizures in TLE mice can occur independent of the hippocampus.

SIGNIFICANCE

This study has shown that the network pathology that evolves in TLE is not localized to the hippocampus; rather, remote brain areas are also recruited. The occurrence of light-induced or noise-induced seizures and epileptic discharges in epileptic mice is a consequence of the activation of nonhippocampal brain areas. This work therefore demonstrates the fundamental role of nonhippocampal epileptic networks in generating epileptic activities with or without the hippocampus in TLE disease.

Reproducible network changes occur in a mouse model of temporal lobe epilepsy but do not correlate with disease severity

Studying the development of brain network disruptions in epilepsy is challenged by the paucity of data before epilepsy onset. Here, we used the unilateral, kainate mouse model of hippocampal epilepsy to investigate brain network changes before and after epilepsy onset and their stability across time. Using 32 epicranial electrodes distributed over the mouse hemispheres, we analyzed EEG epochs free from epileptic activity in 15 animals before and 28 days after hippocampal injection (group 1) and in 20 animals on two consecutive days (d28 and d29, group 2). Statistical dependencies between electrodes were characterized with the debiased-weighted phase lag index. We analyzed: a) graph metric changes from baseline to chronic stage (d28) in group 1; b) their reliability across d28 and d29, in group 2; c) their correlation with epileptic activity (EA: seizure, spike and fast-ripple rates), averaged over d28 and d29, in group 2. During the chronic stage, intra-hemispheric connections of the non-injected hemisphere strengthened, yielding an asymmetrical network in low (4-8 Hz) and high theta (8-12 Hz) bands. The contralateral hemisphere also became more integrated and segregated within the high theta band. Both network topology and EEG markers of EA were stable over consecutive days but not correlated with each other. Altogether, we show reproducible large-scale network modifications after the development of focal epilepsy. These modifications are mostly specific to the non-injected hemisphere. The absence of correlation with epileptic activity does not allow to specifically ascribe these network changes to mechanisms supporting EA or rather compensatory inhibition but supports the notion that epilepsy extends beyond the sole repetition of EA and impacts network that might not be involved in EA generation.

Seizure classification with selected frequency bands and EEG montages: a Natural Language Processing approach

Individualized treatment is crucial for epileptic patients with different types of seizures. The differences among patients impact the drug choice as well as the surgery procedure. With the advance in machine learning, automatic seizure detection can ease the manual time-consuming and labor-intensive procedure for diagnose seizure in the clinical setting. In this paper, we present an electroencephalography (EEG) frequency bands (sub-bands) and montages selection (sub-zones) method for classifier training that exploits Natural Language Processing from individual patients' clinical report. The proposed approach is targeting for individualized treatment. We integrated the prior knowledge from patient's reports into the classifier-building process, mimicking the authentic thinking process of experienced neurologist's when diagnosing seizure using EEG. The keywords from clinical documents are mapped to the EEG data in terms of frequency bands and scalp EEG electrodes. The data of experiments are from the Temple University Hospital EEG seizure corpus, and the dataset is divided based on each group of patients with same seizure type and same recording electrode references. The classifier includes Random Forest, Support Vector Machine and Multi-Layer Perceptron. The classification performance indicates that competitive results can be achieve with a small portion of EEG the data. Using the sub-zones selection for Generalized Seizures (GNSZ) on all three electrodes, data are reduced by nearly 50% while the performance metrics remain at the same level with the whole frequency and zones. Moreover, when selecting by sub-zones and sub-bands together for GNSZ with Linked Ears reference, the data range reduced to 0.3% of whole range, and the performance deviates less than 3% from the results with whole range of data. Results show that using proposed approach may lead to more efficient implementations of the seizure classifier to be executed on power-efficient devices for long lasting real-time seizures detection.

Emerging imaging methods to study whole-brain function in rodent models

In the past decade, the idea that single populations of neurons support cognition and behavior has gradually given way to the realization that connectivity matters and that complex behavior results from interactions between remote yet anatomically connected areas that form specialized networks. In parallel, innovation in brain imaging techniques has led to the availability of a broad set of imaging tools to characterize the functional organization of complex networks. However, each of these tools poses significant technical challenges and faces limitations, which require careful consideration of their underlying anatomical, physiological, and physical specificity. In this review, we focus on emerging methods for measuring spontaneous or evoked activity in the brain. We discuss methods that can measure large-scale brain activity (directly or indirectly) with a relatively high temporal resolution, from milliseconds to seconds. We further focus on methods designed for studying the mammalian brain in preclinical models, specifically in mice and rats. This field has seen a great deal of innovation in recent years, facilitated by concomitant innovation in gene-editing techniques and the possibility of more invasive recordings. This review aims to give an overview of currently available preclinical imaging methods and an outlook on future developments. This information is suitable for educational purposes and for assisting scientists in choosing the appropriate method for their own research question.

Brain injuries can set up an epileptogenic neuronal network

Development of epilepsy or epileptogenesis promotes recurrent seizures. As of today, there are no effective prophylactic therapies to prevent the onset of epilepsy. Contributing to this deficiency of preventive therapy is the lack of clarity in fundamental neurobiological mechanisms underlying epileptogenesis and lack of reliable biomarkers to identify patients at risk for developing epilepsy. This limits the development of prophylactic therapies in epilepsy. Here, neural network dysfunctions reflected by oscillopathies and microepileptiform activities, including neuronal hyperexcitability and hypersynchrony, drawn from both clinical and experimental epilepsy models, have been reviewed. This review suggests that epileptogenesis reflects a progressive and dynamic dysfunction of specific neuronal networks which recruit further interconnected groups of neurons, with this resultant pathological network mediating seizure occurrence, recurrence, and progression. In the future, combining spatial and temporal resolution of neuronal non-invasive recordings from patients at risk of developing epilepsy, together with analytics and computational tools, may contribute to determining whether the brain is undergoing epileptogenesis in asymptomatic patients following brain injury.

