Nonlinear analysis of epileptiform and normal EEG in five partial epileptic patients

Abstract

EEG recording is important in the diagnosis of various neurologic diseases, coma, and toxic-metabolic encephalopathies and is analyzed visually. At present, neurologists analyze EEGs objectively on the assumptions that EEG results from nonlinear dynamics and changes in neuronal complexity can be found with epileptiform discharges. Digital EEGs of the interictal state in five partial epileptic patients without cerebral structural lesion were included in this study. We selected an EEG epoch containing epileptiform discharges (state I) and another epoch without epileptiform discharges (state II) from each patient. We used a software, which was programmed in our laboratory, to perform two nonlinear measures, FD and LI. Statistical analysis was performed by SPSS windows. FD of EKG did not change regardless of the state and was lower than that of EEG. This result suggests that the degree of complexity of EKG is lower than that of EEG. Mean FD of all electrodes of state I was lower than that of state II so the degree of complexity of all electrodes drops with partial epileptiform discharge. This study suggests that the hypersynchrony of partial epilepsy is involved in all channels of brain. Future studies should include nonlinear analysis of various time frames of EEG in order to evaluate neurophysiologic changes in partial epilepsy.