Seizure prediction and its applications

Sex-specific differences in mortality and neurocardiac interactions in the Kv1.1 knockout mouse model of sudden unexpected death in epilepsy (SUDEP)

Sudden unexpected death in epilepsy (SUDEP) is a devastating complication of epilepsy with possible sex-specific risk factors, although the exact relationship between sex and SUDEP remains unclear. To investigate this, we studied Kcna1 knockout (Kcna1-/-) mice, which lack voltage-gated Kv1.1 channel subunits and are widely used as a SUDEP model that mirrors key features in humans. To assess sex differences, we first performed survival analysis, EEG-ECG recordings, seizure threshold testing and retrospective analysis of previous intracardiac pacing data. We then applied a novel modelling approach across organs (organomics) to uncover potential sex-specific differences in brain-heart communication. Our findings revealed female Kcna1-/- mice have significantly longer lifespans than males, suggesting lower SUDEP rates. Although no sex differences were found in seizure frequency, duration, burden, susceptibility or interictal heart rate variability, females showed a higher incidence of bradycardia during spontaneous seizures than males, as well as resistance to inducible ventricular tachyarrhythmias in response to programmed electrical stimulation. Two captured SUDEP events, one per sex, displayed similar patterns of ictal bradycardia in both sexes, progressing to postictal cardiorespiratory failure. Going beyond traditional seizure and cardiac metrics, organomics analysis revealed that seizures affect brain-heart communication differently between sexes. Females exhibited more effective resetting of brain-heart interactions postictally than males. This finding may contribute to the lower SUDEP risk in females and underscores the complex interplay between sex, cardiac function and brain-heart communication in determining SUDEP susceptibility. Furthermore, seizure-resetting measures could represent a promising class of biomarkers for SUDEP risk stratification. KEY POINTS: Female Kcna1-/- mice live longer than males, suggesting lower sudden unexpected death in epilepsy (SUDEP) rates. There are no sex differences in seizure metrics or interictal heart rate variability. Females show more bradycardia during seizures and are resistant to inducible ventricular tachyarrhythmias. Seizures affect brain-heart communication differently between the sexes. Seizures in females reset brain-heart interactions more effectively postictally, potentially lowering SUDEP risk.

Machine Learning in Neurosurgery: Toward Complex Inputs, Actionable Predictions, and Generalizable Translations

Machine learning can predict neurosurgical diagnosis and outcomes, power imaging analysis, and perform robotic navigation and tumor labeling. State-of-the-art models can reconstruct and generate images, predict surgical events from video, and assist in intraoperative decision-making. In this review, we will detail the neurosurgical applications of machine learning, ranging from simple to advanced models, and their potential to transform patient care. As machine learning techniques, outputs, and methods become increasingly complex, their performance is often more impactful yet increasingly difficult to evaluate. We aim to introduce these advancements to the neurosurgical audience while suggesting major potential roadblocks to their safe and effective translation. Unlike the previous generation of machine learning in neurosurgery, the safe translation of recent advancements will be contingent on neurosurgeons' involvement in model development and validation.

Functional connectivity discriminates epileptogenic states and predicts surgical outcome in children with drug resistant epilepsy

Normal brain functioning emerges from a complex interplay among regions forming networks. In epilepsy, these networks are disrupted causing seizures. Highly connected nodes in these networks are epilepsy surgery targets. Here, we assess whether functional connectivity (FC) using intracranial electroencephalography can quantify brain regions epileptogenicity and predict surgical outcome in children with drug resistant epilepsy (DRE). We computed FC between electrodes on different states (i.e. interictal without spikes, interictal with spikes, pre-ictal, ictal, and post-ictal) and frequency bands. We then estimated the electrodes' nodal strength. We compared nodal strength between states, inside and outside resection for good- (n = 22, Engel I) and poor-outcome (n = 9, Engel II-IV) patients, respectively, and tested their utility to predict the epileptogenic zone and outcome. We observed a hierarchical epileptogenic organization among states for nodal strength: lower FC during interictal and pre-ictal states followed by higher FC during ictal and post-ictal states (p < 0.05). We further observed higher FC inside resection (p < 0.05) for good-outcome patients on different states and bands, and no differences for poor-outcome patients. Resection of nodes with high FC was predictive of outcome (positive and negative predictive values: 47-100%). Our findings suggest that FC can discriminate epileptogenic states and predict outcome in patients with DRE.

An automated detection of epileptic seizures EEG using CNN classifier based on feature fusion with high accuracy

BACKGROUND

Epilepsy is a neurological disorder that is usually detected by electroencephalogram (EEG) signals. Since manual examination of epilepsy seizures is a laborious and time-consuming process, lots of automatic epilepsy detection algorithms have been proposed. However, most of the available classification algorithms for epilepsy EEG signals adopted a single feature extraction, in turn to result in low classification accuracy. Although a small account of studies have carried out feature fusion, the computational efficiency is reduced due to too many features, because there are also some poor features that interfere with the classification results.

