Epilepsy Prediction

Patients suffering from epileptic seizures are usually treated with medication and/or surgical procedures. However, in more than 30% of cases, medication or surgery does not effectively control seizure activity. A method that predicts the onset of a seizure before it occurs may prove useful as patients might be alerted to make themselves safe or seizures could be prevented with therapeutic interventions just before they occur. Abnormal neuronal activity, the preictal state, starts a few minutes before the onset of a seizure. In recent years, different methods have been proposed to predict the start of the preictal state. These studies follow some common steps, including recording of EEG signals, preprocessing, feature extraction, classification, and postprocessing. However, online prediction of epileptic seizures remains a challenge as all these steps need further refinement to achieve high sensitivity and low false positive rate. In this paper, we present a comparison of state-of-the-art methods used to predict seizures using both scalp and intracranial EEG signals and suggest improvements to existing methods.

Patients suffering from epileptic seizures are usually treated with medication and/or surgical procedures. However, in more than 30% of cases, medication or surgery does not effectively control seizure activity. A method that predicts the onset of a seizure before it occurs may prove useful as patients might be alerted to make themselves safe or seizures could be prevented with therapeutic interventions just before they occur. Abnormal neuronal activity, the preictal state, starts a few minutes before the onset of a seizure. In recent years, different methods have been proposed to predict the start of the preictal state. These studies follow some common steps, including recording of EEG signals, preprocessing, feature extraction, classification, and postprocessing. However, online prediction of epileptic seizures remains a challenge as all these steps need further refinement to achieve high sensitivity and low false positive rate. In this paper, we present a comparison of state-of-the-art methods used to predict seizures using both scalp and intracranial EEG signals and suggest improvements to existing methods.

Epilepsy is the most prevalent neurological disorder in humans. Epilepsy is characterized by recurrent seizures .This happens due to an abnormality in brain wiring, an imbalance of nerve signaling chemicals called neurotransmitters, or some combination of these factors. Normally, neurons generate electrochemical impulses, act on other neurons, glands, and muscles to generate human thoughts, feelings and actions. Epilepsy is determined by EEG signal recording, which contain valuable information for understanding epilepsy. Epilepsy can create clear disturbance and leaves its signature on standard EEG signals. The detection of seizures occurring in the EEGs is an important component for the diagnosis and treatment of epilepsy. In recent years soft computing based techniques are attracting much attention from the scientific communities as an alternative tool. This technique for detection and classification of epilepsy has been reported by number of researchers. It is an important part o...

Patients suffering from epileptic seizures are usually treated with medication and/or surgical procedures. However, in more than 30% of cases, medication or surgery does not effectively control seizure activity. A method that predicts the onset of a seizure before it occurs may prove useful as patients might be alerted to make themselves safe or seizures could be prevented with therapeutic interventions just before they occur. Abnormal neuronal activity, the preictal state, starts a few minutes before the onset of a seizure. In recent years, different methods have been proposed to predict the start of the preictal state. These studies follow some common steps, including recording of EEG signals, preprocessing, feature extraction, classification, and postprocessing. However, online prediction of epileptic seizures remains a challenge as all these steps need further refinement to achieve high sensitivity and low false positive rate. In this paper, we present a comparison of state-of-the-art methods used to predict seizures using both scalp and intracranial EEG signals and suggest improvements to existing methods.

Seizure is the main factor that affects the quality of life of epileptic patients. The diagnosis of epilepsy, and hence the identification of epileptogenic zone, is commonly made by using continuous Electroencephalogram (EEG) signal monitoring. Seizure identification on EEG signals is made manually by epileptologists and this process is usually very long and error prone. The aim of this paper is to describe an automated method able to detect seizures in EEG signals, using knowledge discovery in database process and data mining methods and algorithms, which can support physicians during the seizure detection process. Our detection method is based on Artificial Neural Network classifier, trained by applying the multilayer perceptron algorithm, and by using a software application, called Training Builder that has been developed for the massive extraction of features from EEG signals. This tool is able to cover all the data preparation steps ranging from signal processing to data analys...

Epilepsy, simply put, is an abnormality in the central nervous system that leads to unplanned-for seizures affecting millions of people worldwide. Medication is the most common treatment for all those suffering from epilepsy, however, this paper introduces the idea of designing an implantable/embedded chip that is to be fed with a machine learning algorithm, specifically Support Vector Machine (SVM) to predict seizure periods prior to their occurrence to be able to notify the patient or suppress the seizure from happening. Since the system's problem is binary classification between pre-ictal and normal periods, determining the right set of features along with the SVM kernel function is the first step in designing the chip. This paper proposes two tracks, a Linear Features-Linear Kernel combination with a set of time-domain hardwareinexpensive features namely; Coastline, Absolute Mean, Root Mean Square, and Standard Deviation. In addition to the Non-Linear Feature-RBF Kernel combination with another timedomain feature namely; Hurst Exponent. This paper also introduces the concept of segmenting the testing data which showed extremely promising results in the Non-Linear Feature-RBF Kernel combination.

