Power spectrum and spectrogram of EEG analysis during general anesthesia: Python-based computer programming analysis

The commonly used principle for measuring the depth of anesthesia involves changes in the frequency components of the electroencephalogram (EEG) under general anesthesia. Therefore, it is essential to construct an effective spectrum and spectrogram to analyze the relationship between the depth of anesthesia and the EEG frequency during general anesthesia. This paper reviews the computer programming techniques for analyzing the spectrum and spectrogram derived from a single-channel EEG recorded during general anesthesia. A periodogram is obtained by repeating a Fourier transform on EEG segments separated by short time intervals, but spectral leakage (i.e., dissociation from the true spectrum) occurs as a consequence of unnatural segmentation and noise. While offsetting the securing of the dynamic range, practical analyses of the adaptation of the window function are explained. Finally, the multitaper method, which can suppress artifacts caused by the edges of the analysis segments, suppress noise, and probabilistically infer values that are close to the real power spectral density, is explained using practical examples of the analysis. All analyses were performed and all graphs plotted using Python under Jupyter Notebook. The analyses demonstrated the effectiveness of Python-based programming under the integrated development environment Jupyter Notebook for constructing an effective spectrum and spectrogram for analyzing the relationship between the depth of anesthesia and EEG frequency analysis in general anesthesia.