METHODS

In order to solve the above problems, an automatic recognition method of epilepsy EEG signals based on feature fusion and selection is proposed in this paper. Firstly, the Approximate Entropy (ApEn), Fuzzy Entropy (FuzzyEn), Sample Entropy (SampEn), and Standard Deviation (STD) mixed features of the subband obtained by the Discrete Wavelet Transform (DWT) decomposition of EEG signals are extracted. Secondly, the random forest algorithm is used for feature selection. Finally, the Convolutional Neural Network (CNN) is used to classify epilepsy EEG signals.

RESULTS

The empirical evaluation of the presented algorithm is performed on the benchmark Bonn EEG datasets and New Delhi datasets. In the interictal and ictal classification tasks of Bonn datasets, the proposed model achieves an accuracy of 99.9%, a sensitivity of 100%, a precision of 99.81%, and a specificity of 99.8%. For the interictal-ictal case of New Delhi datasets, the proposed model achieves a classification accuracy of 100%, a sensitivity of 100%, a specificity of 100%, and a precision of 100%.

CONCLUSION

The proposed model can effectively realize the high-precision automatic detection and classification of epilepsy EEG signals. This model can provide high-precision automatic detection capability for clinical epilepsy EEG detection. We hope to provide positive implications for the prediction of seizure EEG.

Toward automated prediction of sudden unexpected death in epilepsy

Sudden unexpected death in epilepsy (SUDEP) is a devastating yet overlooked complication of epilepsy. The rare and complex nature of SUDEP makes it challenging to study. No prediction or prevention of SUDEP is currently available in a clinical setting. In the past decade, significant advances have been made in our knowledge of the pathophysiologic cascades that lead to SUDEP. In particular, studies of brain, heart, and respiratory functions in both human patients at the epilepsy monitoring unit and animal models during fatal seizures provide critical information to integrate computational tools for SUDEP prediction. The rapid advances in automated seizure detection and prediction algorithms provide a fundamental framework for their adaption in predicting SUDEP. If a SUDEP can be predicted, then there will be a potential for medical intervention to be administered, either by their caregivers or via an implanted device automatically delivering electrical stimulation or medication, and finally save lives from fatal seizures. This article presents recent developments of SUDEP studies focusing on the pathophysiologic basis of SUDEP and computational implications of machine learning techniques that can be adapted and extended for SUDEP prediction. This article also discusses some novel ideas for SUDEP prediction and rescue including principal component analysis and closed-loop intervention.

The performance evaluation of the state-of-the-art EEG-based seizure prediction models

The recurrent and unpredictable nature of seizures can lead to unintentional injuries and even death. The rapid development of electroencephalogram (EEG) and Artificial Intelligence (AI) technologies has made it possible to predict seizures in real-time through brain-machine interfaces (BCI), allowing advanced intervention. To date, there is still much room for improvement in predictive seizure models constructed by EEG using machine learning (ML) and deep learning (DL). But, the most critical issue is how to improve the performance and generalization of the model, which involves some confusing conceptual and methodological issues. This review focuses on analyzing several factors affecting the performance of seizure prediction models, focusing on the aspects of post-processing, seizure occurrence period (SOP), seizure prediction horizon (SPH), and algorithms. Furthermore, this study presents some new directions and suggestions for building high-performance prediction models in the future. We aimed to clarify the concept for future research in related fields and improve the performance of prediction models to provide a theoretical basis for future applications of wearable seizure detection devices.

Central histaminergic signalling, neural excitability and epilepsy

Epilepsy is a common neurological disorder characterized by repeated and spontaneous epileptic seizures and is not well controlled by current medication. Traditional theory suggests that epilepsy results from an imbalance of excitatory glutamate neurons and inhibitory GABAergic neurons. However, new evidence from clinical and preclinical research suggests that histamine in the CNS plays an important role in the modulation of neural excitability and in the pathogenesis of epilepsy. Many histamine receptor ligands have achieved curative effects in animal epilepsy models, among which the histamine H₃ receptor antagonist pitolisant has shown anti-epileptic effects in clinical trials. Recent studies, therefore, have focused on the potential action of histamine receptors to control and treat epilepsy. In this review, we summarize the findings from animal and clinical epilepsy research on the role of brain histamine and its receptors. We also identify current gaps in the research and suggest where further studies are most needed.

Control of epileptic seizures by electrical stimulation: a model-based study

High frequency electrical stimulation of brain is commonly used in research experiments and clinical trials as a modern tool for control of epileptic seizures. However, the mechanistic basis by which periodic external stimuli alter the brain state is not well understood. This study provides a computational insight into the mechanism of seizure suppression by high frequency stimulation (HFS). In particular, a modified version of the Jansen-Rit neural mass model is employed, in which EEG signals can be considered as the input. The proposed model reproduces seizure-like activity in the output during the ictal period of the input signal. By applying a control signal to the model, a wide range of stimulation amplitudes and frequencies are systematically explored. Simulation results reveal that HFS can effectively suppress the seizure-like activity. Our results suggest that HFS has the ability of shifting the operating state of neural populations away from a critical condition. Furthermore, a closed-loop control strategy is proposed in this paper. The main objective has been to considerably reduce the control effort needed for blocking abnormal activity of the brain. Such an energy reduction could be of practical importance, to reduce possible side effects and increase battery life for implanted neurostimulators.