Epilepsy is a common brain disorder that causes patients to face multiple seizures in a single day. Around 65 million people are affected by epilepsy worldwide. Patients with focal epilepsy can be treated with surgery, whereas generalized epileptic seizures can be managed with medications. It has been noted that in more than 30% of cases, these medications fail to control epileptic seizures, resulting in accidents and limiting the patient’s life. Predicting epileptic seizures in such patients prior to the commencement of an oncoming seizure is critical so that the seizure can be treated with preventive medicines before it occurs. Electroencephalogram (EEG) signals of patients recorded to observe brain electrical activity during a seizure can be quite helpful in predicting seizures. Researchers have proposed methods that use machine and/or deep learning techniques to predict epileptic seizures using scalp EEG signals; however, prediction of seizures with increased accuracy is still a...

Patients suffering from epileptic seizures are usually treated with medication and/or surgical procedures. However, in more than 30% of cases, medication or surgery does not effectively control seizure activity. A method that predicts the onset of a seizure before it occurs may prove useful as patients might be alerted to make themselves safe or seizures could be prevented with therapeutic interventions just before they occur. Abnormal neuronal activity, the preictal state, starts a few minutes before the onset of a seizure. In recent years, different methods have been proposed to predict the start of the preictal state. These studies follow some common steps, including recording of EEG signals, preprocessing, feature extraction, classification, and postprocessing. However, online prediction of epileptic seizures remains a challenge as all these steps need further refinement to achieve high sensitivity and low false positive rate. In this paper, we present a comparison of state-of-the-art methods used to predict seizures using both scalp and intracranial EEG signals and suggest improvements to existing methods.

Background: The development of automated seizure detection methods using EEG signals could be of great importance for the diagnosis and the monitoring of patients with epilepsy. These methods are often patient-specific and require high accuracy in detecting seizures but also very low false-positive rates. The aim of this study is to evaluate the performance of a seizure detection method using EEG signals by investigating its performance in correctly identifying seizures and in minimizing false alarms and to determine if it is generalizable to different patients. Methods: We tested the method on about two hours of preictal/ictal and about ten hours of interictal EEG recordings of one patient from the Freiburg Seizure Prediction EEG database using machine learning techniques for data mining. Then, we tested the obtained model on six other patients of the same database. Results: The method achieved very high performance in detecting seizures (close to 100% of correctly classified posit...

Background: The development of automated seizure detection methods using EEG signals could be of great importance for the diagnosis and the monitoring of patients with epilepsy. These methods are often patient-specific and require high accuracy in detecting seizures but also very low false-positive rates. The aim of this study is to evaluate the performance of a seizure detection method using EEG signals by investigating its performance in correctly identifying seizures and in minimizing false alarms and to determine if it is generalizable to different patients. Methods: We tested the method on about two hours of preictal/ictal and about ten hours of interictal EEG recordings of one patient from the Freiburg Seizure Prediction EEG database using machine learning techniques for data mining. Then, we tested the obtained model on six other patients of the same database. Results: The method achieved very high performance in detecting seizures (close to 100% of correctly classified posit...

Epilepsy is a group of neurological disorders identifiable by infrequent but recurrent seizures. Seizure prediction is widely recognized as a significant problem in the neuroscience domain. Developing a Brain-Computer Interface (BCI) for seizure prediction can provide an alert to the patient, providing a buffer time to get the necessary emergency medication or at least be able to call for help, thus improving the quality of life of the patients. A considerable number of clinical studies presented evidence of symptoms (patterns) before seizure episodes and thus, there is large research on seizure prediction, however, there is very little existing literature that illustrates the use of structured processes in machine learning for predicting seizures. Limited training data and class imbalance (EEG segments corresponding to preictal phase, the duration just before the seizure, to about an hour prior to the episode, are usually in a tiny minority) are a few challenges that need to be add...

Epilepsy is a neurological disorder which is characterized by repeated seizures. Although these can be recorded using Electroencephalogram (EEG), it is very difficult task to classify the different states such as normal, preictal and ictal by visual inspection due to the non-linear and non stationary property of the EEG signals and also due to its long term recordings. Thus automatic detection of these states using various techniques has been in the literature for a very long time. We used novel features such as permutation entropy (PE), Tsallis entropy (TsE), and Kolmogorov complexity (KC) along with five different classifiers. We achieved a highest accuracy of 89.33% with Decision tree classifier. Our method being very simple and has fast computation time in comparison with other features in the literature and thus can form as a software tool that can be installed easily and also opens future opportunities towards real time detection and prediction of epileptic seizures.

