Epilepsy EEG Research Articles

Applying Deep Learning for Epilepsy Seizure Detection and Brain Mapping Visualization

Deep Convolutional Neural Network (CNN) has achieved remarkable results in computer vision tasks for end-to-end learning. We evaluate here the power of a deep CNN to learn robust features from raw Electroencephalogram (EEG) data to detect seizures. Seizures are hard to detect, as they vary both inter- and intra-patient. In this article, we use a deep CNN model for seizure detection task on an open-access EEG epilepsy dataset collected at the Boston Children's Hospital. Our deep learning model is able to extract spectral, temporal features from EEG epilepsy data and use them to learn the general structure of a seizure that is less sensitive to variations. For cross-patient EEG data, our method produced an overall sensitivity of 90.00%, specificity of 91.65%, and overall accuracy of 98.05% for the whole dataset of 23 patients. The system can detect seizures with an accuracy of 99.46%. Thus, it can be used as an excellent cross-patient seizure classifier. The results show that our model performs better than the previous state-of-the-art models for patient-specific and cross-patient seizure detection task. The method gave an overall accuracy of 99.65% for patient-specific data. The system can also visualize the special orientation of band power features. We use correlation maps to relate spectral amplitude features to the output in the form of images. By using the results from our deep learning model, this visualization method can be used as an effective multimedia tool for producing quick and relevant brain mapping images that can be used by medical experts for further investigation.

Classification for EEG report generation and epilepsy detection

An automatic generation of medical report method (AGMedRep) was proposed in order to process electroencephalogram (EEG) segments using machine learning (ML) to generate textual reports for epilepsy detection. This method is applied in two phases: (1) predictive model building, and (2) automatic generation of medical reports. In Phase 1, classifiers are built and evaluated, in which the best predictive model is chosen to perform Phase 2 for report generation. The AGMedRep was applied in a set of 500 EEG segments, 100 for each class, in a total of five classes. In the first phase, 90 signal segments for each class were selected for feature extraction and classifier building. Posteriorly, five different report expression were defined, one for each class. In each EEG segment, all possible combination of cross-correlogram, power spectrum, spectrogram (SG), and bispectrogram (BS) were computed for feature extraction, generating a total of 15 different datasets. Afterwards, ML methods, such as decision tree, 1-nearest-neighbor (1NN), naive Bayes, backpropagation based on multilayer perceptron (MLP), and support vector machine (SVM) were applied in each dataset considering a multiclass decomposition approach, generating a total of 75 (15 datasets × 5 ML techniques) classifiers. In the evaluation step, it was found that the “SG” 1NN, “SG and BS” MLP, and “SG and BS” SVM reached, statistically, the best performances for signal classification, but a statistical difference was not found among them. In this sense, the “SG” 1NN model was chosen for report generation due to its lower computation cost, and it proved to be one of the most accurate models. In the second phase, the 50 remaining EEG segments were equally distributed, randomly, into ten folders in order to simulate individual EEG exams. Finally, the chosen classifier was computed in each examination to construct textual reports. As a result, the 1NN model reached an average of 84% accuracy in the report generation. Our results suggest that the AGMedRep can be able to assist medical experts in the identification of patterns related to epileptic events in EEG.

Claims data analyses unable to properly characterize the value of neurologists in epilepsy care.

