Abstracts

Epileptic seizures are network phenomena involving synchronous activity of multiple neurons. Partial seizures originate from focal regions of epileptogenesis and have variable patterns of regional propagation. The influences of these regional networks on the seizure focus and subsequent seizure evolution is not fully understood. To investigate these influences, measures of seizure complexity and propagation are studied using methods derived from time-frequency decomposition. We analyzed data from 61 partial seizures from 8 different patients with mesial temporal onset epilepsy (MTLE) and 117 seizures from 10 patients with neocortical lesional epilepsy (NLE) monitored in the epilepsy monitoring unit for pre-surgery evaluation with intracranial subdural grid arrays in 2000-2003. We applied the matching pursuit (MP) method developed by Mallat and Zhang (1993). The Gabor atom density (GAD) method (Jouny et al. 2003), derived from MP, provides a measure of signal complexity. Propagation index (PI) was defined by the average number of channels involved (GAD larger than threshold). The same threshold was used for all patients within the same group (MTLE,NLE) but different values were used for each group. GAD reveals that mesial temporal onset seizures and neocortical onset seizures have similar complexity patterns when comparing similar types of seizures. But overall MTLE seizures exhibit higher GAD values than NLE seizures (0.63[plusmn]0.29 versus 0.58[plusmn]0.17; p[lt]0.05). In both groups, partial seizures which secondarily generalized have longer duration at the focus than those that did not generalized. For all partial seizures, the extent of seizure propagation max(PI) and the signal complexity of the seizure GAD[sub]max[/sub] at the contact closest to the seizure focus were correlated for both groups. This correlation was stronger for NLE (R²=0.579; F=23.8; p[lt]0.05) than MTLE (R²=0.287; F=158.1; p[lt]0.05). Partial seizures that propagate regionally result in increased complexity of the ictal signal near the seizure focus. This suggests that regional networks are not merely passive pathways for seizure spread, but that these networks have the potential to influence the activity of the seizure focus itself. The possibility that such remote network influences may contribute to focal seizure dynamics and subsequent seizure evolution and spread should be considered. Similarly, interventions (e.g. stimulation) that affect these regional networks somewhat remote from the focus may have the potential to influence seizure evolution and duration. (Supported by NIH grant - NS 33732)