Identifying Refractory Epilepsy Without Structural Abnormalities by Fusing the Common Spatial Patterns of Functional and Effective EEG Networks

Drug refractory epilepsy (RE) is believed to be associated with structural lesions, but some RE patients show no significant structural abnormalities (RE-no-SA) on conventional magnetic resonance imaging scans. Since most of the medically controlled epilepsy (MCE) patients also do not exhibit structural abnormalities, a reliable assessment needs to be developed to differentiate RE-no-SA patients and MCE patients to avoid misdiagnosis and inappropriate treatment. Using resting-state scalp electroencephalogram (EEG) datasets, we extracted the spatial pattern of network (SPN) features from the functional and effective EEG networks of both RE-no-SA patients and MCE patients. Compared to the performance of traditional resting-state EEG network properties, the SPN features exhibited remarkable superiority in classifying these two groups of epilepsy patients, and accuracy values of 90.00% and 80.00% were obtained for the SPN features of the functional and effective EEG networks, respectively. By further fusing the SPN features of functional and effective networks, we demonstrated that the highest accuracy value of 96.67% could be reached, with a sensitivity of 100% and specificity of 92.86%. Overall, these findings not only indicate that the fused functional and effective SPN features are promising as reliable measurements for distinguishing RE-no-SA patients and MCE patients but also may provide a new perspective to explore the complex neurophysiology of refractory epilepsy.

Machine learning algorithms to predict seizure due to acute tramadol poisoning

INTRODUCTION

This study was designed to develop and evaluate machine learning algorithms for predicting seizure due to acute tramadol poisoning, identifying high-risk patients and facilitating appropriate clinical decision-making.

METHODS

Several characteristics of acute tramadol poisoning cases were collected in the Emergency Department (ED) (2013-2019). After selecting important variables in random forest method, prediction models were developed using the Support Vector Machine (SVM), Naïve Bayes (NB), Artificial Neural Network (ANN) and K-Nearest Neighbor (K-NN) algorithms. Area Under the Curve (AUC) and other diagnostic criteria were used to assess performance of models.

RESULTS

In 909 patients, 544 (59.8%) experienced seizures. The important predictors of seizure were sex, pulse rate, arterial blood oxygen pressure, blood bicarbonate level and pH. SVM (AUC = 0.68), NB (AUC = 0.71) and ANN (AUC = 0.70) models outperformed k-NN model (AUC = 0.58). NB model had a higher sensitivity and negative predictive value and k-NN model had higher specificity and positive predictive values than other models.

CONCLUSION

A perfect prediction model may help improve clinicians' decision-making and clinical care at EDs in hospitals and medical settings. SVM, ANN and NB models had no significant differences in the performance and accuracy; however, validated logistic regression (LR) was the superior model for predicting seizure due to acute tramadol poisoning.

Directed Connectivity Analysis of the Neuro-Cardio- and Respiratory Systems Reveals Novel Biomarkers of Susceptibility to SUDEP

Goal: Sudden unexpected death in epilepsy (SUDEP) is the leading cause of epilepsy-related mortality and its pathophysiological mechanisms remain unknown. We set to record and analyze for the first time concurrent electroencephalographic (EEG), electrocardiographic (ECG), and unrestrained whole-body plethysmographic (Pleth) signals from control (WT - wild type) and SUDEP-prone mice (KO- knockout Kcna1 animal model). Employing multivariate autoregressive models (MVAR) we measured all tri-organ effective directional interactions by the generalized partial directed coherence (GPDC) in the frequency domain over time (hours). When compared to the control (WT) animals, the SUDEP-prone (KO) animals exhibited (p < 0.001) reduced afferent and efferent interactions between the heart and the brain over the full frequency spectrum (0-200Hz), enhanced efferent interactions from the brain to the lungs and from the heart to the lungs at high (>90 Hz) frequencies (especially during periods with seizure activity), and decreased feedback from the lungs to the brain at low (<40 Hz) frequencies. These results show that impairment in the afferent and efferent pathways in the holistic neuro-cardio-respiratory network could lead to SUDEP, and effective connectivity measures and their dynamics could serve as novel biomarkers of susceptibility to SUDEP and seizures respectively.