To determine the association of a neurologist visit with health care use and cost outcomes for patients with incident epilepsy. Using health care claims data for individuals insured by United Healthcare from 2001 to 2016, we identified patients with incident epilepsy. The population was defined by an epilepsy/convulsion diagnosis code (ICD codes 345.xx/780.3x, G40.xx/R56.xx), an antiepileptic prescription filled within the succeeding 2 years, and neither criterion met in the 2 preceding years. Cases were defined as patients who had a neurologist encounter for epilepsy within 1 year after an incident diagnosis; a control cohort was constructed with propensity score matching. Primary outcomes were emergency room (ER) visits and hospitalizations for epilepsy. Secondary outcomes included measures of cost (epilepsy related, not epilepsy related, and antiepileptic drugs) and care escalation (including EEG evaluation and epilepsy surgery). After participant identification and propensity score matching, there were 3,400 cases and 3,400 controls. Epilepsy-related ER visits were more likely for cases than controls (year 1: 5.9% vs 2.3%, p < 0.001), as were hospitalizations (year 1: 2.1% vs 0.7%, p < 0.001). Total medical costs for epilepsy care, nonepilepsy care, and antiepileptic drugs were greater for cases (p ≤ 0.001). EEG evaluation and epilepsy surgery occurred more commonly for cases (p ≤ 0.001). Patients with epilepsy who visited a neurologist had greater subsequent health care use, medical costs, and care escalation than controls. This comparison using administrative claims is plausibly confounded by case disease severity, as suggested by higher nonepilepsy care costs. Linking patient-centered outcomes to claims data may provide the clinical resolution to assess care value within a heterogeneous population.

A Seizure-Based Power Reduction SoC for Wearable EEG in Epilepsy

Epilepsy is one of the most common serious brain disorders affecting 1% of the world population. Epileptic seizure events are caused by abnormal excessive neuronal activity in the brain, which may be associated with behavioural changes that severely affect the patients' quality of life. These events are manifested as abnormal activity in electroencephalography (EEG) recordings of individuals with epilepsy. This paper presents the on-chip implementation of an algorithm that, operating on the principle of data selection applied to seizures, would be able to reduce the power consumption of EEG devices, and consequently their size, thereby significantly increasing their usability. In order to reduce the power consumed by the on-chip implementation of the algorithm, mathematical approximations have been carried out to allow for an analog implementation, resulting in the power consumed by the system to be negligible in comparison to other blocks in an EEG device. The system has been fabricated in a 0.18 μm CMOS process, consumes 1.14 μW from a 1.25 V supply and achieves a sensitivity of 98.5% while only selecting 52.5% of the EEG data for transmission.

BOLD-fMRI activity informed by network variation of scalp EEG in juvenile myoclonic epilepsy.

Epilepsy is marked by hypersynchronous bursts of neuronal activity, and seizures can propagate variably to any and all areas, leading to brain network dynamic organization. However, the relationship between the network characteristics of scalp EEG and blood oxygenation level-dependent (BOLD) responses in epilepsy patients is still not well known. In this study, simultaneous EEG and fMRI data were acquired in 18 juvenile myoclonic epilepsy (JME) patients. Then, the adapted directed transfer function (ADTF) values between EEG electrodes were calculated to define the time-varying network. The variation of network information flow within sliding windows was used as a temporal regressor in fMRI analysis to predict the BOLD response. To investigate the EEG-dependent functional coupling among the responding regions, modulatory interactions were analyzed for network variation of scalp EEG and BOLD time courses. The results showed that BOLD activations associated with high network variation were mainly located in the thalamus, cerebellum, precuneus, inferior temporal lobe and sensorimotor-related areas, including the middle cingulate cortex (MCC), supplemental motor area (SMA), and paracentral lobule. BOLD deactivations associated with medium network variation were found in the frontal, parietal, and occipital areas. In addition, modulatory interaction analysis demonstrated predominantly directional negative modulation effects among the thalamus, cerebellum, frontal and sensorimotor-related areas. This study described a novel method to link BOLD response with simultaneous functional network organization of scalp EEG. These findings suggested the validity of predicting epileptic activity using functional connectivity variation between electrodes. The functional coupling among the thalamus, frontal regions, cerebellum and sensorimotor-related regions may be characteristically involved in epilepsy generation and propagation, which provides new insight into the pathophysiological mechanisms and intervene targets for JME.

SYNGAP1 encephalopathy: A distinctive generalized developmental and epileptic encephalopathy.

