Abstracts

RATIONALE: Independent component analysis(ICA)is a new method of analyzing EEGs and there have been reports on successful artifact removal and interictal source localization. We wanted to know whether this method could also be applied to the ictal EEGs for an accurate source localization. METHODS: Eleven ictal EEGs from 7 right medial temporal lobe epilepsy(TLE) patients and 13 EEGs from 7 left TLE patients were analyzed. Digital EEG signals were recorded using the international 10-20 systemwith additional T1 and T2 tracings.Twenty tracings were analyzed by ICA algorithm into independent components. Among them, the tracings which showed changes into rhythmic waves and evolved were considered to represent ictal rhythm and their sources were localized by referring them to the mapping results. RESULTS: A considerable portion of artifacts and ictal rhythms was successfully separated. Among 11 EEGs from the right TLE patients, anterior temporal localization was possible in 5 EEGs and the frontal and temporal independent ictal sources were found in 4 EEGs. The remaining 2 EEGs showed bitemporal ictal rhythm. In 13 EEGs from the left TLE patients, anterior temporal localization was possible in 5 EEGs and frontal and temporal independent ictal sources were found in 4 EEGs. Two EEGs showed ipsilateral diffuse temporal onset and the remaining 2 showed bitemporal onset. CONCLUSIONS: Some ictal rhythms seemingly originating from one source in the conventional EEG are actually mixtures of waves from multiple ictal generators. This study was supported by the fund from Korean Ministry of Health and Welfare.