Prediction of epileptic seizures with convolutional neural networks and functional near-infrared spectroscopy signals

There have been different efforts to predict epileptic seizures and most of them are based on the analysis of electroencephalography (EEG) signals; however, recent publications have suggested that functional Near-Infrared Spectroscopy (fNIRS), a relatively new technique, could be used to predict seizures. The objectives of this research are to show that the application of fNIRS to epileptic seizure detection yields results that are superior to those based on EEG and to demonstrate that the application of deep learning to this problem is suitable given the nature of fNIRS recordings. A Convolutional Neural Network (CNN) is applied to the prediction of epileptic seizures from fNIRS signals, an optical modality for recording brain waves. The implementation of the proposed method is presented in this work. Application of CNN to fNIRS recordings showed an accuracy ranging between 96.9% and 100%, sensitivity between 95.24% and 100%, specificity between 98.57% and 100%, a positive predictive value between 98.52% and 100%, and a negative predictive value between 95.39% and 100%. The most important aspect of this research is the combination of fNIRS signals with the particular CNN algorithm. The fNIRS modality has not been used in epileptic seizure prediction. A CNN is suitable for this application because fNIRS recordings are high dimensional data and they can be modeled as three-dimensional tensors for classification.

Probabilistic prediction of Epileptic Seizures using SVM

In this paper, an algorithm based on the linear Support Vector Machine (SVM) tool was proposed to classify intracranial electroencephalography (iEEG) signals as ictal or interictal to perform human seizure prediction, efficiently. Various univariate linear measures were extracted, and the developed classifier performed adequately well with numerous performance metrics, especially the dataset was suffering from a significant imbalanced class, with the majority of samples representing non-seizure events. The proposed tool was indeed able to forecast accurately such rare events, seizures, from a large set of EEG dataset. In fact, our model can predict some seizures with up to 0.4 probability and about 30-40 minutes in advance. The proposed work employed intracranial EEG recordings of 6 patients in the Freiburg EEG database, totalling trained and tested on 34 seizures of 140-hour-long. It exhibits a sensitivity of 78% and specificity of 100% employing a 2-second-long window with 10-fold cross-validation.

Connectivity and Centrality Characteristics of the Epileptogenic Focus Using Directed Network Analysis

Accurate epileptogenic focus localization is required prior to surgical resection of brain tissue for the treatment of patients with antiepileptic drug-resistant (intractable) epilepsy. This clinical need is only partially fulfilled through a subjective, and at times inconclusive, the evaluation of the recorded electroencephalogram (EEG) at seizures' onset (the so-called gold standard for focus localization in epilepsy). We herein present a novel method of multivariate analysis of the EEG that appears to be very promising for an objective and robust localization of the epileptogenic focus at seizures' onset. Using the measure of generalized partial directed coherence, combined with surrogate data analysis, we first estimated from multichannel intracranial EEG the statistically significant causal interactions between brain regions at the onset of 92 clinical seizures from nine patients with temporal lobe intractable epilepsy. From the networks that were formed based on the thus derived interactions, a set of centrality metrics was estimated per network node (brain site). Brain sites located anatomically within the epileptogenic focus were shown to be associated with greater inward centrality values than non-focal brain regions at high frequencies ( γ band), and particular inward centrality metrics accurately localized the focus in all nine patients. In addition to focus localization from seizure (ictal) onset, the developed novel framework for analysis of EEG could be employed to identify the changes of the focal network over time, peri-ictally and interictally, and thus shed light onto the dynamics of ictogenesis, which could then have a significant impact on automated prediction and closed-loop control of seizures by neuromodulation.

Multivariate Matching Pursuit Decomposition and Normalized Gabor Entropy for Quantification of Preictal Trends in Epilepsy

Quantification of the complexity of signals recorded concurrently from multivariate systems, such as the brain, plays an important role in the study and characterization of their state and state transitions. Multivariate analysis of the electroencephalographic signals (EEG) over time is conceptually most promising in unveiling the global dynamics of dynamical brain disorders such as epilepsy. We employed a novel methodology to study the global complexity of the epileptic brain en route to seizures. The developed measures of complexity were based on Multivariate Matching Pursuit (MMP) decomposition of signals in terms of time–frequency Gabor functions (atoms) and Shannon entropy. The measures were first validated on simulation data (Lorenz system) and then applied to EEGs from preictal (before seizure onsets) periods, recorded by intracranial electrodes from eight patients with temporal lobe epilepsy and a total of 42 seizures, in search of global trends of complexity before seizures onset. Out of five Gabor measures of complexity we tested, we found that our newly defined measure, the normalized Gabor entropy (NGE), was able to detect statistically significant (p < 0.05) nonlinear trends of the mean global complexity across all patients over 1 h periods prior to seizures’ onset. These trends pointed to a slow decrease of the epileptic brain’s global complexity over time accompanied by an increase of the variance of complexity closer to seizure onsets. These results show that the global complexity of the epileptic brain decreases at least 1 h prior to seizures and imply that the employed methodology and measures could be useful in identifying different brain states, monitoring of seizure susceptibility over time, and potentially in seizure prediction.

Seizure detection: do current devices work? And when can they be useful?

PURPOSE OF REVIEW

The unpredictability and apparent randomness of epileptic seizures is one of the most vexing aspects of epilepsy. Methods or devices capable of detecting seizures may help prevent injury or even death and significantly improve quality of life. Here, we summarize and evaluate currently available, unimodal, or polymodal detection systems for epileptic seizures, mainly in the ambulatory setting.

