Use of Artificial Neural Networks for Classification Intracraneal EEG Signals fron Epileptic Patients

摘要

An epileptic seizure is an episode of abnormal electrical brain activity that might involve partial or total consciousness loss and/or involuntary movements among others uncomfortable sensations in people who suffer it. At least 1% population worldwide suffer from epilepsy, therefore research aimed to develop new techniques to treat this condition is essential. In the present work an Artificial Neural Network (ANN) implementation is proposed by using the Back-propagation Algorithm (BP) which is used to perform a classification of real Intracranial Electroencephalogram (iEEG) signals into interictal or ictal following these steps: (i) Data reduction by extracting 7 features of the signal, (ii) Implementing the Backpropagation algorithm in a DSP platform with the computed features as input and the signals classification, ictal or interictal as output.