Biomimetic Deep Learning Networks With Applications to Epileptic Spasms and Seizure Prediction.

OBJECTIVE: In this study, we present a novel biomimetic deep learning network for epileptic spasms and seizure prediction and compare its performance with state-of-the-art conventional machine learning models. METHODS: Our proposed model incorporates modular Volterra kernel convolutional networks and bidirectional recurrent networks in combination with the phase amplitude cross-frequency coupling features derived from scalp EEG. They are applied to the standard CHB-MIT dataset containing focal epilepsy episodes as well as two other datasets from the Montefiore Medical Center and the University of California Los Angeles that provide data of patients experiencing infantile spasm (IS) syndrome. RESULTS: Overall, in this study, the networks can produce accurate predictions (100%) and significant detection latencies (10 min). Furthermore, the biomimetic network outperforms conventional ones by producing no false positives. SIGNIFICANCE: Biomimetic neural networks utilize extensive knowledge about processing and learning in the electrical networks of the brain. Predicting seizures in adults can improve their quality of life. Epileptic spasms in infants are part of a particular seizure type that needs identifying when suspicious behaviors are noticed in babies. Predicting epileptic spasms within a given time frame (the prediction horizon) suggests their existence and allows an epileptologist to flag an EEG trace for future review.

EEG-based spatiotemporal dynamics of fast ripple networks and hubs in infantile epileptic spasms.

OBJECTIVE: Infantile epileptic spasms (IS) are epileptic seizures that are associated with increased risk for developmental impairments, adult epilepsies, and mortality. Here, we investigated coherence-based network dynamics in scalp EEG of infants with IS to identify frequency-dependent networks associated with spasms. We hypothesized that there is a network of increased fast ripple connectivity during the electrographic onset of clinical spasms, which is distinct from controls. METHODS: We retrospectively analyzed peri-ictal and interictal EEG recordings of 14 IS patients. The data was compared with 9 age-matched controls. Wavelet phase coherence (WPC) was computed between 0.2 and 400 Hz. Frequency- and time-dependent brain networks were constructed using this coherence as the strength of connection between two EEG channels, based on graph theory principles. Connectivity was evaluated through global efficiency (GE) and channel-based closeness centrality (CC), over frequency and time. RESULTS: GE in the fast ripple band (251-400 Hz) was significantly greater following the onset of spasms in all patients (P < 0.05). Fast ripple networks during the first 10s from spasm onset show enhanced anteroposterior gradient in connectivity (posterior > central > anterior, Kruskal-Wallis P < 0.001), with maximum CC over the centroparietal channels in 10/14 patients. Additionally, this anteroposterior gradient in CC connectivity is observed during spasms but not during the interictal awake or asleep states of infants with IS. In controls, anteroposterior gradient in fast ripple CC was noted during arousals and wakefulness but not during sleep. There was also a simultaneous decrease in GE in the 5-8 Hz range after the onset of spasms (P < 0.05), of unclear biological significance. SIGNIFICANCE: We identified an anteroposterior gradient in the CC connectivity of fast ripple hubs during spasms. This anteroposterior gradient observed during spasms is similar to the anteroposterior gradient in the CC connectivity observed in wakefulness or arousals in controls, suggesting that this state change is related to arousal networks.

Ictal ECG-based assessment of sudden unexpected death in epilepsy.

Introduction: Previous case-control studies of sudden unexpected death in epilepsy (SUDEP) patients failed to identify ECG features (peri-ictal heart rate, heart rate variability, corrected QT interval, postictal heart rate recovery, and cardiac rhythm) predictive of SUDEP risk. This implied a need to derive novel metrics to assess SUDEP risk from ECG. Methods: We applied Single Spectrum Analysis and Independent Component Analysis (SSA-ICA) to remove artifact from ECG recordings. Then cross-frequency phase-phase coupling (PPC) was applied to a 20-s mid-seizure window and a contour of -3 dB coupling strength was determined. The contour centroid polar coordinates, amplitude (alpha) and angle (theta), were calculated. Association of alpha and theta with SUDEP was assessed and a logistic classifier for alpha was constructed. Results: Alpha was higher in SUDEP patients, compared to non-SUDEP patients (p < 0.001). Theta showed no significant difference between patient populations. The receiver operating characteristic (ROC) of a logistic classifier for alpha resulted in an area under the ROC curve (AUC) of 94% and correctly classified two test SUDEP patients. Discussion: This study develops a novel metric alpha, which highlights non-linear interactions between two rhythms in the ECG, and is predictive of SUDEP risk.

