Detecção e caracterização de crises epilépticas através de estimulação elétrica

Abstract

Epilepsy is a highly debilitating disease, marked by the imprevisibility of its seizures. Seizure prediction is an essencial tool for a more ecient control over the disease, which is, currently, treated in a continuous fashion, lacking a method to reliably foresee the moment of the seizure and treat the disease only when needed. Recently, a new approach for seizure prediction has been suggested, based on electrical stimulation as means of probing the neural network to assess the status of the brain. The present work shows preliminary studies in signal processing of EEG data, colected in an experiment where rats are submitted to pro-epileptic drug infusion (PTZ) together with electrical stimulation in the brain. Validating the control groups, it has been found that the interictal discharges slowly synchronize with the stimulus, settling in different moments, but altogether with great temporal coupling, which raises the possibility of developing a closed-loop device for prediction and supression of epileptic seizures.