Wavelet Transform-Based Algorithm for Single Spike-and-Wave Discharges Detection in Epileptic Patients' Electroencephalogram

The aim of current research is to develop the algorithm for on-line detection of spike-and-wave discharges (SWDs), which are of a peculiar waveform known to appear in epileptic patients' electroencephalogram. In this study, the wavelet transform was used as a backbone of the algorithm. Considering the possibility of applying the magnetic stimulation therapy in the future, 100% specificity was intended to grant the safety of the patient. Extracted by clinician from the thirty minutes four-lead EEG data recording fifty-four SWD samples and fifteen pseudo-SWD samples were employed. Our results of the simulated on-line detection showed the perfect distinction of SWD from the pseudo-SWD, however, few cases of over detection (not SWD, not pseudo-SWD), made the specificity and safety of developed algorithm to become a subject of discussion.