Detection of characteristic waves of sleep EEG by continuous wavelet transform

In recent years, sleep disturbances on day to day life, has been a major problem. Sleep staging is clinically important in the assessment of sleep quality. The usual method for sleep staging is time consuming and laborious procedure since it depends on the experience and visual inspection of sleep experts. Different characteristic waves like k-complex, sleep spindles, vertex and slow waves are often found in EEG signal during different sleep stages. Detection of these characteristic waveforms in the EEG is an important component of sleep staging. These characteristic waves are having typical signature waveforms, which are non-stationary in nature. Therefore, in this work, an attempt was made to identify these characteristic waves using continuous wavelet transform. Finally, it is emphasized that the features extracted from the time frequency analysis is quite robust in identifying these characteristic waves.