Seizure provocation in EEG recordings: A data-driven approach

Area of Science:

• Clinical Neurophysiology
• Epileptology
• Diagnostic Imaging

Background:

• Video-electroencephalography (video-EEG) is crucial for epilepsy diagnosis.
• Provocation methods are used to elicit seizures during video-EEG.
• Bilateral tonic-clonic seizures are a potential risk during these procedures.

Purpose of the Study:

• To evaluate the diagnostic yield of various provocation methods in short-term video-EEG.
• To assess the risks associated with these provocation techniques, particularly bilateral tonic-clonic seizures.

Main Methods:

• Data were extracted from a large database of short-term video-EEGs with standardized annotations (SCORE).
• Analysis included 2742 paroxysmal clinical episodes from 11,919 EEGs.
• Provocation methods studied included hyperventilation, intermittent photic stimulation (IPS), and sleep.

Main Results:

• 54% of epileptic seizures were provoked.
• Hyperventilation was effective for absence seizures (55%).
• IPS provoked myoclonic (25%) and most bilateral tonic-clonic seizures (55%).
• Sleep precipitated 43% of focal seizures.
• Bilateral tonic-clonic seizures (16 total) were primarily provoked by IPS or sleep.
• Latency to bilateral tonic-clonic seizures after photoparoxysmal EEG response was ≤3 seconds in most cases.

Conclusions:

• Provocation methods significantly enhance the diagnostic utility of short-term video-EEG.
• The risk of bilateral tonic-clonic seizures is low but cannot be mitigated by a 3-second cutoff for IPS.
• Tailoring provocation strategies based on seizure semiology is recommended for optimal diagnostic yield.