ObjectiveTo delineate the epileptology, a key part of the SYNGAP1 phenotypic spectrum, in a large patient cohort.MethodsPatients were recruited via investigators' practices or social media. We included patients with (likely) pathogenic SYNGAP1 variants or chromosome 6p21.32 microdeletions incorporating SYNGAP1. We analyzed patients' phenotypes using a standardized epilepsy questionnaire, medical records, EEG, MRI, and seizure videos.ResultsWe included 57 patients (53% male, median age 8 years) with SYNGAP1 mutations (n = 53) or microdeletions (n = 4). Of the 57 patients, 56 had epilepsy: generalized in 55, with focal seizures in 7 and infantile spasms in 1. Median seizure onset age was 2 years. A novel type of drop attack was identified comprising eyelid myoclonia evolving to a myoclonic-atonic (n = 5) or atonic (n = 8) seizure. Seizure types included eyelid myoclonia with absences (65%), myoclonic seizures (34%), atypical (20%) and typical (18%) absences, and atonic seizures (14%), triggered by eating in 25%. Developmental delay preceded seizure onset in 54 of 56 (96%) patients for whom early developmental history was available. Developmental plateauing or regression occurred with seizures in 56 in the context of a developmental and epileptic encephalopathy (DEE). Fifty-five of 57 patients had intellectual disability, which was moderate to severe in 50. Other common features included behavioral problems (73%); high pain threshold (72%); eating problems, including oral aversion (68%); hypotonia (67%); sleeping problems (62%); autism spectrum disorder (54%); and ataxia or gait abnormalities (51%).ConclusionsSYNGAP1 mutations cause a generalized DEE with a distinctive syndrome combining epilepsy with eyelid myoclonia with absences and myoclonic-atonic seizures, as well as a predilection to seizures triggered by eating.

EEG endophenotypes in autism spectrum disorder

ObjectivesThe association between autism spectrum disorder (ASD) and epilepsy is well-known. Abnormalities on electroencephalography (EEG) results have been reported in patients with ASD without a history of seizures. However, little is known about the relationship between abnormalities on EEG results and the core features of ASD. The purpose of the study was to determine the relationship between the presence of epilepsy and/or abnormalities on EEG results and disease-associated impairments in young children with ASD. MethodsData were collected from medical records at Cincinnati Children's Hospital Medical Center (CCHMC) of patients with well-characterized ASD. Patients were subdivided into three groups: ASD without epilepsy but with abnormal EEG results, ASD without epilepsy and normal EEG results, and ASD with epilepsy. Developmental (Mullen Scales of Early Learning (MSEL)), adaptive (Vineland Adaptive Behavior Scales (VABS)), behavioral (Child Behavior Checklist), and language (Preschool Language Scales (PLS)) assessments, along with birth and developmental histories, medications, and medical comorbidities were collected. Electroencephalography data were abstracted from reports and included presence, characterization, and location of abnormalities. ResultsAnalysis was performed on 443 patients with ASD. Seventy patients (15.8%) had epilepsy at the time of ASD diagnosis. Out of 372 patients with ASD and no epilepsy, 95 (25.5%) had an abnormal EEG result (67.4% epileptiform, 36.8% other abnormalities). Majority of epileptiform discharges were focal (83%) and most commonly seen in the left temporal region. The group with abnormal EEG results exhibited more impaired adaptive functioning when compared with the group with normal EEG results (p < 0.05). The group with abnormal EEG results was more similar to the group with epilepsy, differing only in expressive language (p < 0.01) and fine motor (p < 0.05) skills on the Mullen Scales. The group with epilepsy exhibited lower scores in all areas of developmental and adaptive functioning compared with the group with normal EEG results (p < 0.05). At the time of analysis, 13 patients (8 in the group with abnormal EEG results, 5 in the group with normal EEG results) developed epilepsy at a mean age of 10.5 years ± 3.3 years. ConclusionsThe presence of an abnormal EEG result or epilepsy in the setting of ASD suggests worse developmental and adaptive functioning. Further analysis will help to clarify associations and offer insight into treatment for this subpopulation without epilepsy but with abnormal EEG results.