Interregional phase-amplitude coupling between theta rhythm in the nucleus tractus solitarius and high-frequency oscillations in the hippocampus during REM sleep in rats.

Cross-frequency coupling (CFC) between theta and high-frequency oscillations (HFOs) is predominant during active wakefulness, REM sleep and behavioral and learning tasks in rodent hippocampus. Evidence suggests that these state-dependent CFCs are linked to spatial navigation and memory consolidation processes. CFC studies currently include only the cortical and subcortical structures. To our knowledge, the study of nucleus tractus solitarius (NTS)-cortical structure CFC is still lacking. Here we investigate CFC in simultaneous local field potential recordings from hippocampal CA1 and the NTS during behavioral states in freely moving rats. We found a significant increase in theta (6-8 Hz)-HFO (120-160 Hz) coupling both within the hippocampus and between NTS theta and hippocampal HFOs during REM sleep. Also, the hippocampal HFOs were modulated by different but consistent phases of hippocampal and NTS theta oscillations. These findings support the idea that phase-amplitude coupling is both state- and frequency-specific and CFC analysis may serve as a tool to help understand the selective functions of neuronal network interactions in state-dependent information processing. Importantly, the increased NTS theta-hippocampal HFO coupling during REM sleep may represent the functional connectivity between these two structures which reflects the function of the hippocampus in visceral learning with the sensory information provided by the NTS. This gives a possible insight into an association between the sensory activity and REM-sleep dependent memory consolidation.

A Peri-Ictal EEG-Based Biomarker for Sudden Unexpected Death in Epilepsy (SUDEP) Derived From Brain Network Analysis.

Sudden unexpected death in epilepsy (SUDEP) is the leading seizure-related cause of death in epilepsy patients. There are no validated biomarkers of SUDEP risk. Here, we explored peri-ictal differences in topological brain network properties from scalp EEG recordings of SUDEP victims. Functional connectivity networks were constructed and examined as directed graphs derived from undirected delta and high frequency oscillation (HFO) EEG coherence networks in eight SUDEP and 14 non-SUDEP epileptic patients. These networks were proxies for information flow at different spatiotemporal scales, where low frequency oscillations coordinate large-scale activity driving local HFOs. The clustering coefficient and global efficiency of the network were higher in the SUDEP group pre-ictally, ictally and post-ictally (p < 0.0001 to p < 0.001), with features characteristic of small-world networks. These results suggest that cross-frequency functional connectivity network topology may be a non-invasive biomarker of SUDEP risk.

Altered neocortical oscillations and cellular excitability in an in vitro Wwox knockout mouse model of epileptic encephalopathy.

Loss of function mutations of the WW domain-containing oxidoreductase (WWOX) gene are associated with severe and fatal drug-resistant pediatric epileptic encephalopathy. Epileptic seizures are typically characterized by neuronal hyperexcitability; however, the specific contribution of WWOX to that hyperexcitability has yet to be investigated. Using a mouse model of neuronal Wwox-deletion that exhibit spontaneous seizures, in vitro whole-cell and field potential electrophysiological characterization identified spontaneous bursting activity in the neocortex, a marker of the underlying network hyperexcitability. Spectral analysis of the neocortical bursting events highlighted increased phase-amplitude coupling, and a propagation from layer II/III to layer V. These bursts were NMDAR and gap junction dependent. In layer II/III pyramidal neurons, Wwox knockout mice demonstrated elevated amplitude of excitatory post-synaptic currents, whereas the frequency and amplitude of inhibitory post-synaptic currents were reduced, as compared to heterozygote and wild-type littermate controls. Furthermore, these neurons were depolarized and demonstrated increased action potential frequency, sag current, and post-inhibitory rebound. These findings suggest WWOX plays an essential role in balancing neocortical excitability and provide insight towards developing therapeutics for those suffering from WWOX disorders.