RECENT FINDINGS

There are two broad categories of detection devices: EEG-based and non-EEG-based systems. Wireless wearable EEG devices are now available both in research and commercial arenas. Neuro-stimulation devices are currently evolving and initial experiences of these show potential promise. As for non-EEG devices, different detecting systems show different sensitivity according to the different patient and seizure types. Regardless, when used in combination, these modalities may complement each other to increase positive predictive value. Although some devices with high sensitivity are promising, practical widespread use of such detection systems is still some way away. More research and experience are needed to evaluate the most efficient and integrated systems, to allow for better approaches to detection and prediction of seizures. The concept of closed-loop systems and prompt intervention may substantially improve quality of life for patients and carers.

Predictability and Resetting in a Case of Convulsive Status Epilepticus

In this case study, we present evidence of resetting of brain dynamics following convulsive status epilepticus (SE) that was treated successfully with antiepileptic medications (AEDs). The measure of effective inflow (EI), a novel measure of network connectivity, was applied to the continuously recorded multichannel intracranial stereoelectroencephalographic (SEEG) signals before, during and after SE. Results from this analysis indicate trends of progressive reduction of EI over hours up to the onset of SE, mainly at sites of the epileptogenic focus with reversal of those trends upon successful treatment of SE by AEDs. The proposed analytical framework is promising for elucidation of the pathology of neuronal network dynamics that could lead to SE, evaluation of the efficacy of SE treatment strategies, as well as the development of biomarkers for susceptibility to SE.

Scn2a deletion improves survival and brain-heart dynamics in the Kcna1-null mouse model of sudden unexpected death in epilepsy (SUDEP)

People with epilepsy have greatly increased probability of premature mortality due to sudden unexpected death in epilepsy (SUDEP). Identifying which patients are most at risk of SUDEP is hindered by a complex genetic etiology, incomplete understanding of the underlying pathophysiology and lack of prognostic biomarkers. Here we evaluated heterozygous Scn2a gene deletion (Scn2a+/-) as a protective genetic modifier in the Kcna1 knockout mouse (Kcna1-/-) model of SUDEP, while searching for biomarkers of SUDEP risk embedded in electroencephalography (EEG) and electrocardiography (ECG) recordings. The human epilepsy gene Kcna1 encodes voltage-gated Kv1.1 potassium channels that act to dampen neuronal excitability whereas Scn2a encodes voltage-gated Nav1.2 sodium channels important for action potential initiation and conduction. SUDEP-prone Kcna1-/- mice with partial genetic ablation of Nav1.2 channels (i.e. Scn2a+/-; Kcna1-/-) exhibited a two-fold increase in survival. Classical analysis of EEG and ECG recordings separately showed significantly decreased seizure durations in Scn2a+/-; Kcna1-/- mice compared with Kcna1-/- mice, without substantial modification of cardiac abnormalities. Novel analysis of the EEG and ECG together revealed a significant reduction in EEG-ECG association in Kcna1-/- mice compared with wild types, which was partially restored in Scn2a+/-; Kcna1-/- mice. The degree of EEG-ECG association was also proportional to the survival rate of mice across genotypes. These results show that Scn2a gene deletion acts as protective genetic modifier of SUDEP and suggest measures of brain-heart association as potential indices of SUDEP susceptibility.

Graph Theory at the Service of Electroencephalograms

The brain is one of the largest and most complex organs in the human body and EEG is a noninvasive electrophysiological monitoring method that is used to record the electrical activity of the brain. Lately, the functional connectivity in human brain has been regarded and studied as a complex network using EEG signals. This means that the brain is studied as a connected system where nodes, or units, represent different specialized brain regions and links, or connections, represent communication pathways between the nodes. Graph theory and theory of complex networks provide a variety of measures, methods, and tools that can be useful to efficiently model, analyze, and study EEG networks. This article is addressed to computer scientists who wish to be acquainted and deal with the study of EEG data and also to neuroscientists who would like to become familiar with graph theoretic approaches and tools to analyze EEG data.

Low Frequency Electrical Stimulation Either Prior to Or after Rapid Kindling Stimulation Inhibits the Kindling-Induced Epileptogenesis

Objective. Studies are ongoing to find appropriate low frequency stimulation (LFS) protocol for treatment of epilepsy. The present study aimed at assessing the antiepileptogenesis effects of LFS with the same protocol applied either just before or immediately after kindling stimulations. Method. This experimental animal study was conducted on adult Wistar rats (200 ± 20 g) randomly divided into kindle (n = 7), LFS + Kindle (n = 6), and Kindle + LFS groups (n = 6). All animals underwent rapid kindling procedure and four packages of LFS (1 Hz) with 5 min interval were applied either immediately before (LFS-K) or after kindling stimulation (K-LFS). The after discharge duration (ADD), daily stages of kindling, and kindling seizure stage and number of stimulations required to reach each stage were compared between the three groups using two-way analysis of variance (ANOVA) followed by Tukey post hoc and one-way ANOVA, and Kruskal-Wallis test, respectively. Results. LFS in both protocols significantly decreased the ADD (p < 0.05) and daily seizure stages (p < 0.05) and increased the number of stimulations required to achieve stage 3 and stages 4 and 5 of kindling compared with the kindle group (stage 2: p > 0.05, stages 3 to 5: p < 0.05). Conclusion. Although LFS-K showed more inhibiting effect than K-LFS, the difference was not statistically significant.