MCA Based Epilepsy EEG Classification Using Time Frequency Domain Features.

In this work, we proposed a morphological component analysis (MCA) based method for epilepsy classification using the explicit dictionary of independent redundant transforms to decomposes the electroencephalogram (EEG) by considering it's morphology. Output components of MCA are represented into analytical form by using Hilbert transform. Then features, parameter's ratio of bandwidth square, mean square frequency and fractional contributions to dominant frequency were extracted to discriminate epilepsy EEG by support vector machine (SVM). These features have shown classification results comparable to previous works.

F179. Notched delta, epilepsy and Angelman’s syndrome genetics

Introduction Angelman syndrome (AS) is a neurogenetic disorder resulting from a deletion of the maternal 15q11.2-13.1 region, mutation of maternal UBE3A gene, paternal uniparental disomy of chromosome 15, or methylation imprinting defect. Epilepsy affects the majority of the patients. Notched delta (ND) is characterized by ill-defined mixed spikes among a diffuse delta slowing, giving a “notched” appearance seen in AS. The aim of this study was to evaluate the prevalence of ND and epilepsy among the known genetic mechanisms. Methods Patients with AS seen at the Mayo Clinic Medical Genetics Clinics between September 2006 and September 2017 were identified. Medical records were reviewed for epilepsy history and EEG findings. The first 4 available EEG reports were reviewed for presence of ND. Patients were classified into 4 genetic groups: 1. microdeletion, 2.uniparental disomy (UPD), 3.UBE3A mutation, or 4.methylation defect. Patients without research consent (n = 1), inadequate epilepsy or EEG records (n = 3) were excluded. The age at seizure onset and the prevalence of ND was calculated in each genetic group. Fisher exact tests and nonparametric Wilcoxon rank sum test were used. Results Twenty-four patients were included (male 42%, median age 14.1 years, IQR 8.4–25.8). Genetic etiology was microdeletion 13, UPD 1, UBE3A mutation 4, and methylation defect 6. Epilepsy was present in 21/24 (88%) patients. There was no difference in the prevalence of epilepsy among the different genetic groups (microdeletion 12/13, UPD 1/1, UBE3A mutation 3/4, methylation defect 5/6, p > 0.05). Patients without seizure were younger than those with epilepsy (No seizure median age 7.4 years, IQR 2.6–8.4 vs Epilepsy 17.3 years, IQR 9.6–27.5, p = 0.03). Among epilepsy patients, the median age of seizure onset was 2 years, IQR 0.5–3.7. Patients with microdeletion have earlier age of seizure onset compared to the other groups: microdeletion, n = 12, median age of seizure onset 1.3 years, IQR 0.5–2.0; UPD, n = 1, age of seizure onset 3.0 years; UBE3A mutation, n = 3, median age of seizure onset 4.3 years, IQR 2.7–5.2; methylation defect, n = 5, median age of seizure onset, 3.8 years IQR 1.7–5.1; p = 0.03. ND was recognized in 15/24 (62%) patients at any time, including 8/13 chromosomal microdeletion, 1/1 UPD, 2/4 UBE3A mutation, 4/6 methylation defect. The median age at which ND was first recognized was 9.2 years, IQR 2.83–18.2. Prevalence of epilepsy was no different between patients with or without ND (13/15 patient ND and 8/9 patients without ND, p > 0.05). Conclusion Epilepsy was invariably present in the majority of the different genetic groups of AS patients, with earlier age of onset in the chromosome 15 microdeletion group. Patients with seizures were older than those without which may suggest an age related increased risk for developing epilepsy. ND was recognized in nearly two-thirds of all patients, with no specific distribution among the different genetic etiologies and did not influence the prevalence of epilepsy.