The Kainic Acid Models of Temporal Lobe Epilepsy

Experimental models of epilepsy are useful to identify potential mechanisms of epileptogenesis, seizure genesis, comorbidities, and treatment efficacy. The kainic acid (KA) model is one of the most commonly used. Several modes of administration of KA exist, each producing different effects in a strain-, species-, gender-, and age-dependent manner. In this review, we discuss the advantages and limitations of the various forms of KA administration (systemic, intrahippocampal, and intranasal), as well as the histologic, electrophysiological, and behavioral outcomes in different strains and species. We attempt a personal perspective and discuss areas where work is needed. The diversity of KA models and their outcomes offers researchers a rich palette of phenotypes, which may be relevant to specific traits found in patients with temporal lobe epilepsy.

Rat Locomotion Detection Based on Brain Functional Directed Connectivity from Implanted Electroencephalography Signals

Previous findings have suggested that the cortex involved in walking control in freely locomotion rats. Moreover, the spectral characteristics of cortical activity showed significant differences in different walking conditions. However, whether brain connectivity presents a significant difference during rats walking under different behavior conditions has yet to be verified. Similarly, whether brain connectivity can be used in locomotion detection remains unknown. To address those concerns, we recorded locomotion and implanted electroencephalography signals in freely moving rats performing two kinds of task conditions (upslope and downslope walking). The Granger causality method was used to determine brain functional directed connectivity in rats during these processes. Machine learning algorithms were then used to categorize the two walking states, based on functional directed connectivity. We found significant differences in brain functional directed connectivity varied between upslope and downslope walking. Moreover, locomotion detection based on brain connectivity achieved the highest accuracy (91.45%), sensitivity (90.93%), specificity (91.3%), and F1-score (91.43%). Specifically, the classification results indicated that connectivity features in the high gamma band contained the most discriminative information with respect to locomotion detection in rats, with the support vector machine classifier exhibiting the most efficient performance. Our study not only suggests that brain functional directed connectivity in rats showed significant differences in various behavioral contexts but also proposed a method for classifying the locomotion states of rat walking, based on brain functional directed connectivity. These findings elucidate the characteristics of neural information interaction between various cortical areas in freely walking rats.

Application of a flexible polymer microECoG array to map functional coherence in schizophrenia model

Anatomically, connections form the fundamental brain network, functionally the different types of oscillatory electric activities are creating a temporarily connected fraction of the anatomical connectome generating an output to the motor system. Schizophrenia can be considered as a connectome disease, in which the sensory input generates a schizophrenia specific temporary connectome and the signal processing becomes diseased showing hallucinations and adverse behavioral reactions. In this work, flexible, 32-channel polymer microelectrode arrays fabricated by the authors are used to map the functional coherence on large cortical areas during physiological activities in a schizophrenia model in rats.-Fabrication of a flexible microECoG array is shown.-Protocol to use a flexible microECoG is demonstrated to characterize connectome diseases in rats.-Customized method to analyze the functional coherence between different cortical areas during visually evoked potential is detailed.-R-based implementation of the analysis method is presented.

Understanding the dynamics of biological and neural oscillator networks through exact mean-field reductions: a review

Many biological and neural systems can be seen as networks of interacting periodic processes. Importantly, their functionality, i.e., whether these networks can perform their function or not, depends on the emerging collective dynamics of the network. Synchrony of oscillations is one of the most prominent examples of such collective behavior and has been associated both with function and dysfunction. Understanding how network structure and interactions, as well as the microscopic properties of individual units, shape the emerging collective dynamics is critical to find factors that lead to malfunction. However, many biological systems such as the brain consist of a large number of dynamical units. Hence, their analysis has either relied on simplified heuristic models on a coarse scale, or the analysis comes at a huge computational cost. Here we review recently introduced approaches, known as the Ott-Antonsen and Watanabe-Strogatz reductions, allowing one to simplify the analysis by bridging small and large scales. Thus, reduced model equations are obtained that exactly describe the collective dynamics for each subpopulation in the oscillator network via few collective variables only. The resulting equations are next-generation models: Rather than being heuristic, they exactly link microscopic and macroscopic descriptions and therefore accurately capture microscopic properties of the underlying system. At the same time, they are sufficiently simple to analyze without great computational effort. In the last decade, these reduction methods have become instrumental in understanding how network structure and interactions shape the collective dynamics and the emergence of synchrony. We review this progress based on concrete examples and outline possible limitations. Finally, we discuss how linking the reduced models with experimental data can guide the way towards the development of new treatment approaches, for example, for neurological disease.

Controversies on the network theory of epilepsy: Debates held during the ICTALS 2019 conference

Debates on six controversial topics on the network theory of epilepsy were held during two debate sessions, as part of the International Conference for Technology and Analysis of Seizures, 2019 (ICTALS 2019) convened at the University of Exeter, UK, September 2-5 2019. The debate topics were (1) From pathologic to physiologic: is the epileptic network part of an existing large-scale brain network? (2) Are micro scale recordings pertinent for defining the epileptic network? (3) From seconds to years: do we need all temporal scales to define an epileptic network? (4) Is it necessary to fully define the epileptic network to control it? (5) Is controlling seizures sufficient to control the epileptic network? (6) Does the epileptic network want to be controlled? This article, written by the organizing committee for the debate sessions and the debaters, summarizes the arguments presented during the debates on these six topics.