The Concept of Effective Inflow: Application to Interictal Localization of the Epileptogenic Focus From iEEG

GOAL

Accurate determination of the epileptogenic focus is of paramount diagnostic and therapeutic importance in epilepsy. The current gold standard for focus localization is from ictal (seizure) onset and thus requires the occurrence and recording of multiple typical seizures of a patient. Localization of the focus from seizure-free (interictal) periods remains a challenging problem, especially in the absence of interictal epileptiform activity.

METHODS

By exploring the concept of effective inflow, we developed a focus localization algorithm (FLA) based on directed connectivity between brain sites. Subsequently, using the measure of generalized partial directed coherence over a broad frequency band in FLA for the analysis of interictal periods from long-term (days) intracranial electroencephalographic signals, we identified the brain region that is the most frequent receiver of maximal effective inflow from other brain regions.

RESULTS

In six out of nine patients with temporal lobe epilepsy, the thus identified brain region was a statistically significant outlier (p < 0.01) and coincided with the clinically assessed epileptogenic focus. In the remaining three patients, the clinically assessed focus still exhibited the highest inflow, but it was not deemed an outlier (p > 0.01).

CONCLUSIONS

These findings suggest that the epileptogenic focus is a region of intense influence from other regions interictally, possibly as a mechanism to keep it under control in seizure-free periods.

SIGNIFICANCE

The developed framework is expected to assist with the accurate epileptogenic focus localization, reduce hospital stay and healthcare cost, and provide guidance to treatment of epilepsy via resective surgery or neuromodulation.

Suppression of seizures based on the multi-coupled neural mass model

Epilepsy is one of the most common serious neurological disorders, which affects approximately 1% of population in the world. In order to effectively control the seizures, we propose a novel control methodology, which combines the feedback linearization control (FLC) with the underlying mechanism of epilepsy, to achieve the suppression of seizures. The three coupled neural mass model is constructed to study the property of the electroencephalographs (EEGs). Meanwhile, with the model we research on the propagation of epileptiform waves and the synchronization of populations, which are taken as the foundation of our control method. Results show that the proposed approach not only yields excellent performances in clamping the pathological spiking patterns to the reference signals derived under the normal state but also achieves the normalization of the pathological parameter, where the parameters are estimated from EEGs with Unscented Kalman Filter. The specific contribution of this paper is to treat the epilepsy from its pathogenesis with the FLC, which provides critical theoretical basis for the clinical treatment of neurological disorders.

Jump neural network for real-time prediction of glucose concentration

Prediction of the future value of a variable is of central importance in a wide variety of fields, including economy and finance, meteorology, informatics, and, last but not least important, medicine. For example, in the therapy of Type 1 Diabetes (T1D), in which, for patient safety, glucose concentration in the blood should be maintained in a defined normoglycemic range, the ability to forecast glucose concentration in the short-term (with a prediction horizon of around 30 min) might be sufficient to reduce the incidence of hypoglycemic and hyperglycemic events. Neural Network (NN) approaches are suitable for prediction purposes because of their ability to model nonlinear dynamics and handle in their inputs signals coming from different domains. In this chapter we illustrate the design of a jump NN glucose prediction algorithm that exploits past glucose concentration data, measured in real-time by a minimally invasive continuous glucose monitoring (CGM) sensor, and information on ingested carbohydrates, supplied by the patient himself or herself. The methodology is assessed by tuning the NN on data of ten T1D individuals and then testing it on a dataset of ten different subjects. Results with a prediction horizon of 30 min show that prediction of glucose concentration in T1D via NN is feasible and sufficiently accurate. The average time anticipation obtained is compatible with the generation of preventive hypoglycemic and hyperglycemic alerts and the improvement of artificial pancreas performance.

The stereotactic approach for mapping epileptic networks: a prospective study of 200 patients

Object

Stereoelectroencephalography (SEEG) is a methodology that permits accurate 3D in vivo electroclinical recordings of epileptiform activity. Among other general indications for invasive intracranial electroencephalography (EEG) monitoring, its advantages include access to deep cortical structures, its ability to localize the epileptogenic zone when subdural grids have failed to do so, and its utility in the context of possible multifocal seizure onsets with the need for bihemispheric explorations. In this context, the authors present a brief historical overview of the technique and report on their experience with 2 SEEG techniques (conventional Leksell frame-based stereotaxy and frameless stereotaxy under robotic guidance) for the purpose of invasively monitoring difficult-to-localize refractory focal epilepsy.

