Detection of K-Complex in Sleep EEG Signal using Support Vector Machine

The K-complex is a transient EEG waveform that contributes to the assessment of sleep stages. Due to the non stationary and non linear behaviour of the brain signals, it is very difficult to study the characteristics of these transients manually.The main difficulty of the automated K-complex detection problem has been the lack of specific characterization and the close similar at to other EEG waves. We present a detection approach based on feature extraction and Support Vector machine (SVM) that provides good agreement with visual K-complex recognition. In this case, Independent Component Analysis (ICA) is used for the denoising purpose.The performance of this method is considerably efficient & is a hardware independent solution for the biomedical signal processing field. Index Term — EEG Signal, ,K-Complex, Sleep Stages , Fast ICA Algorithm , Golay Filter, ,Support Vector machine (SVM) ——————————  ——————————