Frequency evolution during tonic-clonic seizures.

By using the Short Time Fourier Transform, we analyzed the EEG frequency evolution during tonic-clonic seizures on 18 scalp recordings corresponding to 7 patients admitted for Video-EEG monitoring. This information was correlated with clinical findings observed in the video recordings. From the time-frequency plots, we recognized patterns related with brain activity even when embedded in a background of muscle artifacts. In 13/18 seizures we found a clear frequency dynamics characterized by an activity originally localized at about 8 Hz, later slowing down to about 1.5 Hz. In the remaining cases muscle artifacts hinder the disclosure of a clear frequency evolution. The clonic phases started when the main frequency slowed down to about 3 Hz. We conclude that the Short Time Fourier Transform is very useful for a quantitative analysis of epileptic seizures, especially when muscle artifacts contaminate the recordings. We further conclude that the clonic phase starts as a response to brain activity that can be only established when brain oscillations are slow enough to be followed by the muscles.