Application of Wavelet Transform in Analysis of EEG Signal Recorded During Motor Imagery Tasks

Snimanjem i analizom EEG signala moguće je detektirati namjeru voljnog pokreta tijela. Prije pojave mišićne aktivnosti dolazi do pojave evociranih potencijala karakterističnih frekvencijskih svojstava. Pokazano je da do pojave istih takvih evociranih potencijala dolazi i ako iz nekog razloga se izvršavanje pokreta ne dogodi. U ovom radu, objašnjeni su evocirani potencijali zamišljanja pokreta, valićna transformacija kao metoda analize signala te je razvijen algoritam klasifikacije EEG signala snimljenih tijekom zamišljanja pokreta lijeve i desne ruke. Za obradu EEG signala korištena je diskretna valićna transformacija (engl. Discrete Wavelet Transformation, DWT), a klasifikacija je provedena linearnom diskriminantnom analizom (engl. Linear discriminant analysis, LDA).By recording and analyzing EEG signals it is possible to detect the intention of voluntary movements. Before the occurrence of muscle activity, evoked potentials appear as a part of the brain electrical activity. It has been shown that those evoked potentials appear even when the intended movement is not executed for some reason. This paper explains evoked potentials that appear as a result of motor imagery and the wavelet transform as a method of signal analysis. It also describes an algorithm for classification of signals recorded during right and left hand motor imagery. The discrete wavelet transform (DWT) was used to analyze the EEG signals and the signal classification was done using linear discriminant analysis (LDA)