Wavelet-based spectral analysis of the electroencephalogram under brain ischemia

Time-varying spectral analysis is frequently used for studying changes of the electroencephalogram (EEG) under brain ischemia caused by blockage of blood supply. The wavelet transform (WT) can be interpreted as a tool for frequency representation which uses varying bandwidth and time window length for different frequencies. In contrast to the Short-Time Fourier Transform (STFT), this enables a better resolution for high frequencies, even if long temporal windows are necessary to process low frequencies. This study aims to present preliminary results with wavelet transform applied to EEG recordings under carotid endarterectomy with special focus on the spectral changes during the transition to the pathological state. It is observed that high frequency power drops in two steps with a delay of approximately 40 seconds between them.