Prognostication of EEG Signal using Wavelets

Abstract

A comprehensive quantitative analysis of electroencephalogram signals is carried out. Due to the non-stationary nature of EEG signals the visual investigation of EEG information is restrictively tedious and wasteful, regardless of the possibility that the master clinician reads the data ten times quicker than the recording speed. The visual assessment needs quantitative investigation which can reveal concealed characters of the data. Wavelet provide a solution and provides functions for synthesizing and analyzing signals, pictures as well as information that show general conduct punctuated with sudden changes. Properties from the accessible database are separated and investigation of signal with various wavelets is prepared to identify and foresee the kind of disorders. This article proposes a method for reliable detection of different types of disorders by using different wavelets as haar wavelet, Shannon wavelet etc to the database signals and on the basis of which we are able to prognosticate. Further the type of disorder is predicted using Artificial Neural Network classifiers.