Naslov
Detection of chaotic EEG in traumatic cerebral coma: the value of Fourier transform technique

Autori
Sekulić, Ante ; Išgum, Velimir

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
Proc. of the 8th International IMEKO Conference on Measurement in Clinical Medicine, BMI'98 / Magjarević, Ratko - Zagreb : Hrvatsko društvo za komunikacije, računarstvo, elektroniku, mjerenja I automatiku (KoREMA), 1998, 3.45-3.48

ISBN
953-6037-26-2

Skup
8th International IMEKO Conference on Measurement in Clinical Medicine, BMI'98

Mjesto i datum
Dubrovnik, Hrvatska, 16.09.1998. - 19.09.1998

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
EEG; coma; Fowier transfonn; neurodynamics; nonlinear analysis

Sažetak
Nonlinear evolution equations play a central role in applied mathematics and mathematical physics. These equations can describe many natural phenomena including human EEG and neurodynamics. Fourier transform technique is well known in solving and investigating linear phenomena. This technique can be also used in nonlinear context. In this way Fourier transform technique and spectral analysis of human EEG can help in investigation of periodic, quasiperiodic or chaotic time series i.e. linear or nonlinear context of neurodynamics. The later is observed during epi1eptic seizures as chaotic time series of EEG. The aim of this paper is to show the value of Fourier transform technique in nonlinear analysis of human EEG and detection of epi1eptic seizures in traumatic coma. EEG was recorded in 8 patients with traumatic coma after severe head injury. Visually inspected periods of EEG with the best 2 signal to noise ratio were used for non-linear analysis. This analysis included: construction of phase space, Fourier transform technique for spectral analysis, autocorrelation function and maximal Lyapunov exponent. Stationary stochastic time series of EEG have had clear picture of the main frequency in spectral w analysis. Quasiperiodicity was accompanied by two or more main frequencies. Onset of chaotic time series was without the main frequency. In epi1eptic seizures chaotic r time series of EEG were observed. Chaotic time serie of EEG have had the following characteristics: trajectories in the phase space with wide distribution, no main n. frequency in spectral analysis by Fourier transform technique, autocorrelation function with characteristics t of nonstationary time serie and very positive maximal Lyapunov exponent. Fourier transform technique used e for spectral analysis of EEG can be a good indicator of the onset of chaotic time serie i.e. onset of epileptic seizure in traumatic coma.

Izvorni jezik
Engleski

Znanstvena područja
Elektrotehnika, Kliničke medicinske znanosti

POVEZANOST RADA