Variant Intestinal-Cell Kinase in Juvenile Myoclonic Epilepsy

BackgroundIn juvenile myoclonic epilepsy, data are limited on the genetic basis of networks promoting convulsions with diffuse polyspikes on electroencephalography (EEG) and the subtle microscopic brain dysplasia called microdysgenesis.MethodsUsing Sanger sequencing, we sequenced the exomes of six members of a large family affected with juvenile myoclonic epilepsy and confirmed cosegregation in all 37 family members. We screened an additional 310 patients with this disorder for variants on DNA melting-curve analysis and targeted real-time DNA sequencing of the gene encoding intestinal-cell kinase (ICK). We calculated Bayesian logarithm of the odds (LOD) scores for cosegregating variants, odds ratios in case–control associations, and allele frequencies in the Genome Aggregation Database. We performed functional tests of the effects of variants on mitosis, apoptosis, and radial neuroblast migration in vitro and conducted video-EEG studies in mice lacking a copy of Ick.ResultsA variant, K305T (c.914A→C), cosegregated with epilepsy or polyspikes on EEG in 12 members of the family affected with juvenile myoclonic epilepsy. We identified 21 pathogenic ICK variants in 22 of 310 additional patients (7%). Four strongly linked variants (K220E, K305T, A615T, and R632X) impaired mitosis, cell-cycle exit, and radial neuroblast migration while promoting apoptosis. Tonic–clonic convulsions and polyspikes on EEG resembling seizures in human juvenile myoclonic epilepsy occurred more often in knockout heterozygous mice than in wild-type mice (P=0.02) during light sleep with isoflurane anesthesia.ConclusionsOur data provide evidence that heterozygous variants in ICK caused juvenile myoclonic epilepsy in 7% of the patients included in our analysis. Variant ICK affects cell processes that help explain microdysgenesis and polyspike networks observed on EEG in juvenile myoclonic epilepsy. (Funded by the National Institutes of Health and others.)

Effects of Levetiracetam and Sulthiame on EEG in benign epilepsy with centrotemporal spikes: A randomized controlled trial

PurposeBECTS (benign childhood epilepsy with centrotemporal spikes) is associated with characteristic EEG findings. This study examines the influence of anti-convulsive treatment on the EEG. MethodsIn a randomized controlled trial including 43 children with BECTS, EEGs were performed prior to treatment with either Sulthiame or Levetiracetam as well as three times under treatment. Using the spike-wave-index, the degree of EEG pathology was quantified. The EEG before and after initiation of treatment was analyzed. Both treatment arms were compared and the EEG of the children that were to develop recurrent seizures was compared with those that were successfully treated. ResultsRegardless of the treatment agent, the spike-wave-index was reduced significantly under treatment. There were no differences between the two treatment groups. In an additional analysis, the EEG characteristics of the children with recurrent seizures differed statistically significant from those that did not have any further seizures. ConclusionBoth Sulthiame and Levetiracetam influence the EEG of children with BECTS. Persistent EEG pathologies are associated with treatment failures.

Deep Convolution Neural Network and Autoencoders-Based Unsupervised Feature Learning of EEG Signals

Epilepsy is a health problem that seriously affects the quality of humans for many years. Therefore, it is important to accurately analyze and recognize epilepsy based on EEG signals, and for a long time, researchers have attempted to extract new features from the signals for epilepsy recognition. However, it is very difficult to select useful features from a large number of them in this diagnostic application. As the development of artificial intelligence progresses, unsupervised feature learning based on the deep learning model can obtain features that can better describe identified objects from unlabeled data. In this paper, the deep convolution network and autoencoders-based model, named as AE-CDNN, is constructed in order to perform unsupervised feature learning from EEG in epilepsy. We extract features by AE-CDNN model and classify the features based on two public EEG data sets. Experimental results showed that the classification results of features obtained by AE-CDNN are more optimal than features obtained by principal component analysis and sparse random projection. Using several common classifiers to classify features obtained by AE-CDNN model results in high accuracy and not inferior to the research results from most recent studies. The results also showed that the features of AE-CDNN model are clear, effective, and easy to learn. These features can speed up the convergence and reduce the training times of classifiers. Therefore, the AE-CDNN model can be effectively applied to feature extraction of EEG in epilepsy.

