Mapping Postictal Aphasia through Signal Complexity: A Stereo-Electroencephalography Study

The postictal period provides an opportunity to investigate the pathophysiology underlying aphasia and recovery following epileptic seizures. This study examines postictal aphasia in stereo-electroencephalography (SEEG)-explored patients to identify brain regions associated with task-specific language deficits using signal complexity analysis.

We evaluated video-SEEG-recorded focal seizures with and without postictal aphasia in patients with SEEG-confirmed hemispheric language dominance. SEEG traces were analyzed using permutation entropy (PE), with the postictal period quantified by a PE-based metric, the Postictal Alteration Time (PAT). Brain region PAT was correlated with language function recovery (eg, repetition). Electro-clinical recuperation was also assessed within the dorsal-ventral language stream framework. Additionally, a bedside testing battery was developed to evaluate postictal aphasia severity and task-specific deficits.

A total of 322 seizures from 98 patients were analyzed. Seizures with postictal aphasia had longer PAT than those without. Task-specific language recovery correlated with regional PAT (eg, naming - middle temporal gyrus). Moreover, the dorsal stream recovered faster than the ventral stream. Additionally, the Postictal Aphasia Scale (PAS) was developed, evaluating naming, reading, repetition, and comprehension (verbal and written) and automatic speech. Higher PAS scores (indicating milder deficits) correlated with faster regional complexity recovery. At 5 and 10 minutes postictally, PAS revealed a global aphasia pattern, with comprehension deficits gradually resolving. By 15 minutes, aphasia was primarily production-related, particularly affecting naming.

This study provides new insights into the pathophysiology of postictal aphasia and introduces PAS as a tool for assessing postictal aphasia severity and domain-specific deficits, aiding surgical planning and rehabilitation. ANN NEUROL 2025;98:1080-1095.