BACKGROUND: The dynamic characteristics of electroencephalogram (EEG) include a decrease in the chaotic dimension, the correlation dimen sion, the Lyapunov exponent, the chaotic complexity, the freedom of EEG and an enhanced synchronization and periodicity of the EEG from several minutes to tens of minutes before epileptic seizures. All these characteristics prefigure the forthcoming seizures. Some studies have proven that the non linear dynamical system can be used as a feasible approach to explore the potential variables for describing the chaos portrait of EEG. OBJECTIVE: To analyze the electric characteristics of EEG signal in the epileptic seizures in rat model by investigating the nonlinear dynamical variables, such as the approximate entropy (ApEn) and correlation dimen sion. DESIGN: Observational and experimental study based on animals. SETTING: Department of Medical Engineering, Department of Gastroen terology, Second Artilleryman General Hospital of Chinese PLA; Department of Physics, Faulty of Biomedical Engineering, Fourth Military Medical Uni versity of Chinese PLA. MATERIALS: From September 2001 to January 2002, this study was conducted at the Complexity Laboratory of the Biomedical Department of the Fourth Military Medical University of Chinese PLA. Six male SD rats,weighing 150- 200 g, were selected.INTERVENTIONS: After intraperitoneal injection of chloral hydrate (0. 5 mL), the male SD rats were deeply anesthetized. When their EEG signal became stable, bemegride injection was diluted at 1:1 with saline and was given on a volume of 0.5 mL to the rats intraperitoneally. After a while,the epileptic seizures started marked by a spasm with a deep roar. The entire epileptic seizures were recorded. According to the shape of EEG waves and the corresponding symptoms of the rats during their seizures, data of the four phases, referring to normal condition, preictal phase, ictal phase and postictal phases of epileptic seizures, were selected for nonlinear analysis. The variations of the ApEn and the correlation dimension were calculated.MAIN OUTCOME MEASURES: In the four phases of the seizures, before seizures, preictal phase, ictal phase and postictal phases, the changes in the ApEn and correlation dimension were observed.RESULTS: All the 6 rats entered the statistical procedure. During epilepsy, the ApEn and correlation dimension of the EEG signal in ictal phases (0. 447 ±0. 126, 2. 166 ±0. 377) decreased significantly while those in preictal phases(0. 807 ±0. 182, 4. 773 ±0. 319) and postictal phases (1. 241 ±0. 125, 6. 042 ±0. 373) (t = -3. 984to 17. 902, P ＜0. 01). The ApEn and the correlation dimension of the EEG signal in preictal and ictal phases had significant difference with those observed under normal conditions (1.313 ± 0. 090, 6. 405 ± 0. 694) (t = -5. 228 to 12. 740, P ＜ 0. 01 ).CONCLUSION: The changes in ApEn and correlation dimension showed by nonlinear dynamical approach in this study reflect the characteristics of EEG signals in preictal time, ictal time and postictal timeof the epileptic seizures and the differences among them. Additionally, they also reveal the laws in the changes of the complex ictal EEG signal.

In this paper,a pulsed magnetic fields stimulation instrument is designed and realized,which provides a pulsed magnetic field with the range of maximal intensity from 0.01~2T,frequency from 0.2~100Hz and time width of pulse from 0.01~1ms.The instrument,controlled by the hand,foot or itself,can display stimulus intensity and times and output trigger signals with different waveforms to make measuring devices operate synchronistically.

The aim of this study was to develop a new method of epileptic prediction using nonlinear dynamic theory. When rat was falling sickness, its EEG was researched by using approximate entropy and correlation dimension. The results showed the approximate entropy and correlation dimension during epileptic seizure are obviously lower than those before seizure and after seizure. The span of time before seizure is a special phase. Before the seizure symptom appeared, the complexity of EEG had begun declining. Thus, the outbreak of epilepsy could be predicted in short time using nonlinear dynamic methods.
Algorithms ; Animals ; Disease Models, Animal ; Electroencephalography ; Entropy ; Epilepsy ; diagnosis ; physiopathology ; Male ; Nonlinear Dynamics ; Rats ; Rats, Sprague-Dawley ; Signal Processing, Computer-Assisted