We compared episodic memory performance in patients with focal epilepsy to healthy controls (HCs) using a naturalistic audiovisual task.

Continuous audiovisual information presented in films resembles the constant stream of information humans process and remember every day. Event segmentation, the chunking of continuous experience into meaningful units, is a robust feature of human cognition and facilitates episodic memory. Within this framework, we characterized film recall performance, as measured by event and sequence recall in surgical epilepsy patients, patients with well-controlled epilepsy, and healthy subjects. We hypothesized that patients with poorly controlled seizures involving the hippocampus would demonstrate poorer event and event sequence recall compared to the other subject groups.

Three groups of subjects are being recruited: surgical epilepsy patients (N=15), patients with well-controlled epilepsy (N=15), and HCs (N=15). Subjects are eligible if they are (1) between the ages of 18-65; (2) fluent in English; and (3) are capable of informed consent. All subjects will (a) watch 3 blocks of 5 short films, each 2-5 minutes in length, and then (b) describe each film in as much detail as possible. The spoken recall data will be
compared with a transcription of the actual film event log and scored by a reviewer blinded to patient diagnosis. Percentage event and event sequence recall will be compared across the 3 subject groups with a one-way ANOVA with a post-hoc t-test comparison.

Our preliminary findings indicate that temporal lobe epilepsy patients with wellcontrolled seizures demonstrated similar event and film level performance compared to HCs but worse event sequence memory. Data collection for surgical epilepsy patients is ongoing.

Naturalistic tasks like film event recall may provide an ecologically valid and sensitive measure of episodic memory. In particular, event sequence memory may be a sensitive indicator of hippocampal dysfunction.