Analysis on Electroencephalogram (EEG) signal can provide important information related to the clinical pathology of epilepsy. Detecting the onset, prediction and type of seizures based on EEG signals is very important to determine an appropriate treatment for the patients. However, EEGs have the high complexity with non-linear and non-stationary characteristics; hence, an analysis will be very difficult to do through a visual inspection. Signal processing applications are, therefore, needed to make the interpretation easier. In this study, we proposed a method for EEG analysis based on signal complexity for the epileptic EEG classification. The Renyi entropy was used to extract the data of EEG features, which consist of seizure, interictal and normal features. Then, these features became the input to a classification algorithm. SVM (Support vector machine) classifier was applied to determine the type of that epileptic EEG signal and achieved accuracy of 85 %. This study can be a re...

One instrument to record the activity of brainwave in a specific time is called Electroencephalography (EEG). EEG signal can be used to analyze the epilepsy disease. Brainwave of seizure patient has a low frequency with a tighter pattern than brainwave of normal people. We use data from Temple University Hospital Seizure Corpus (TUSZ) that represents an accurate clinical condition characterization. Based on neurologist report, several types of seizure can be found in the dataset. In this research, we classify three types of seizure, Generalized Non-Specific Seizure (GNSZ), Focal Non-Specific Seizure (FNSZ) and Tonic-Clonic Seizure (TCSZ). We added a normal EEG signal, so we have four classes to be classified using Support Vector Machine (SVM). The training dataset consists from 120 data (20 GNSZ, 50 FNSZ, 25 TCSZ and 25 Normal), while the evaluation dataset is 90 datasets (20 GNSZ, 50 FNSZ, 5 TCSZ and 15 Normal). We observe the combination of three feature extraction method, Mel Fre...

Brain big data empowered by intelligent analysis provide an unrivalled opportunity to probe the dynamics of the brain in disorder. A typical example is to identify evolving synchronization patterns from multivariate electroencephalography (EEG) routinely superimposed with intensive noise in epilepsy research and practice. Under the circumstance of insufficient a priori knowledge of subject dependency on domain problem, it becomes even more important to adaptively classify the synchronization dynamics to accurately characterize the intrinsic nature of seizure activities represented by the EEG. This paper first measures the global maximal information coefficient (MIC) of all EEG data channels to form a time sequence of correlation matrices. A lightweight VGGNet (Visual Geometry Group) is designed to adapt to the need to prune massive EEG datasets. The VGGNet characterizes the synchronization dynamics captured in the correlation matrices and then automatically identifies the seizure states of the EEG. Experiments are performed over the Children's Hospital Boston-Massachusetts Institute of Technology (CHB-MIT) scalp EEG dataset to evaluate the proposed approach. Seizure states can be identified with an accuracy, sensitivity, and specificity of [98.13% ± 0.24%], [98.85% ± 0.51%], and [97.47% ± 0.36%], respectively; the resulting performance is superior to those of most existing methods over the same dataset. The approach directly applies to raw EEG analysis, which holds great potential for handling brain big data.

We present a machine learning framework aimed at increasing performance seizure detection systems. Despite the diversity of ML-based available, the methodology to select an optimal classifier or ensemble model for seizure research is not commonly known. Our study is aimed at bridging this gap by showing use of a statistically guided machine learning framework that has delivered reliable on-field performance. We empirically examined performances of 11 different classification algorithms, with additional 13 variants, executing 300 machine learning models against 6 fixed preictal windows (ranging from 5 minutes to 60 minutes) and 3 different sets of features as part of our study. Using a base of 55 subjects, of which we used 43 subjects to train and validate multiple ensemble models and classifiers, we achieved upto 98.3% F1 score, with 94.8% chance level performance on our data, and a False Positive Rate (FPR) of 0.0156/hr. We applied the top ensembles and classifier variants on 12 co...

Received: 12 December 2020 Accepted: 8 February 2021 Epilepsy is the most common form of neurological disease. Patients with epilepsy may experience seizures of a certain duration with or without provocation. Epilepsy analysis can be done with an electroencephalogram (EEG) examination. Observation of qualitative EEG signals generates high cost and often confuses due to the nature of the non-linear EEG signal and noise. In this study, we proposed an EEG signal processing system for EEG seizure detection. The signal dynamics approach to normal and seizure signals' characterization became the main focus of this study. Spectral Entropy (SpecEn) and fractal analysis are used to estimate the EEG signal dynamics and used as feature sets. The proposed method is validated using a public EEG dataset, which included preictal, ictal, and interictal stages using the Naïve Bayes classifier. The test results showed that the proposed method is able to generate an ictal detection accuracy of up ...

