Abstract
OBJECTIVE: There is evidence for a pre-ictal state, distinct from the interictal state, that can be detected minutes before the onset of a clinical seizure. Analysis of scalp EEG signals do not reliably distinguish the pre-ictal EEG from the interictal EEG. Changes in cerebral blood flow and oxygenation occur several minutes before EEG evidence of a seizure. Local increases in cerebral activity are associated with increases in brain temperature. Brain temperature changes are reflected in the tympanic membrane temperature. This exploratory study investigated whether a brain temperature change could be detected noninvasively in the immediate pre-ictal period. METHODS: Patients with focal seizures undergoing inpatient video-EEG telemetry had epitympanic temperatures recorded. Pre-ictal temperature changes deviating from the interictal temperature were studied. RESULTS: Data was available for 25 seizures in 12 patients. For a given seizure, the mean temperature at seizure onset exceeded the interictal temperature by 0.31 °C. Peri-ictal peak temperature exceeded the interictal temperature by a mean of 0.37 °C. Duration of temperature rise was 1081 s. CONCLUSIONS: Pre-ictal temperature increases were detected non-invasively in focal onset seizures. SIGNIFICANCE: This finding provides the basis for development of temperature-based technology for seizure warning in ambulatory patients with refractory epilepsy.