Introduction
Diagnosing epilepsy is often time-consuming, partially due to the limited sensitivity of the routine electroencephalogram (EEG). Therefore, there is a need for additional diagnostic measures. There is usually a small brain area responsible for the seizure onset, although it cannot always be localized. Transcranial magnetic stimulation (TMS) enables quantification of the brain’s excitability. Previous studies have shown an increased excitability in epilepsy patients (Badawy et al., 2010). When TMS is applied while recording EEG, a characteristic waveform-the TMS evoked potential (TEP)-is induced in the EEG. A previous study showed that TEP consists of an early part, which is always present, and a late part, that was present in 9 out of 11 epilepsy patients, and not in healthy subjects (Valentin et al., 2008).

Objectives
To investigate late TEP responses and the spread of induced activity over the cortex in healthy subjects and epilepsy patients.

Materials and methods
TMS/EEG was recorded in healthy controls and adult epilepsy patients using a Magstim Rapid2 stimulator and a 64-channel EEG amplifier (ANT Neuro, Enschede). TMS was targeted at the left and right motor cortex. We administered 75 pulses at an intensity of 110% motor threshold for both targets. The TEP was obtained by averaging over all TMS pulses, and the baseline power was then subtracted from the late response power. The values for all trials before and after the TMS pulse were then compared using a student t-test. Increases in power of >1 μV in the 9 electrodes surrounding the stimulation point, with a significance level of p < 0.01, were regarded as a late response.

Results
At present, 18 healthy subjects (11 males, mean age 28 years) and 10 epilepsy patients (3 males, mean age 24 years) have been included. Nine patients were taking anti-epileptic drugs. In all healthy controls and epilepsy patients we found an early TEP, and three patients and five healthy subjects showed a late response.

Conclusion
Initial results show that the late responses are not sufficient to reliably differentiate between healthy subjects and epilepsy patients. We are currently analysing the activity spread data. In addition, more patient measurements have been scheduled.