Epilepsy is a chronic neurological disease characterized by a large electrical explosion that is excessive and uncontrolled, as defined by the world health organization. It is an anomaly that affects people of all ages. An electroencephalogram (EEG) of the brain activity is a widely known method designed as a reference dedicated to study epileptic seizures and to record the changes in brain electrical activity. Therefore, the prediction and early detection of epilepsy is necessary to provide timely preventive interventions that allow patients to be relieved from the harmful consequences of epileptic seizures. Despite decades of research, the prediction of these seizures with accuracy remains an unresolved problem. In this article, we have proposed five deep learning models on intracranial electroencephalogram (iEEG) datasets with the aim of automatically predicting epileptic seizures. The proposed models are based on the Convolutional Neural Network (CNN) model, the fusion of the tw...

A Vehicular Ad-hoc Network (VANET) is considered as an advanced network mechanism of wireless vehicle nodes. By dynamic changeable topologies and symmetric links of network, high challenges van be evaluated by VANET protocols. The vehicles can send effective information of traffic and road status. Proper and appreciating routing mechanism is necessary for the configuration of nodes in VANET. Many network simulators have intrinsic features to characterize it from others. To make communication within the network with different routing protocols, the most ordinary simulator tool is NS 2.34. In this research work, three types of routing protocols namely on demand routing protocol (AODV, AOMDV, DSR), proactive routing protocol (DSDV) and geographical routing protocol (GPSR) have been experimented by altering values of number of nodes and speed of the nodes. Different performance metrics namely average end to end delay, average throughput, normalized routing load, packet delivery ratio ha...

Abstract— Temporal seizures due to Hippocampal origins are very common amongst epileptic patients. This article presents a novel seizure prediction approach based on a combination of graph and Chaos theory. The early identification of seizure signature allows for various preventive measures to be undertaken. The proposed non-linear approach consists of observing a high correlation level between any pair of electrodes for the lower frequencies and a decrease in the Lyapunov index (chaos or entropy) for the higher frequencies. The Fast Fourier Transform (FFT) and statistical analysis tools were used to determine threshold levels for the lower frequencies. A graph topology characterizes seizure signatures for each patient. These graphs are used to monitor both Delta and Gamma behaviors. In order to validate the proposed approach, six patients from both sexes and various age groups with temporal epilepsies originating from the hippocampal area were studied using the Freiburg Database. A...

Electroencephalogram (EEG), a record of electrical signal to represent the human brain activity, has great potential for the diagnosis to treatment of mental disorder and brain diseases such as epileptic seizure. Features extraction and classification of EEG signals is the crucial task to detect the stage of ictal (i.e., seizure period) and interictal (i.e., period between seizures) signals for the treatment and precaution of the epileptic patient. However, existing seizure and nonseizure feature extraction techniques are not good enough for the classification of ictal and interictal EEG signals considering their non-abruptness phenomena and inconsistency in different brain locations. In this paper, we present a new approach for feature extraction and classification by exploiting temporal correlation within an EEG signal for better seizure detection as any abruptness in the temporal correlation within a signal represents the transition of a phenomenon. In the proposed methods we divide an EEG signal into a number of epochs and arrange them into two-dimensional matrix and then apply different transformation/decomposition to extract a number of statistical features. These features are then used as an input to least square support vector machine to classify ictal and interictal EEG signals. Experimental results show that the proposed methods outperform the existing state-of-the-art method for better classification in terms of sensitivity, specificity, and accuracy with greater consistence of ictal and interictal period of epilepsy for benchmark datasets and different brain locations.

Electroencephalogram (EEG), a record of electrical signal to represent the human brain activity, has great potential for the diagnosis to treatment of mental disorder and brain diseases such as epileptic seizure. Features extraction and classification of EEG signals is the crucial task to detect the stage of ictal (i.e., seizure period) and interictal (i.e., period between seizures) signals for the treatment and precaution of the epileptic patient. However, existing seizure and non-seizure feature extraction techniques are not good enough for the classification of ictal and interictal EEG signals considering their non-abruptness phenomena and inconsistency in different brain locations. In this paper, we present a new approach for feature extraction and classification by exploiting temporal correlation within an EEG signal for better seizure detection as any abruptness in the temporal correlation within a signal represents the transition of a phenomenon. In the proposed methods we di...