Glial Modulation of Electrical Rhythms in a Neuroglial Network Model of Epilepsy.

OBJECTIVE: An important EEG-based biomarker for epilepsy is the phase-amplitude cross-frequency coupling (PAC) of electrical rhythms; however, the underlying pathways of these pathologic markers are not always clear. Since glial cells have been shown to play an active role in neuroglial networks, it is likely that some of these PAC markers are modulated via glial effects. METHODS: We developed a 4-unit hybrid model of a neuroglial network, consisting of 16 sub-units, that combines a mechanistic representation of neurons with an oscillator-based Cognitive Rhythm Generator (CRG) representation of glial cells-astrocytes and microglia. The model output was compared with recorded generalized tonic-clonic patient data, both in terms of PAC features, and state classification using an unsupervised hidden Markov model (HMM). RESULTS: The neuroglial model output showed PAC features similar to those observed in epileptic seizures. These generated PAC features were able to accurately identify spontaneous epileptiform discharges (SEDs) as seizure-like states, as well as a postictal-like state following the long-duration SED, when applied to the HMM machine learning algorithm trained on patient data. The evolution profile of the maximal PAC during the SED compared well with patient data, showing similar association with the duration of the postictal state. CONCLUSION: The hybrid neuroglial network model was able to generate PAC features similar to those observed in ictal and postictal epileptic states, which has been used for state classification and postictal state duration prediction. SIGNIFICANCE: Since PAC biomarkers are important for epilepsy research and postictal state duration has been linked with risk of sudden unexplained death in epilepsy, this model suggests glial synaptic effects as potential targets for further analysis and treatment.

Delta-gamma phase-amplitude coupling as a biomarker of postictal generalized EEG suppression.

Postictal generalized EEG suppression is the state of suppression of electrical activity at the end of a seizure. Prolongation of this state has been associated with increased risk of sudden unexpected death in epilepsy, making characterization of underlying electrical rhythmic activity during postictal suppression an important step in improving epilepsy treatment. Phase-amplitude coupling in EEG reflects cognitive coding within brain networks and some of those codes highlight epileptic activity; therefore, we hypothesized that there are distinct phase-amplitude coupling features in the postictal suppression state that can provide an improved estimate of this state in the context of patient risk for sudden unexpected death in epilepsy. We used both intracranial and scalp EEG data from eleven patients (six male, five female; age range 21-41 years) containing 25 seizures, to identify frequency dynamics, both in the ictal and postictal EEG suppression states. Cross-frequency coupling analysis identified that during seizures there was a gradual decrease of phase frequency in the coupling between delta (0.5-4 Hz) and gamma (30+ Hz), which was followed by an increased coupling between the phase of 0.5-1.5 Hz signal and amplitude of 30-50 Hz signal in the postictal state as compared to the pre-seizure baseline. This marker was consistent across patients. Then, using these postictal-specific features, an unsupervised state classifier-a hidden Markov model-was able to reliably classify four distinct states of seizure episodes, including a postictal suppression state. Furthermore, a connectome analysis of the postictal suppression states showed increased information flow within the network during postictal suppression states as compared to the pre-seizure baseline, suggesting enhanced network communication. When the same tools were applied to the EEG of an epilepsy patient who died unexpectedly, ictal coupling dynamics disappeared and postictal phase-amplitude coupling remained constant throughout. Overall, our findings suggest that there are active postictal networks, as defined through coupling dynamics that can be used to objectively classify the postictal suppression state; furthermore, in a case study of sudden unexpected death in epilepsy, the network does not show ictal-like phase-amplitude coupling features despite the presence of convulsive seizures, and instead demonstrates activity similar to postictal. The postictal suppression state is a period of elevated network activity as compared to the baseline activity which can provide key insights into the epileptic pathology.

Transitions between neocortical seizure and non-seizure-like states and their association with presynaptic glutamate release.

