Frontal electroencephalogram analysis with ensemble empirical mode decomposition during the induction of general anesthesia. (30th November 2016)

• Record Type: Journal Article
• Authors: Tsai, Feng-Fang; Hu, Xiyuan; Lin, Yi-Shiuan; Peng, Chung-Kang; Fan, Shou-Zen
• Is Part Of: Biomedical physics & engineering express. Volume 2:Number 6(2016)
• Publication Date: 2016-11-30
• Subjects: ensemble empirical mode decomposition -- frontal electroencephalogram -- Hilbert–Huang transformation; Medical physics -- Periodicals; Biophysics -- Periodicals; Biomedical engineering -- Periodicals; Medical sciences -- Periodicals; 610.153
• DOI: 10.1088/2057-1976/2/6/065004
• Languages: English
• ISSNs: 2057-1976

Abstract

Abstract: Background. Ensemble empirical mode decomposition (EEMD) was proposed for decomposing electroencephalography (EEG) signals into intrinsic mode functions (IMFs), which obtain instantaneous frequency data and work well with data that are nonstationary and nonlinear. Hilbert–Huang Transformation (HHT) was used in this study to convert IMFs into spectrograms, which are useful for observation. We recorded EEG signals through a bispectral index (BIS) monitor for EEMD analysis, and calculated the energy change after HHT. Methods. A total of 19 patients who had received general anesthesia were included. The EEG signals were recorded by physiological monitor with BIS electrode strip and saved in a portal computer. The frequency changes of the IMF spectrograms were compared during the induction period, and the raw data energy changes were quantified with moving window standard deviation every 10 s. Results. The second IMF, with an initial spectrum of approximately 10–30 Hz, was focused between 10 and 15 Hz after the patient was anesthetized. All patients presented with decreased frequency and bandwidth focusing in the second IMF and indicated energy gathering. Conclusions. We found energy gathering in IMF2 after patient was anesthetized. The results suggest that examining IMFs rather than EEG signals was more useful for determining the particular bandwidth changes in which synchronization phenomena occur. With this method, it is easily to observe the separate energy changes of IMFs within the EEG signals.