Epilepsy is a chronic neurological disease characterized by a large electrical explosion that is excessive and uncontrolled, as defined by the world health organization. It is an anomaly that affects people of all ages. An electroencephalogram (EEG) of the brain activity is a widely known method designed as a reference dedicated to study epileptic seizures and to record the changes in brain electrical activity. Therefore, the prediction and early detection of epilepsy is necessary to provide timely preventive interventions that allow patients to be relieved from the harmful consequences of epileptic seizures. Despite decades of research, the prediction of these seizures with accuracy remains an unresolved problem. In this article, we have proposed five deep learning models on intracranial electroencephalogram (iEEG) datasets with the aim of automatically predicting epileptic seizures. The proposed models are based on the Convolutional Neural Network (CNN) model, the fusion of the tw...

Deep learning (DL) has been expensively applied in multiple fields like computer vision, speech recognition and natural language processing. The field of Epileptic seizure prediction didn’t receive the deserved attention by DL community, even though, deep neural networks can handle the challenging task of onsets prediction whilst achieving the highest rates of sensitivity, despite the complex nature of EEG signals. In the literature, this issue was addressed differently most of the time using handcrafted temporal and spectral features, machine learning techniques and rarely deep learning with extracted features. In this paper, we introduce an LSTM model designed to address the chaotic nature of an EEG signal in order to predict pre-ictal and inter-ictal states. Our model is evaluated on the publicly available CHBMIT database. We achieved an average sensitivity rate of 0.84 using a Raw EEG data segment as input to the LSTM model.

Epilepsy is the most common neurological disorder and an accurate forecast of seizures would help to overcome the patient's uncertainty and helplessness. In this contribution, we present and discuss a novel methodology for the classification of intracranial electroencephalography (iEEG) for seizure prediction. Contrary to previous approaches, we categorically refrain from an extraction of hand-crafted features and use a convolutional neural network (CNN) topology instead for both the determination of suitable signal characteristics and the binary classification of preictal and interictal segments. Three different models have been evaluated on public datasets with long-term recordings from four dogs and three patients. Overall, our findings demonstrate the general applicability. In this work we discuss the strengths and limitations of our methodology.

The Epilepsies are a common, chronic neurological disorder affecting more than 50 million individuals across the globe. It is characterized by unprovoked, recurring (similar or different type) seizures which are commonly diagnosed through clinical EEGs. Good-quality, open-access and free EEG data can act as a catalyst for on-going state-of-the-art (SOTA) research works for detection, prediction and management of epilepsy and seizures. They can also aid in improving the quality of life (QOL) of these diseased individuals and contribute research in healthcare multimedia, data analytics and Artificial Intelligence (AI) in personalized medicine. This paper presents widely used, available, open and free EEG datasets available for epilepsy and seizure diagnosis. A brief comparison and discussion of open and private datasets has also been done. Such datasets will help in development and evaluation of automatic computer-aided system in healthcare.

It is proved that electroencephalographic activity or other electrophysiological measures of brain function might provide a new non-muscular channel for sending messages and commands to the external world – a brain–computer interface (BCI). BCI for seizure prediction provides a means of controlling epilepsy in medically refractory patients. The site of epileptogenicity cannot be resected but yet can be defined sufficiently to be selectively influenced by strategically implanted electrodes. The goal is to provide these patients, who may be completely paralyzed, with basic communication capabilities so that they can express their wishes to caregivers or even operate word processing programs or neuroprostheses. By leveraging the potential of cloud computing and deep learning, we develop and deploy BCI seizure prediction and localization from scalp EEG and ECoG big data. First, a new method for epileptic seizure prediction and localization of the seizure focus is studied. Secondly, an e...

Seizure is the main factor that affects the quality of life of epileptic patients. The diagnosis of epilepsy, and hence the identification of epileptogenic zone, is commonly made by using continuous Electroencephalogram (EEG) signal monitoring. Seizure identification on EEG signals is made manually by epileptologists and this process is usually very long and error prone. The aim of this paper is to describe an automated method able to detect seizures in EEG signals, using knowledge discovery in database process and data mining methods and algorithms, which can support physicians during the seizure detection process. Our detection method is based on Artificial Neural Network classifier, trained by applying the multilayer perceptron algorithm, and by using a software application, called Training Builder that has been developed for the massive extraction of features from EEG signals. This tool is able to cover all the data preparation steps ranging from signal processing to data analys...

For several decades, the detection of epileptic seizures has been an active research topic. The performance of current patient-specific algorithms is satisfactory. However, due to significant variability of EEG data between patients, cross-subject seizure characterization and detection remains a challenging task. The purpose of this study is to propose and investigate a modified convolutional neural network (CNN) architecture based on separable depth-wise convolution for effective automatic cross-subject seizure detection. The architecture is conceived with a reduced number of trainable parameters to reduce the model complexity and storage requirements to easily deploy it in connected devices for real-time seizure detection. The performance of the proposed method is evaluated on two public datasets collected in the Children’s Hospital Boston and the University of Bonn respectively. The method achieves the highest sensitivity-false positive rate/h of 91.93%–0.005, 100%–0.057 for the ...

