Localization of seizure onset area from intracranial non-seizure EEG by exploiting locally enhanced synchrony

For as many as 30% of epilepsy patients, seizures are poorly controlled with medication alone. For some of these patients surgery may be an option: the brain region responsible for seizure onset may be removed surgically. However, this requires accurate delineation of the seizure onset region. Currently, the key to making this determination is seizure EEG. Therefore, EEG recordings must continue until enough seizures are obtained to determine the onset region; this may take about 5 days to several weeks. In some cases these recordings must be done using invasive electrodes, a procedure that includes substantial risk, discomfort and cost. In this paper, techniques are developed that use periods of intracranial non-seizure ("rest") EEG to localize epileptogenic networks. Analysis of intracranial EEG (recorded by surface and/or depth electrodes) of 6 epileptic patients shows that certain EEG channels and hence cortical regions are consistently more synchronous ("hypersynchronous" ) compared to others. It is shown that hypersynchrony seems to strongly correlate with the seizure onset zone; this phenomenon may in the long term allow to determine the seizure onset area(s) from non-seizure EEG, which in turn would enable shorter hospitalizations or even avoidance of semi-chronic implantations all-together.