Brain Seizure Detection and Classification Using EEG Signals

Electroencephalogram (EEG) remains the most immediate, simple, and rich source of information for understanding phenomena related to brain electrical activities. The objective of the book is to analyze the EEG signals to observe abnormalities of brain activities called epileptic seizure. Seizure is a neurological disorder in which too many neurons are excited at the same time and are triggered by brain injury or by chemical imbalance. The seizures are predominantly characterized by unpredictable interruptions of normal brain function. A seizure occurs when too many nerve cells in the brain “fire” too quickly causing an “electrical storm.” The EEG signals recorded from epileptic patients are analyzed for monitoring extracting behavior of signals during onset seizures. Epileptic seizure detection still poses challenges in the field of accurate seizure detection and prediction of seizures. Mostly these techniques are analyzed on the basis of detection and classification accuracy, sensitivity and specificity. Brain Seizure Detection and Classification Using Electroencephalographic Signals presents EEG signal processing and analysis with high performance feature extraction. The Time and Frequency Domain (TFD), Wavelet Transform (WT) and Empirical Mode Decomposition (EMD) are optimized feature extraction methods presented by the authors. The book also covers the feature selection method based on One-way ANOVA along with high performance machine learning classifiers for classification of EEG signals in normal and epileptic EEG signals. In addition, the authors also present new methods of feature extraction, including Singular Spectrum-Empirical Wavelet Transform (SSEWT) for improved classification of seizures in significant seizure-types, specifically epileptic and Non-Epileptic Seizures (NES). The performance of the system will be compared with existing methods of feature extraction using Wavelet Transform (WT) and Empirical Wavelet Transform (EWT). The machine learning classifiers are used for classification of EEG signal in normal, epileptic seizure, non-epileptic seizure, and thus non-epileptic patients. One of the major new contributions of the book is identification of non-epileptic patients using SSEWT. Presents EEG signal processing and analysis with high performance feature extractionDiscusses recent trends in seizure detection, prediction and classification methodologiesClassification of epileptic and non-epileptic seizures is still a demanding issue, and misdiagnosing NES leads to the unnecessary use of antiepileptic medication, which can worsen NES and affect learning or working abilityThe authors present new guidance and technical discussion in these areasPresents a variety of feature-extraction methods, including Time and Frequency Domain (TFD), Wavelet Transform (WT), Empirical Mode Decomposition (EMD), and feature selection methods based on One-way ANOVA along with high performance machine learning classifiers for classification of EEG signals in normal and epileptic EEG signals. Presents new methods of feature extraction developed by the authors, including Singular Spectrum-Empirical WaveletTransform (SSEWT) for improved classification of seizures in significant seizure-types, specifically epileptic and NonEpileptic Seizures (NES)