The patients suffering from refractory epilepsy can improve their quality of life through seizure forecasting. Though the initial findings were positive, and there is no single attribute has been originating and proficient in personally symbolizing the brain crescendos during the period of seizure transition. Recently, “Cross-frequency Phase Amplitude Coupling” has been projected and considered a forerunner of the activity of seizure. This study elaborates the presence of statistical important variance in mean amplitude coupling distribution phase between the interictal and preictal seizure states in animals (dogs) with intracranial electrodes bilateral implant. According to results, a change of statistic significant (p < 0.05) of the amplitude coupling phase in the preictal phase period. This change is basically correlated with the implanted electrodes position and specifically more important within frequency bands of high-gamma. These discoveries prominent the bilateral iEEG po...

Deep learning (DL) has been expensively applied in multiple fields like computer vision, speech recognition and natural language processing. The field of Epileptic seizure prediction didn’t receive the deserved attention by DL community, even though, deep neural networks can handle the challenging task of onsets prediction whilst achieving the highest rates of sensitivity, despite the complex nature of EEG signals. In the literature, this issue was addressed differently most of the time using handcrafted temporal and spectral features, machine learning techniques and rarely deep learning with extracted features. In this paper, we introduce an LSTM model designed to address the chaotic nature of an EEG signal in order to predict pre-ictal and inter-ictal states. Our model is evaluated on the publicly available CHBMIT database. We achieved an average sensitivity rate of 0.84 using a Raw EEG data segment as input to the LSTM model.

We propose a robust method for automated detection of epileptic seizures using intracranial electroencephalogram (iEEG) recordings with two electrodes. The state-of-the-art seizure detection methods suffer from high number of false detections, even when designed to be patient-specific. The solution reported here aims to achieve very low false detection rate, while providing a high sensitivity. Two adjacent iEEG recordings are subtracted from each other to make the bipolar iEEG signal. The values achieved from singular value decomposition (SVD) of the bipolar iEEG signal are used as measure. A threshold is subsequently applied on the measure. Results indicate robustness of the proposed measure for seizure detection. The method is applied on 5 invasive recordings containing 54 seizures in 780 hours of multichannel iEEG recordings. On average, the results revealed 85.2% sensitivity and a very low false detection rate of 0.02 per hour in long-term continuous iEEG recordings.

Scarcity of good quality electroencephalography (EEG) data is one of the roadblocks for accurate seizure prediction. This work proposes a deep convolutional generative adversarial network (DCGAN) to generate synthetic EEG data. Another objective of our study is to use transferlearning (TL) for evaluating the performance of four well-known deep-learning (DL) models to predict epileptic seizure. Methods: We proposed an algorithm that generate synthetic data using DCGAN trained on real EEG data in a patient-specific manner. We validate quality of generated data using one-class SVM and a new proposal namely convolutional epileptic seizure predictor (CESP). We evaluate performance of VGG16, VGG19, ResNet50, and Inceptionv3 trained on augmented data using TL with average time of 10 min between true prediction and seizure onset samples. Results: The CESP model achieves sensitivity of 78.11% and 88.21%, and false prediction rate of 0.27/h and 0.14/h for training on synthesized and testing on real Epilepsyecosystem and CHB-MIT datasets, respectively. Using TL and augmented data, Inceptionv3 achieved highest accuracy with sensitivity of 90.03% and 0.03 FPR/h. With the proposed data augmentation method prediction results of CESP model and Inceptionv3 increased by 4-5% as compared to state-of-the-art augmentation techniques. Conclusion: The performance of CESP shows that synthetic data acquired association between features and labels very well and by using the augmented data CESP predicted better than chance level for both datasets. Significance: The proposed DCGAN can be used to generate synthetic data to increase the prediction performance and to overcome good quality data scarcity issue.

Epilepsy affected patient experiences more than one frequency seizures which can not be treated with medication or surgical procedures in 30% of the cases. Therefore, an early prediction of these seizures is inevitable for these cases to control them with therapeutic interventions. Methods: In recent years, researchers have proposed multiple deep learning based methods for detection of preictal state in electroencephalogram (EEG) signals, however, accurate detection of start of preictal state remains a challenge. We propose a novel ensemble classifier based method that gets the comprehensive feature set as input and combines three different classifiers to detect the preictal state. Results: We have applied the proposed method on the publicly available scalp EEG dataset CHBMIT of 22 subjects. An average accuracy of 94.31% with sensitivity and specificity of 94.73% and 93.72% respectively has been achieved with the method proposed in this study. Conclusions: Proposed study utilizes the preprocessing techniques for noise removal, combines deep learning based and handcrafted features and an ensemble classifier for detection of start of preictal state. Proposed method gives better results in terms of accuracy, sensitivity, and specificity.