The transition between seizure and non-seizure states in neocortical epileptic networks is governed by distinct underlying dynamical processes. Based on the gamma distribution of seizure and inter-seizure durations, over time, seizures are highly likely to self-terminate; whereas, inter-seizure durations have a low chance of transitioning back into a seizure state. Yet, the chance of a state transition could be formed by multiple overlapping, unknown synaptic mechanisms. To identify the relationship between the underlying synaptic mechanisms and the chance of seizure-state transitions, we analyzed the skewed histograms of seizure durations in human intracranial EEG and seizure-like events (SLEs) in local field potential activity from mouse neocortical slices, using an objective method for seizure state classification. While seizures and SLE durations were demonstrated to have a unimodal distribution (gamma distribution shape parameter >1), suggesting a high likelihood of terminating, inter-SLE intervals were shown to have an asymptotic exponential distribution (gamma distribution shape parameter <1), suggesting lower probability of cessation. Then, to test cellular mechanisms for these distributions, we studied the modulation of synaptic neurotransmission during, and between, the in vitro SLEs. Using simultaneous local field potential and whole-cell voltage clamp recordings, we found a suppression of presynaptic glutamate release at SLE termination, as demonstrated by electrically- and optogenetically-evoked excitatory postsynaptic currents (EPSCs), and focal hypertonic sucrose application. Adenosine A1 receptor blockade interfered with the suppression of this release, changing the inter-SLE shape parameter from asymptotic exponential to unimodal, altering the chance of state transition occurrence with time. These findings reveal a critical role for presynaptic glutamate release in determining the chance of neocortical seizure state transitions.

Pannexin-1 Deficiency Decreases Epileptic Activity in Mice.

OBJECTIVE: Pannexin-1 (Panx1) is suspected of having a critical role in modulating neuronal excitability and acute neurological insults. Herein, we assess the changes in behavioral and electrophysiological markers of excitability associated with Panx1 via three distinct models of epilepsy. Methods Control and Panx1 knockout C57Bl/6 mice of both sexes were monitored for their behavioral and electrographic responses to seizure-generating stimuli in three epilepsy models-(1) systemic injection of pentylenetetrazol, (2) acute electrical kindling of the hippocampus and (3) neocortical slice exposure to 4-aminopyridine. Phase-amplitude cross-frequency coupling was used to assess changes in an epileptogenic state resulting from Panx1 deletion. RESULTS: Seizure activity was suppressed in Panx1 knockouts and by application of Panx1 channel blockers, Brilliant Blue-FCF and probenecid, across all epilepsy models. In response to pentylenetetrazol, WT mice spent a greater proportion of time experiencing severe (stage 6) seizures as compared to Panx1-deficient mice. Following electrical stimulation of the hippocampal CA3 region, Panx1 knockouts had significantly shorter evoked afterdischarges and were resistant to kindling. In response to 4-aminopyridine, neocortical field recordings in slices of Panx1 knockout mice showed reduced instances of electrographic seizure-like events. Cross-frequency coupling analysis of these field potentials highlighted a reduced coupling of excitatory delta-gamma and delta-HF rhythms in the Panx1 knockout. SIGNIFICANCE: These results suggest that Panx1 plays a pivotal role in maintaining neuronal hyperexcitability in epilepsy models and that genetic or pharmacological targeting of Panx1 has anti-convulsant effects.

Classification of Scalp EEG States Prior to Clinical Seizure Onset.

OBJECTIVE: To investigate the feasibility of improving the performance of an EEG-based multistate classifier (MSC) previously proposed by our group. RESULTS: Using the random forest (RF) classifiers on the previously reported dataset of patients, but with three improvements to classification logic, the specificity of our alarm algorithm improves from 82.4% to 92.0%, and sensitivity from 87.9% to 95.2%. DISCUSSION: The MSC could be a useful approach for seizure-monitoring both in the clinic and at home. METHODS: Three improvements to the MSC are described. Firstly, an additional check using RF outputs is made prior to alarm to confirm increasing probability of a seizure onset state. Secondly, a post-alarm detection horizon that accounts for the seizure state duration is implemented. Thirdly, the alarm decision window is kept constant.

