Morphology-based automatic seizure detector for intracerebral EEG recordings

Identifying targets for preventing epilepsy using systems biology of the human brain

Approximately one third of all epilepsy patients are resistant to current therapeutic treatments. Some patients with focal forms of epilepsy benefit from invasive surgical approaches that can lead to large surgical resections of human epileptic neocortex. We have developed a systems biology approach to take full advantage of these resections and the brain tissues they generate as a means to understand underlying mechanisms of neocortical epilepsy and to identify novel biomarkers and therapeutic targets. In this review, we will describe our unique approach that has led to the development of a 'NeuroRepository' of electrically-mapped epileptic tissues and associated data. This 'Big Data' approach links quantitative measures of ictal and interictal activities corresponding to a specific intracranial electrode to clinical, imaging, histological, genomic, proteomic, and metabolomic measures. This highly characterized data and tissue bank has given us an extraordinary opportunity to explore the underlying electrical, cellular, and molecular mechanisms of the human epileptic brain. We describe specific examples of how an experimental design that compares multiple cortical regions with different electrical activities has led to discoveries of layer-specific pathways and how these can be 'reverse translated' from animal models back to humans in the form of new biomarkers and therapeutic targets. This article is part of the special issue entitled 'New Epilepsy Therapies for the 21st Century - From Antiseizure Drugs to Prevention, Modification and Cure of Epilepsy'.

Automatic Seizure Detection Based on Morphological Features Using One-Dimensional Local Binary Pattern on Long-Term EEG

Epileptic neurological disorder of the brain is widely diagnosed using the electroencephalography (EEG) technique. EEG signals are nonstationary in nature and show abnormal neural activity during the ictal period. Seizures can be identified by analyzing and obtaining features of EEG signal that can detect these abnormal activities. The present work proposes a novel morphological feature extraction technique based on the local binary pattern (LBP) operator. LBP provides a unique decimal value to a sample point by weighing the binary outcomes after thresholding the neighboring samples with the present sample point. These LBP values assist in capturing the rising and falling edges of the EEG signal, thus providing a morphologically featured discriminating pattern for epilepsy detection. In the present work, the variability in the LBP values is measured by calculating the sum of absolute difference of the consecutive LBP values. Interquartile range is calculated over the preprocessed EEG signal to provide dispersion measure in the signal. For classification purpose, K-nearest neighbor classifier is used, and the performance is evaluated on 896.9 hours of data from CHB-MIT continuous EEG database. Mean accuracy of 99.7% and mean specificity of 99.8% is obtained with average false detection rate of 0.47/h and sensitivity of 99.2% for 136 seizures.

Exploring human epileptic activity at the single-neuron level

Today, localization of the seizure focus heavily relies on EEG monitoring (scalp or intracranial). However, current technology enables much finer resolutions. The activity of hundreds of single neurons in the human brain can now be simultaneously explored before, during, and after a seizure or in association with an interictal discharge. This technology opens up new horizons to understanding epilepsy at a completely new level. This review therefore begins with a brief description of the basis of the technology, the microelectrodes, and the setup for their implantation in patients with epilepsy. Using these electrodes, recent studies provide novel insights into both the time domain and firing patterns of epileptic activity of single neurons. In the time domain, seizure-related activity may occur even minutes before seizure onset (in its current, EEG-based definition). Seizure-related neuronal interactions exhibit complex heterogeneous dynamics. In the seizure-onset zone, changes in firing patterns correlate with cell loss; in the penumbra, neurons maintain their spike stereotypy during a seizure. Hence, investigation of the extracellular electrical activity is expected to provide a better understanding of the mechanisms underlying the disease; it may, in the future, serve for a more accurate localization of the seizure focus; and it may also be employed to predict the occurrence of seizures prior to their behavioral manifestation in order to administer automatic therapeutic interventions.

Robust deep network with maximum correntropy criterion for seizure detection

Effective seizure detection from long-term EEG is highly important for seizure diagnosis. Existing methods usually design the feature and classifier individually, while little work has been done for the simultaneous optimization of the two parts. This work proposes a deep network to jointly learn a feature and a classifier so that they could help each other to make the whole system optimal. To deal with the challenge of the impulsive noises and outliers caused by EMG artifacts in EEG signals, we formulate a robust stacked autoencoder (R-SAE) as a part of the network to learn an effective feature. In R-SAE, the maximum correntropy criterion (MCC) is proposed to reduce the effect of noise/outliers. Unlike the mean square error (MSE), the output of the new kernel MCC increases more slowly than that of MSE when the input goes away from the center. Thus, the effect of those noises/outliers positioned far away from the center can be suppressed. The proposed method is evaluated on six patients of 33.6 hours of scalp EEG data. Our method achieves a sensitivity of 100% and a specificity of 99%, which is promising for clinical applications.

Ngram-derived pattern recognition for the detection and prediction of epileptic seizures

This work presents a new method that combines symbol dynamics methodologies with an Ngram algorithm for the detection and prediction of epileptic seizures. The presented approach specifically applies Ngram-based pattern recognition, after data pre-processing, with similarity metrics, including the Hamming distance and Needlman-Wunsch algorithm, for identifying unique patterns within epochs of time. Pattern counts within each epoch are used as measures to determine seizure detection and prediction markers. Using 623 hours of intracranial electrocorticogram recordings from 21 patients containing a total of 87 seizures, the sensitivity and false prediction/detection rates of this method are quantified. Results are quantified using individual seizures within each case for training of thresholds and prediction time windows. The statistical significance of the predictive power is further investigated. We show that the method presented herein, has significant predictive power in up to 100% of temporal lobe cases, with sensitivities of up to 70-100% and low false predictions (dependant on training procedure). The cases of highest false predictions are found in the frontal origin with 0.31-0.61 false predictions per hour and with significance in 18 out of 21 cases. On average, a prediction sensitivity of 93.81% and false prediction rate of approximately 0.06 false predictions per hour are achieved in the best case scenario. This compares to previous work utilising the same data set that has shown sensitivities of up to 40-50% for a false prediction rate of less than 0.15/hour.

Noninvasive imaging of the high frequency brain activity in focal epilepsy patients

High-frequency (HF) activity represents a potential biomarker of the epileptogenic zone in epilepsy patients, the removal of which is considered to be crucial for seizure-free surgical outcome. We proposed a high frequency source imaging (HFSI) approach to noninvasively image the brain sources of the scalp-recorded HF EEG activity. Both computer simulation and clinical patient data analysis were performed to investigate the feasibility of using the HFSI approach to image the sources of HF activity from noninvasive scalp EEG recordings. The HF activity was identified from high-density scalp recordings after high-pass filtering the EEG data and the EEG segments with HF activity were concatenated together to form repetitive HF activity. Independent component analysis was utilized to extract the components corresponding to the HF activity. Noninvasive EEG source imaging using realistic geometric boundary element head modeling was then applied to image the sources of the pathological HF brain activity. Five medically intractable focal epilepsy patients were studied and the estimated sources were found to be concordant with the surgical resection or intracranial recordings of the patients. The present study demonstrates, for the first time, that source imaging from the scalp HF activity could help to localize the seizure onset zone and provide a novel noninvasive way of studying the epileptic brain in humans. This study also indicates the potential application of studying HF activity in the presurgical planning of medically intractable epilepsy patients.