Abstract— Temporal seizures due to Hippocampal origins are very common amongst epileptic patients. This article presents a novel seizure prediction approach based on a combination of graph and Chaos theory. The early identification of seizure signature allows for various preventive measures to be undertaken. The proposed non-linear approach consists of observing a high correlation level between any pair of electrodes for the lower frequencies and a decrease in the Lyapunov index (chaos or entropy) for the higher frequencies. The Fast Fourier Transform (FFT) and statistical analysis tools were used to determine threshold levels for the lower frequencies. A graph topology characterizes seizure signatures for each patient. These graphs are used to monitor both Delta and Gamma behaviors. In order to validate the proposed approach, six patients from both sexes and various age groups with temporal epilepsies originating from the hippocampal area were studied using the Freiburg Database. A...

An epileptic seizure is a sign of abnormal activity in the human brain. Electroencephalogram (EEG) is a standard tool that has been used vastly for detection of seizure activities. Many methods have been developed to help the neurophysiologists to detect the seizure activities with high accuracy. Most of them rely on the features extracted in the time, frequency, or time-frequency domains. The performance of the proposed methods is related to the performance of the features extracted from EEG recordings. Deep neural networks enable learning directly on the data without the domain knowledge needed to construct a feature set. This approach has been hugely successful in almost all machine learning applications. We propose a new framework that also learns directly from the data, without extracting a feature set. We proposed an original deep-learning-based method to classify EEG recordings. The EEG signal is segmented into 4 s segments and used to train the long- and short-term memory ne...

Epilepsy patients experience challenges in daily life due to precautions they have to take in order to cope with this condition. When a seizure occurs, it might cause injuries or endanger the life of the patients or others, especially when they are using heavy machinery, e.g., deriving cars. Studies of epilepsy often rely on electroencephalogram (EEG) signals in order to analyze the behavior of the brain during seizures. Locating the seizure period in EEG recordings manually is difficult and time consuming; one often needs to skim through tens or even hundreds of hours of EEG recordings. Therefore, automatic detection of such an activity is of great importance. Another potential usage of EEG signal analysis is in the prediction of epileptic activities before they occur, as this will enable the patients (and caregivers) to take appropriate precautions. In this paper, we first present an overview of seizure detection and prediction problem and provide insights on the challenges in this area. Second, we cover some of the state-of-the-art seizure detection and prediction algorithms and provide comparison between these algorithms. Finally, we conclude with future research directions and open problems in this topic.

The instantaneous frequency has been frequently employed as a feature for the detection of the oscillatory type of seizures in electroencephalogram signals. However, seizures appearing as spikes cannot be analyzed using the Instantaneous frequency. In this study, we propose a new time-frequency feature that exploits both the instantaneous frequency and group delay to detect seizures with both spike and oscillatory characteristics in elec-troencephalogram (EEG) recordings. It is demonstrated that the computational cost of extracting the proposed feature is less than the features of similar performance. To evaluate the performance of the proposed feature, we employed a newborn EEG database. The proposed feature in combination with other features achieves a classification accuracy of 99%, which is 0.75% greater than the existing methods.

Patients suffering from epileptic seizures are usually treated with medication and/or surgical procedures. However, in more than 30% of cases, medication or surgery does not effectively control seizure activity. A method that predicts the onset of a seizure before it occurs may prove useful as patients might be alerted to make themselves safe or seizures could be prevented with therapeutic interventions just before they occur. Abnormal neuronal activity, the preictal state, starts a few minutes before the onset of a seizure. In recent years, different methods have been proposed to predict the start of the preictal state. These studies follow some common steps, including recording of EEG signals, preprocessing, feature extraction, classification, and postprocessing. However, online prediction of epileptic seizures remains a challenge as all these steps need further refinement to achieve high sensitivity and low false positive rate. In this paper, we present a comparison of state-of-the-art methods used to predict seizures using both scalp and intracranial EEG signals and suggest improvements to existing methods.

—The Hilbert Huang Transform (HHT) has been used extensively in the time-frequency analysis of electroencephalogra-phy (EEG) signals and Brain-Computer Interfaces. Most studies utilizing the HHT for extracting features in seizure prediction have used intracranial EEG recordings. Invasive implants in the cortex have unknown long term consequences and pose the risk of complications during surgery. This added risk dimension makes them unsuitable for continuous monitoring as would be the requirement in a Body Area Network. We present an HHT-based system on Field Programmable Gate Array (FPGA) for predicting epileptic seizures using scalp EEG. We use bandwidth features of Intrinsic Mode Functions and obtain a classification accuracy of close to 100% using patient-specific classifiers in software. Details of FPGA implementation are also given.

