Fractal Analysis of EEG Signals in the Brain of Epileptic Rats, with and without Biocompatible Implanted Neuroreservoirs

Abstract

Current epilepsy rates in Mexico are 4% (SERSAME-Health Ministry), of which 80% correspond to Temporal Lobe Epilepsy (TLE). Antiepileptic drug administration is systemic, meaning that 90% of the active agent is lost between administration and delivery to the epileptic focus in the brain. Severe toxic secondary effects may occur as a result. The present study is aimed at developing an alternative antiepileptic drug delivery system. In this study, a sol-gel nanostructured titania device, in which valproic acid (VPA) has been encapsulated. This is a nanoparticulate device, which is biocompatible with brain tissue. Stereotactic surgery was used to implant the reservoirs in the temporal lobe of Wistar rats, using chemical kindling, which was used to induce epilepsy. The reservoir was designed to release the drug at a constant rate over a period of at least one year. A functional study was performed on the efficiency of drug delivery in order to evaluate the effect on spontaneous and induced neuron electrical activity. A new discovery, which is presented here, shows that in the case of damaged brain tissue, as is the case in epilepsy, the accumulation of red globules, oxygen transportation results in the formation of calcium carbonate crystals which surround the epileptic focus. Because these crystals have a specific polarization, we propose to characterize their influence on the EEG using statistical methods. The electrical activity was measured by electroencephalography using 5 healthy rats without and 5 rats with an implanted VPA/device. Cerebral signals describe the complex behavior of the brain dynamics as a function of time. Fractal algorithms are sensitive to fluctuations and lead to the analysis and characterization of this kind of complex phenomena. A systematic study of these EEG’s was made in order to observe the variation of signals during seizures and on the controlled rate of release of VPA. We have estimated the Hurst exponent (H) to measure long range-dependence. Preliminary results show that for the control group, signal behavior is persistent (H>0.5), while for the epileptic group antipersistency was observed (H<0.5), with variations due seizure stages. During the protection period using VPA, preliminary results show that values tend to reach original behavior, as the crisis is stabilized.