Epilepsy and autism spectrum disorders in children

The problem of epilepsy comorbidity with autism spectrum disorders in children is discussed. The incidence data of autism spectrum disorders in epilepsy, epileptiform discharges on the EEG in autism spectrum disorders and epilepsy in autism spectrum disorders are reviewed. The following types of epilepsy and autism co-occurrence are discussed: both conditions are independent, have different causes and may co-occur by chance; epilepsy and autism are associated, both being independent consequences of the same genetic disorder or early cerebral damage; autism is caused by the epileptic process which interferes with the function of specific brain networks involved in the development of communication and social behavior; autism is a result of the withdrawal reaction in the epileptic child. The known genetic causes of epilepsy and autism spectrum disorders comorbidity are provided. The practical issues are discussed, in particular the rational indication of antiepileptic drugs to the children suffering autism spectrum disorders.

MRI Brain Findings in Adults with Lesional Refractory Epilepsy and Correlation to Surgical Outcome

Introduction: Refractory epilepsy which is defined as failure of adequate trials of two tolerated, appropriately chosen and used antiepileptic drug schedules (whether as monotherapies or in combination) to achieve sustained seizure freedom is common in patients with structural brain lesions including acquired disorders and genetic abnormalities. High resolution Magnetic Resonance Imaging MRI of the brain has proven its precision as a diagnostic tool for recognition of different structural lesions underlying medically intractable seizures. Objective: To recognize common MRI lesions in a series of adult patients with refractory epilepsy admitted to the epilepsy monitoring unit at Prince Sultan Military Medical City PSMMC for pre surgical evaluation for epilepsy surgery with correlation to surgical outcome and to compare our local data with the international literature. Material and methods: 245 patients (100 Males and 145 Females; 14-53 years) with refractory epilepsy were included in this retrospective analysis. They presented with partial with or without complex partial seizures 112 (46.0%), partial with secondary generalized tonic clonic seizures 76 (31.0%) or generalized seizures in 57 (23.0%) of patients. Clinical diagnosis of epilepsy and seizure classification were based on the revised criteria of the International League Against Epilepsy ILAE. Structural neuroimaging MRI brain, functional neuroimaging which include Interictal Fluorodeoxyglucose Positron Emission Computed Tomography FDG-PET, Ictal Technetium-99m hexamethyl-propylene amine oxime Single Photon Emission Computed Tomography 99 m HMPAO SPECT, Electroencephalography EEG recording, epilepsy history and neurological examination were performed. MRI brain imaging epilepsy protocol used a 1.5 or 3 Tesla MRI scanner. All patients included in this study received appropriate epilepsy surgery and post-operative seizure control was followed in the epilepsy clinic with six-month post-operative inter ictal EEG, follow up MRI brain after epilepsy surgery were performed in all patients and 50 patients had additional video-EEG recording postoperative during the follow up period. Epilepsy surgery seizure control outcome was classified according to Engel Classification system. All patients were followed for at least two years post-operatively to assess seizure control. Pre-operative MRI diagnosis was correlated with the epilepsy surgery seizure control outcome. Results: MRI detected different structural brain abnormalities in 245 (100%) patients, including temporal lobe location in 142 (58%) patients, frontal lobe location in 74 (30%) patients and parieto-occipital lobes location in 29 (12%) patients. On MRI hippocampal sclerosis HS is diagnosed in 86 (35%) patients, cerebral tumors in 74 (30%) patients and among the cerebral tumors MRI suggested the diagnosis of developmental tumors that is; glio-neural tumors in 45 out of 74 (61%) of tumor patients, malformations of cortical development MCD in 42 (17%) patients, vascular malformations in 15 (6%) patients, dual pathologies in 12 (5%) patients and remote gliotic lesions in 16 (7%) patients. The histopathological diagnosis confirmed the MRI brain diagnosis in all included patients. At 2 years of post-operative follow-up 196 (80%) patients were classified as Engel Class I, 37 (15%) patients Engel Class II, 6 (3%) patients Engel Class III and 6 (3%) patients as Class IV seizure freedom. In the class IV group two patients failed epilepsy surgery because they had high grade astrocytoma with postoperative tumor recurrence and four patients failed epilepsy surgery due to incomplete resection of epileptogenic lesions. Patients with MRI diagnosis of HS 82 out of 86 (95%) patients, low grade tumor 59 out of 74 (80%) patients and vascular malformation 12 out of 15 (80%) had best epilepsy surgery outcome. Patients with high grade tumors 2 out 3 (67%) patients and incomplete surgical resection of epileptogenic lesions 4 out of 5 (80%) patients had the worst epilepsy surgery outcome. Conclusion: This study revealed that MRI brain structural lesions were commonly associated with refractory epilepsy. Temporal lobe structural brain lesions were most common lesions in adult epilepsy patients with refractory epilepsy referred for epilepsy surgery. The presence of HS, low grade tumors and vascular malformations correlated with best surgical outcome while high grade tumors and incomplete surgical resection of the lesions correlated with worst surgical outcome. Our results were consistent with the international reported literature.