Methods

Over a period of 4 years, the authors prospectively identified 200 patients with refractory epilepsy who collectively underwent 2663 tailored SEEG electrode implantations for invasive intracranial EEG monitoring and extraoperative mapping. The first 122 patients underwent conventional Leksell frame-based SEEG electrode placement; the remaining 78 patients underwent frameless stereotaxy under robotic guidance, following acquisition of a stereotactic ROSA robotic device at the authors' institution. Electrodes were placed according to a preimplantation hypothesis of the presumed epileptogenic zone, based on a standardized preoperative workup including video-EEG monitoring, MRI, PET, ictal SPECT, and neuropsychological assessment. Demographic features, seizure semiology, number and location of implanted SEEG electrodes, and location of the epileptogenic zone were recorded and analyzed for all patients. For patients undergoing subsequent craniotomy for resection, the type of resection and procedure-related complications were prospectively recorded. These results were analyzed and correlated with pathological diagnosis and postoperative seizure outcomes.

Results

The epileptogenic zone was confirmed by SEEG in 154 patients (77%), of which 134 (87%) underwent subsequent craniotomy for epileptogenic zone resection. Within this cohort, 90 patients had a minimum follow-up of at least 12 months; therein, 61 patients (67.8%) remained seizure free, with an average follow-up period of 2.4 years. The most common pathological diagnosis was focal cortical dysplasia Type I (55 patients, 61.1%). Per electrode, the surgical complications included wound infection (0.08%), hemorrhagic complications (0.08%), and a transient neurological deficit (0.04%) in a total of 5 patients (2.5%). One patient (0.5%) ultimately died due to intracerebral hematoma directly ensuing from SEEG electrode placement.

Conclusions

Based on these results, SEEG methodology is safe, reliable, and effective. It is associated with minimal morbidity and mortality, and serves as a practical, minimally invasive approach to extraoperative localization of the epileptogenic zone in patients with refractory epilepsy.

Epileptic focus localization based on resting state interictal MEG recordings is feasible irrespective of the presence or absence of spikes

OBJECTIVE

To investigate whether epileptogenic focus localization is possible based on resting state connectivity analysis of magnetoencephalographic (MEG) data.

METHODS

A multivariate autoregressive (MVAR) model was constructed using the sensor space data and was projected to the source space using lead field and inverse matrix. The generalized partial directed coherence was estimated from the MVAR model in the source space. The dipole with the maximum information inflow was hypothesized to be within the epileptogenic focus.

RESULTS

Applying the focus localization algorithm (FLA) to the interictal MEG recordings from five patients with neocortical epilepsy, who underwent presurgical evaluation for the identification of epileptogenic focus, we were able to correctly localize the focus, on the basis of maximum interictal information inflow in the presence or absence of interictal epileptic spikes in the data, with three out of five patients undergoing resective surgery and being seizure free since.

CONCLUSION

Our preliminary results suggest that accurate localization of the epileptogenic focus may be accomplished using noninvasive spontaneous "resting-state" recordings of relatively brief duration and without the need to capture definite interictal and/or ictal abnormalities.

SIGNIFICANCE

Epileptogenic focus localization is possible through connectivity analysis of resting state MEG data irrespective of the presence/absence of spikes.

Seizure detection, seizure prediction, and closed-loop warning systems in epilepsy

Nearly one-third of patients with epilepsy continue to have seizures despite optimal medication management. Systems employed to detect seizures may have the potential to improve outcomes in these patients by allowing more tailored therapies and might, additionally, have a role in accident and SUDEP prevention. Automated seizure detection and prediction require algorithms which employ feature computation and subsequent classification. Over the last few decades, methods have been developed to detect seizures utilizing scalp and intracranial EEG, electrocardiography, accelerometry and motion sensors, electrodermal activity, and audio/video captures. To date, it is unclear which combination of detection technologies yields the best results, and approaches may ultimately need to be individualized. This review presents an overview of seizure detection and related prediction methods and discusses their potential uses in closed-loop warning systems in epilepsy.

Real-time prediction of disordered breathing events in people with obstructive sleep apnea

PURPOSE

Conventional therapies for obstructive sleep apnea (OSA) are effective but suffer from poor patient adherence and may not fully alleviate major OSA-associated cardiovascular risk factors or improve certain aspects of quality of life. Predicting the onset of disordered breathing events in OSA patients may lead to improved strategies for treating OSA and inform our understanding of underlying disease mechanisms. In this work, we describe a deployable system capable of performing real-time predictions of sleep disordered breathing events in patients diagnosed with OSA, providing a novel approach for gaining insight into OSA pathophysiology, discovering population subgroups, and improving therapies.

METHODS

LArge Memory STorage and Retrieval artificial neural networks with 864 different configurations were applied to polysomnogram records from 64 patients. Wavelet transforms, measures of entropy, and other statistics were applied to six physiological signals to provide network inputs. Approximate statistical tests were used to determine the best performing network for each patient. The most important predictors of disordered breathing events in OSA patients were determined by analyzing internal network parameters.

RESULTS

The average optimized individual prediction sensitivity and specificity were 0.81 and 0.77, respectively. Predictions were better than random guessing for all OSA patients. Analysis of internal network parameters revealed a high degree of heterogeneity among disordered breathing event predictors and may reveal patient subgroups.