Epilepsy as a manifestation of a multistate network of oscillatory systems.

The human brain, largely accepted as the most complex biological system known, is still far from being understood in its parts or as a whole. More specifically, biological mechanisms of epileptic states and state transitions are not well understood. Here, we explore the concept of the epilepsy as a manifestation of a multistate network composed of coupled oscillatory units. We also propose that functional coupling between neuroglial elements is a dynamic process, characterized by temporal changes both at short and long time scales. We review various experimental and modelling data suggesting that epilepsy is a pathological manifestation of such a multistate network - both when viewed as a coupled oscillatory network, and as a system of multistate stable state attractors. Based on a coupled oscillators model, we propose a significant role for glial cells in modulating hyperexcitability of the neuroglial networks of the brain. Also, using these concepts, we explain a number of observable phenomena such as propagation patterns of bursts within a seizure in the isolated intact hippocampus in vitro, postictal generalized suppression in human encephalographic seizure data, and changes in seizure susceptibility in epileptic patients. Based on our conceptual model we propose potential clinical applications to estimate brain closeness to ictal transition by means of active perturbations and passive measures during on-going activity.

Cross-Frequency Coupling Features of Postictal Generalized EEG Suppression State.

In patients with epilepsy, convulsive seizures are often followed by a postictal generalized EEG suppression (PGES) state characterized by reduced background activity. Recent studies found a correlation between seizure termination state and PGES duration, and suggested that PGES is the result of the cessation of neuronal activity. To test that assertion, we investigated ten seizure records obtained from intracranial EEG (iEEG) from six patients, four of which had Engel Class 1 surgical outcome. In each case expert neurologists identified the most likely seizure onset electrode. We found the iEEG equivalent of PGES and an artifact-free preictal quiescent state of the same window size. Using index of cross-frequency coupling (ICFC) we identified the degree of coupling and dominant frequency bands involved in PGES and preictal quiescent states, and quantified the areas of high ICFC. We found that there was an increase in the degree of coupling between the 0.5-1.5Hz with high gamma frequency bands in the PGES states. We found that among all of the patients, as well as in Engel Class 1 patients specifically, the change in the quantified area of high ICFC was significant (p <; 0.05) between PGES and preictal quiescent states. Furthermore, we were able to identify whether a recording was from a depth or subdural electrode, or whether it was from seizure onset zone or not using ICFC markers in PGES. This suggests that there are frequency coupling markers that successfully identify PGES and that there are underlying dynamics that occur in this seemingly quiet postictal state.

Coupled Oscillators Model of Hyperexcitable Neuroglial Networks.

Glial populations within neuronal networks of the brain have recently gained much interest in the context of hyperexcitability and epilepsy. In this paper, we present an oscillator-based neuroglial model capable of generating Spontaneous Electrical Discharges (SEDs) in hyperexcitable conditions. The network is composed of 16 coupled Cognitive Rhythm Generators (CRGs), which are oscillator-based mathematical constructs previously described by our research team. CRGs are well-suited for modeling assemblies of excitable cells, and in this network, each represents one of the following populations: excitatory pyramidal cells, inhibitory interneurons, astrocytes, and microglia. We investigated various pathways leading to hyperexcitability, and our results suggest an important role for astrocytes and microglia in the generation of SEDs of various durations. Analysis of the resultant SEDs revealed two underlying duration distributions with differing properties. Particularly, short and long SEDs are associated with deterministic and random underlying processes, respectively. The mesoscale of this model makes it well-suited for (a) the elucidation of glia-related hypotheses in hyperexcitable conditions, (b) use as a testing platform for neuromodulation purposes, and (c) a hardware implementation for closed-loop neuromodulation.

Neuro-Glial Network Model Of Postictal Generalized EEG Suppression (PGES).

