A PRINCIPAL COMPONENT MODEL OF THE TIME SERIES OF THE BROAD-BAND EEG FREQUENCY SPECTRA AND OF THE EEG SPECTRAL POWER DENSITIES

Publisher Summary This chapter presents a principal component model of the time series of the broad-band EEG frequency spectra and of the EEG spectral power densities. The changes of EEG signals because of toxic agents and drugs are studied using power spectra computed via the discrete Fourier Transform (FFT). The chapter presents a summarizing graph of the power spectra, drawn as compressed spectral arrays of the EEG signal of a rat before and during a lethal CO intoxication: on the x-axis is frequency, on the y-axis is power density, and on the z-axis is time in minutes. The principal component model is applicable even in a simpler situation, when the broadband EEG frequency analysis is used. The chapter describes the broad-band EEG frequency analysis using the active RC-filters. The system is alibrated by sinusoidal signals, and the results are expressed as the average mean voltages over the 5 s epochs.