Automatic Spike and Wave identification in the EEG of epilectic patients for the prediction of epilectic seizures

To analyse epileptic seizures, EEG signals are fundamental for neurologists to diagnose pathological events and evaluate medical treatments. Doing so is not very time-efficient, so an automatic method is required for faster analysis. The basis for this method is the pathological event of the spike and wave, which is indicative for epilepsy. The objective of this study is the ideation of a spike and wave detector based on a set of prototypes spike and waves and match filters to support the diagnosis of epilepsy and predict the incoming attack with the detection of spike and waves in EEG signals. The first section provides a brief anatomical, physiological and biological introduction to the central nervous system, essential to understand the fundaments of the epileptic seizures. A brief description of seizures and their classifications is also provided, along with their pathological waveforms shown in EEG recordings. The second section is dedicated to the state of the art, exploring the many algorithms dedicated to the identification of spike and waves in EEG recordings. The third section describes the dataset used in this study and each individual step of the algorithm developed for this research. The results achieved are explained in the fourth section.