CONCLUSIONS

We report the first practical system to predict individual disordered breathing events in a heterogeneous group of patients diagnosed with OSA. The pattern of disordered breathing predictors suggests variable underlying pathophysiological mechanisms and highlights the need for an individualized approach to OSA diagnosis, therapy, and management.

Jump neural network for online short-time prediction of blood glucose from continuous monitoring sensors and meal information

Several real-time short-term prediction methods, based on time-series modeling of past continuous glucose monitoring (CGM) sensor data have been proposed with the aim of allowing the patient, on the basis of predicted glucose concentration, to anticipate therapeutic decisions and improve therapy of type 1 diabetes. In this field, neural network (NN) approaches could improve prediction performance handling in their inputs additional information. In this contribution we propose a jump NN prediction algorithm (horizon 30 min) that exploits not only past CGM data but also ingested carbohydrates information. The NN is tuned on data of 10 type 1 diabetics and then assessed on 10 different subjects. Results show that predictions of glucose concentration are accurate and comparable to those obtained by a recently proposed NN approach (Zecchin et al. (2012) [26]) having higher structural and algorithmical complexity and requiring the patient to announce the meals. This strengthen the potential practical usefulness of the new jump NN approach.

Peri-ictal ECG changes in childhood epilepsy: implications for detection systems

Early detection of seizures could reduce associated morbidity and mortality and improve the quality of life of patients with epilepsy. In this study, the aim was to investigate whether ictal tachycardia is present in focal and generalized epileptic seizures in children. We sought to predict in which type of seizures tachycardia can be identified before actual seizure onset. Electrocardiogram segments in 80 seizures were analyzed in time and frequency domains before and after the onset of epileptic seizures on EEG. These ECG parameters were analyzed to find the most informative ones that can be used for seizure detection. The algorithm of Leutmezer et al. was used to find the temporal relationship between the change in heart rate and seizure onset. In the time domain, the mean RR shows a significant difference before compared to after onset of the seizure in focal seizures. This can be observed in temporal lobe seizures as well as frontal lobe seizures. Calculation of mean RR interval has a high specificity for detection of ictal heart rate changes. Preictal heart rate changes are observed in 70% of the partial seizures. Ictal heart rate changes are present only in partial seizures in this childhood epilepsy study. The changes can be observed in temporal lobe seizures as well as in frontal lobe seizures. Heart rate changes precede seizure onset in 70% of the focal seizures, making seizure detection and closed-loop systems a possible therapeutic alternative in the population of children with refractory epilepsy.

PHASE SYNCHRONIZATION OF NEURONAL NOISE IN MOUSE HIPPOCAMPAL EPILEPTIFORM DYNAMICS

Organized brain activity is the result of dynamical, segregated neuronal signals that may be used to investigate synchronization effects using sophisticated neuroengineering techniques. Phase synchrony analysis, in particular, has emerged as a promising methodology to study transient and frequency-specific coupling effects across multi-site signals. In this study, we investigated phase synchronization in intracellular recordings of interictal and ictal epileptiform events recorded from pairs of cells in the whole (intact) mouse hippocampus. In particular, we focused our analysis on the background noise-like activity (NLA), previously reported to exhibit complex neurodynamical properties. Our results show evidence for increased linear and nonlinear phase coupling in NLA across three frequency bands [theta (4–10 Hz), beta (12–30 Hz) and gamma (30–80 Hz)] in the ictal compared to interictal state dynamics. We also present qualitative and statistical evidence for increased phase synchronization in the theta, beta and gamma frequency bands from paired recordings of ictal NLA. Overall, our results validate the use of background NLA in the neurodynamical study of epileptiform transitions and suggest that what is considered "neuronal noise" is amenable to synchronization effects in the spatiotemporal domain.

Resetting of brain dynamics: epileptic versus psychogenic nonepileptic seizures

We investigated the possibility of differential diagnosis of patients with epileptic seizures (ES) and patients with psychogenic nonepileptic seizures (PNES) through an advanced analysis of the dynamics of the patients' scalp EEGs. The underlying principle was the presence of resetting of brain's preictal spatiotemporal entrainment following onset of ES and the absence of resetting following PNES. Long-term (days) scalp EEGs recorded from five patients with ES and six patients with PNES were analyzed. It was found that: (1) Preictal entrainment of brain sites was reset at ES (P<0.05) in four of the five patients with ES, and not reset (P=0.28) in the fifth patient. (2) Resetting did not occur (p>0.1) in any of the six patients with PNES. These preliminary results in patients with ES are in agreement with our previous findings from intracranial EEG recordings on resetting of brain dynamics by ES and are expected to constitute the basis for the development of a reliable and supporting tool in the differential diagnosis between ES and PNES. Finally, we believe that these results shed light on the electrophysiology of PNES by showing that occurrence of PNES does not assist patients in overcoming a pathological entrainment of brain dynamics. This article is part of a Supplemental Special Issue entitled The Future of Automated Seizure Detection and Prediction.