Epileptic seizure detection in EEG signal using machine learning techniques.

Epilepsy is a well-known nervous system disorder characterized by seizures. Electroencephalograms (EEGs), which capture brain neural activity, can detect epilepsy. Traditional methods for analyzing an EEG signal for epileptic seizure detection are time-consuming. Recently, several automated seizure detection frameworks using machine learning technique have been proposed to replace these traditional methods. The two basic steps involved in machine learning are feature extraction and classification. Feature extraction reduces the input pattern space by keeping informative features and the classifier assigns the appropriate class label. In this paper, we propose two effective approaches involving subpattern based PCA (SpPCA) and cross-subpattern correlation-based PCA (SubXPCA) with Support Vector Machine (SVM) for automated seizure detection in EEG signals. Feature extraction was performed using SpPCA and SubXPCA. Both techniques explore the subpattern correlation of EEG signals, which helps in decision-making process. SVM is used for classification of seizure and non-seizure EEG signals. The SVM was trained with radial basis kernel. All the experiments have been carried out on the benchmark epilepsy EEG dataset. The entire dataset consists of 500 EEG signals recorded under different scenarios. Seven different experimental cases for classification have been conducted. The classification accuracy was evaluated using tenfold cross validation. The classification results of the proposed approaches have been compared with the results of some of existing techniques proposed in the literature to establish the claim.

Cyclic alternating pattern and interictal epileptiform discharges during morning sleep deprived EEG in temporal lobe epilepsy

Cyclic alternating pattern and interictal epileptiform discharges during morning sleep deprived EEG in temporal lobe epilepsy

Designed Filter with CCII+ and Analysis of EEG for Epilepsy and Alzheimer

Designed Filter with CCII+ and Analysis of EEG for Epilepsy and Alzheimer

Potentially Pathological Alpha-Pattern as a Variant of Vigilance EEG in Drug-Resistant Epilepsy

As a result of pathomorphosis affecting the mechanisms of electrical activity generation interictal EEG may show reduced epileptiform changes whereas clinically apparent epileptic seizures may be present. In these cases patterns of dominant alpha activity are sometimes recorded on the scalp. In this study variations of alpha activity in patients with refractory epilepsy are classified. A group of 50 refractory epilepsy patients aged between 20 and 55 years who were submitted to Polenov Russian Scientific Research Institute of Neurosurgery in 2014-2017 was included in this study. They underwent scalp EEG as a part of their presurgical assessment. In 12 cases patterns of potentially pathological alpha activity were observed. Three variations of alpha-patterns were described: 1) alpha-rhythm with decreased regional diversity and a marked synchronization in temporal areas; 2) alpha-rhythm with reduced epileptiform complexes integrated into the spindles, 3) decelerated non-rhythmic alpha activity distorted by the higher frequency components. Distinguished varieties of potentially pathological alpha-activity according to their order here represent gradual functional decline of normal thalamo-cortical interaction. Considering clinical manifestation of drug-resistant epilepsy with frequent seizures in these patients, reported varieties of alpha activity can not be interpreted as Landolt’s syndrome (forced normalization of EEG). Invasive electrocorticographic monitoring demonstrated that bursts of sharpened polyphasic waves coinciding with alpha-rhythm on scalp EEG are consistent with epileptic discharges on the brain cortex surface. This allows to think of these components as correlates of epileptic activity. Therefore, on a number of occasions in patients with epilepsy a dissonance between clinical signs and electroencephalographic patterns recorded during restful wakefulness may be observed, when epileptiform components are absent or reduced to nonspecific complexes.