A graphical and analytical description of epileptic seizures based on amplitude modulation and frequency modulation components of intracranial EEG (iEEG) is proposed. This representation allows the characterization of seizures and their different stages from the iEEG by means of triangles whose vertexes are the centroids (c m ) of the signal during preictal, ictal and postictal periods. The centroid is the point defined by the average values of instantaneous amplitude and frequency, a i and f i respectively. Data were obtained from 8 patients with recurrent epilepsy, 170 records were processed, 62 of which were seizures and 108 interictal signals. Results show that the centroids of the ictal periods are located in a region of the space a i -f i distant from the centroids corresponding to interictal and postictal periods. This original representation of epileptic seizures can facilitate the visualization of stagetransitions and discrimination between the different stages of the iEEG signal. An additional advantage of the method is that the information contained in the signal is synthesized significantly.

When dealing with seizure detection/prediction problems, there are three main performance metrics that must be optimized: false positive rate, false negative rate, detection delay or, if the problem is seizure prediction, it is desirable to obtain the greatest prediction time achievable. Tuning specific extracted features to individual patients can lead to improved results. The processing window length is also an important parameter whose optimization may significantly affect performance. In this study we propose an approach for selecting the window length for the particular detection/prediction problem. This approach is applicable to other feature parameters suitable for tuning or optimization.

Abstract— Temporal seizures due to Hippocampal origins are very common amongst epileptic patients. This article presents a novel seizure prediction approach based on a combination of graph and Chaos theory. The early identification of seizure signature allows for various preventive measures to be undertaken. The proposed non-linear approach consists of observing a high correlation level between any pair of electrodes for the lower frequencies and a decrease in the Lyapunov index (chaos or entropy) for the higher frequencies. The Fast Fourier Transform (FFT) and statistical analysis tools were used to determine threshold levels for the lower frequencies. A graph topology characterizes seizure signatures for each patient. These graphs are used to monitor both Delta and Gamma behaviors. In order to validate the proposed approach, six patients from both sexes and various age groups with temporal epilepsies originating from the hippocampal area were studied using the Freiburg Database. A...

Epilepsy is one of the common neurological disorders characterized by a sudden and recurrent malfunction of the brain that is termed "seizure", affecting over 50 million individuals worldwide. The Electroencephalogram (EEG) is the most influential technique in detection of epileptic seizures. In recent years, many research works have been devoted to the detection of epileptic seizures based on analysis of EEG signals. Despite remarkable work on seizure detection, there is no generic seizure detection scheme which performs reasonably well for different patients and different brain locations. In this paper we present a generic approach for feature extraction of preictal (pre-stage of seizure onset) and interictal (period between seizures) EEG signals using empirical mode decomposition (EMD) along with discrete cosine transformation (DCT) by exploit temporal correlation for detection of preictal phase of epileptic seizure. Then least square support vector machine is applied on the features for classifications. Results demonstrate that our proposed method outperforms the state-of-the-art methods in terms of sensitivity, specificity and accuracy to classify preictal and interictal EEG signals to the benchmark dataset extracted from different brain locations of different patients.

Automated seizure prediction has a potential in epilepsy monitoring, diagnosis, and rehabilitation. Electroencephalogram (EEG) is widely used for seizure detection and prediction. This paper proposes a new seizure prediction approach based on spatiotemporal relationship of EEG signals using phase correlation. This measures the relative change between current and reference vectors of EEG signals which can be used to identify preictal/ictal (before the actual seizure onset/ actual seizure period) and interictal (period between adjacent seizures) EEG signals to predict the seizure. The experiments show that the proposed method is less sensitive to artifacts and provides higher prediction accuracy (i.e., 91.95%) and lower number of false alarms compared to the state-of-the-art methods using intracranial EEG signals in different brain locations of 21 patients from a benchmark data set.

Temporal seizures due to Hippocampal origins are very common amongst epileptic patients. This article presents a novel seizure prediction approach based on a combination of graph and Chaos theory. The early identification of seizure signature allows for various preventive measures to be undertaken. The proposed non-linear approach consists of observing a high correlation level between any pair of electrodes for the lower frequencies and a decrease in the Lyapunov index (chaos or entropy) for the higher frequencies.

When dealing with seizure detection/prediction problems, there are three main performance metrics that must be optimized: false positive rate, false negative rate, detection delay or, if the problem is seizure prediction, it is desirable to obtain the greatest prediction time achievable. Tuning specific extracted features to individual patients can lead to improved results. The processing window length is also an important parameter whose optimization may significantly affect performance. In this study we propose an approach for selecting the window length for the particular detection/prediction problem. This approach is applicable to other feature parameters suitable for tuning or optimization.