Over the past couple of decades, glial cells have been highlighted as active agents in hyperexcitability of neuronal networks, specifically playing key roles in seizure onset and termination. In particular, microglia have been suggested to have both neuroprotective and neurotoxic effects on the brain. Investigation into seizure termination is of particular interest, as it is sometimes followed by a postictal generalized EEG suppression (PGES) - a low activity state that is potentially associated with sudden unexpected death in epilepsy. In this study, we attempt to link glial effects - synaptic pruning and astrocytic potassium clearance - to the duration of spontaneous epileptiform discharges (SEDs) as well as interSED intervals (iSEDs). We build upon an earlier model of a neuroglial network by translating it into the cortical paradigm and including microglial units. Preliminary findings of our model demonstrated that the duration of SEDs is largely determined by the astrocytic potassium clearance, whereas iSEDs significantly increased with microglial-driven synaptic pruning. In our model, astrocytic potassium clearance itself did not bring a PGES-like state, whereas microglial effects did, which suggests a potential biomarker for PGES phenomena.

Classification of Pre-Clinical Seizure States Using Scalp EEG Cross-Frequency Coupling Features.

OBJECTIVE: This work proposes a machine-learning based system for a scalp EEG that flags an alarm in advance of a clinical seizure onset. METHODS: EEG recordings from 12 patients with drug resistant epilepsy were marked by an expert neurologist for clinical seizure onset. Scalp EEG recordings consisted of 56 seizures and 9.67 h of interictal periods. Data from six patients were reserved for testing, and the rest was split into training and testing sets. A global spatial average of a cross-frequency coupling (CFC) index, , was extracted in 2 s windows, and used as the feature for the machine learning. A multistage state classifier (MSC) based on random forest algorithms was trained and tested on these data. Training was conducted to classify three states: interictal baseline, and segments prior to and following EG onset. Classifier performance was assessed using a receiver-operating characteristic (ROC) analysis. RESULTS: The MSC produced an alarm 45 16 s in advance of a clinical seizure onset across seizures from the 12 patients. It performed with a sensitivity of 87.9%, a specificity of 82.4%, and an area-under-the-ROC of 93.4%. On patients for whom it received training, performance metrics increased. Performance metrics did not change when the MSC used reduced electrode ring configurations. CONCLUSION: Using the scalp , the MSC produced an alarm in advance of a clinical seizure onset for all 12 patients. Patient-specific training improved the specificity of classification. SIGNIFICANCE: The MSC is noninvasive, and demonstrates that CFC features may be suitable for use in a home-based seizure monitoring system.

Electrographic and pharmacological characterization of a progressive epilepsy phenotype in female MeCP2-deficient mice.

Rett Syndrome is a neurodevelopmental disorder caused primarily by mutations in the gene encoding Methyl-CpG-binding protein 2 (MECP2). Spontaneous epileptiform activity is a common co-morbidity present in Rett syndrome, and hyper-excitable neural networks are present in MeCP2-deficient mouse models of Rett syndrome. In this study we conducted a longitudinal assessment of spontaneous cortical electrographic discharges in female MeCP2-deficient mice and defined the pharmacological responsiveness of these discharges to anti-convulsant drugs. Our data show that cortical discharge activity in female MeCP2-deficient mice progressively increases in severity as the mice age, with discharges being more frequent and of longer durations at 19-24 months of age compared to 3 months of age. Semiologically and pharmacologically, this basal discharge activity in female MeCP2-deficient mice displayed electroclinical properties consistent with absence epilepsy. Only rarely were convulsive seizures observed in these mice at any age. Since absence epilepsy is infrequently observed in Rett syndrome patients, these results indicate that the predominant spontaneous electroclinical phenotype of MeCP2-deficient mice we examined does not faithfully recapitulate the most prevalent seizure types observed in affected patients.

Low-to-High Cross-Frequency Coupling in the Electrical Rhythms as Biomarker for Hyperexcitable Neuroglial Networks of the Brain.