Dynamic coupling between fMRI local connectivity and interictal EEG in focal epilepsy: A wavelet analysis approach.

Simultaneous scalp EEG-fMRI recording is a noninvasive neuroimaging technique for combining electrophysiological and hemodynamic aspects of brain function. Despite the time-varying nature of both measurements, their relationship is usually considered as time-invariant. The aim of this study was to detect direct associations between scalp-recorded EEG and regional changes of hemodynamic brain connectivity in focal epilepsy through a time-frequency paradigm. To do so, we developed a voxel-wise framework that analyses wavelet coherence between dynamic regional phase synchrony (DRePS, calculated from fMRI) and band amplitude fluctuation (BAF) of a target EEG electrode with dominant interictal epileptiform discharges (IEDs). As a proof of concept, we applied this framework to seven patients with focal epilepsy. The analysis produced patient-specific spatial maps of DRePS-BAF coupling, which highlight regions with a strong link between EEG power and local fMRI connectivity. Although we observed DRePS-BAF coupling proximate to the suspected seizure onset zone in some patients, our results suggest that DRePS-BAF is more likely to identify wider 'epileptic networks'. We also compared DRePS-BAF with standard EEG-fMRI analysis based on general linear modelling (GLM). There was, in general, little overlap between the DRePS-BAF maps and GLM maps. However, in some subjects the spatial clusters revealed by these two analyses appeared to be adjacent, particularly in medial posterior cortices. Our findings suggest that (1) there is a strong time-varying relationship between local fMRI connectivity and interictal EEG power in focal epilepsy, and (2) that DRePS-BAF reflect different aspects of epileptic network activity than standard EEG-fMRI analysis. These two techniques, therefore, appear to be complementary. Hum Brain Mapp 38:5356-5374, 2017. © 2017 Wiley Periodicals, Inc.

Cyclic alternating pattern and interictal epileptiform discharges during morning sleep after sleep deprivation in temporal lobe epilepsy.

Sleep deprivation (SD) increases the occurrence of interictal epileptiform discharges (IED) compared to basal EEG in temporal lobe epilepsy (TLE). In adults, EEG after SD is usually performed in the morning after SD. We aimed to evaluate whether morning sleep after SD bears additional IED-inducing effects compared with nocturnal physiological sleep, and whether changes in sleep stability (described by the cyclic alternating pattern-CAP) play a significant role. Adult patients with TLE underwent in-lab night polysomnography (n-PSG) and, within 7days from n-PSG, they underwent also a morning EEG after night SD (SD-EEG). We included only TLE patients in which both recordings showed IED. SD-EEG consisted of waking up patients at 2:00 AM and performing video EEG at 8:00 AM. For both recordings, we obtained the following markers for the first sleep cycle: IED/h (Spike Index, SI), sleep macrostructure, microstructure (NREM CAP rate; A1, A2 and A3 Indices), and SI association with CAP variables. The macrostructure of the first sleep cycle was similar in n-PSG and morning SD-EEG, whereas CAP rate and SI were significantly higher in SD-EEG. SI increase was selectively associated with CAP phases. SD increases the instability of morning recovery sleep compared with n-PSG, and particularly enhances CAP A1 phases, which are associated with the majority of IED. Thus, higher instability of morning recovery sleep may account at least in part for the increased IED yield in SD-EEG in TLE patients.