OBJECTIVE: One of the features used in the study of hyperexcitablility is high-frequency oscillations (HFOs, >80 Hz). HFOs have been reported in the electrical rhythms of the brain's neuroglial networks under physiological and pathological conditions. Cross-frequency coupling (CFC) of HFOs with low-frequency rhythms was used to identify pathologic HFOs in the epileptogenic zones of epileptic patients and as a biomarker for the severity of seizure-like events in genetically modified rodent models. We describe a model to replicate reported CFC features extracted from recorded local field potentials (LFPs) representing network properties. METHODS: This study deals with a four-unit neuroglial cellular network model where each unit incorporates pyramidal cells, interneurons, and astrocytes. Three different pathways of hyperexcitability generation-Na - ATPase pump, glial potassium clearance, and potassium afterhyperpolarization channel-were used to generate LFPs. Changes in excitability, average spontaneous electrical discharge (SED) duration, and CFC were then measured and analyzed. RESULTS: Each parameter caused an increase in network excitability and the consequent lengthening of the SED duration. Short SEDs showed CFC between HFOs and theta oscillations (4-8 Hz), but in longer SEDs the low frequency changed to the delta range (1-4 Hz). CONCLUSION: Longer duration SEDs exhibit CFC features similar to those reported by our team. SIGNIFICANCE: First, Identifying the exponential relationship between network excitability and SED durations; second, highlighting the importance of glia in hyperexcitability (as they relate to extracellular potassium); and third, elucidation of the biophysical basis for CFC coupling features.

Prediction of antiepileptic drug treatment outcomes using machine learning.

OBJECTIVE: Antiepileptic drug (AED) treatments produce inconsistent outcomes, often necessitating patients to go through several drug trials until a successful treatment can be found. This study proposes the use of machine learning techniques to predict epilepsy treatment outcomes of commonly used AEDs. APPROACH: Machine learning algorithms were trained and evaluated using features obtained from intracranial electroencephalogram (iEEG) recordings of the epileptiform discharges observed in Mecp2-deficient mouse model of the Rett Syndrome. Previous work have linked the presence of cross-frequency coupling (I CFC) of the delta (2-5 Hz) rhythm with the fast ripple (400-600 Hz) rhythm in epileptiform discharges. Using the I CFC to label post-treatment outcomes we compared support vector machines (SVMs) and random forest (RF) machine learning classifiers for providing likelihood scores of successful treatment outcomes. MAIN RESULTS: (a) There was heterogeneity in AED treatment outcomes, (b) machine learning techniques could be used to rank the efficacy of AEDs by estimating likelihood scores for successful treatment outcome, (c) I CFC features yielded the most effective a priori identification of appropriate AED treatment, and (d) both classifiers performed comparably. SIGNIFICANCE: Machine learning approaches yielded predictions of successful drug treatment outcomes which in turn could reduce the burdens of drug trials and lead to substantial improvements in patient quality of life.

Epileptogenic Source Imaging Using Cross-Frequency Coupled Signals From Scalp EEG.

OBJECTIVE: The epileptogenic zone (EZ) is a brain region containing the sources of seizure genesis. Removal of the EZ is associated with cessation of seizures after resective surgical procedures, as measured by Engel Class I score. This study describes a novel EEG (electroencephalography) source imaging (ESI) method which uses cross-frequency coupled potential signals (SCFC) derived from scalp EEG. METHODS: Scalp EEG were recorded from ten patients (20 seizures) suffering from epilepsy. The S CFC were constructed from the phase and amplitude of the lower and higher frequency rhythms at electrographic seizure onset. ESI was then performed using the SCFC. Validation of the technique was facilitated by forward and inverse computer modeling of known cortical sources, and the correspondence of the ESI with EZ in resected regions of patients. RESULTS: For ten seizures sampled at or above 500 Hz from four patients, all estimated sources lay within the resected region, emphasizing the clinical importance of higher sampling rates. The SCFC demonstrated significant advantages over the "raw" scalp EEG, indicating its robust noise performance. Modeling investigations indicated that a signal-to-noise ratio above 0.2 was sufficient to achieve successful localization regarding EMG artifacts. CONCLUSION: The association of the estimated sources to the EZ suggests that cross-frequency coupling is a feature of the brain's neural networks, not of artifactual activity. The SCFC can effectively extract brain signals from a noisy background. SIGNIFICANCE: We propose this approach to enhance the placement of intracranial electrode for surgical intervention.