A Gamified Framework for Multimodal Memory Assessment in Adults and Children with Epilepsy
Abstract number :
2.499
Submission category :
11. Behavior/Neuropsychology/Language / 11B. Pediatrics
Year :
2025
Submission ID :
1411
Source :
www.aesnet.org
Presentation date :
12/7/2025 12:00:00 AM
Published date :
Authors :
Presenting Author: Taylor Shade, BS – Emory University School of Medicine
Noah Okada, BS – Cal Tech
Carla Ammons, PhD – Childrens Healthcare of Atlanta
Evan Brady, BS – Emory University School of Medicine
Lydia Swinehart, BS – Emory University School of Medicine
Molly Winston, PhD – Childrens Healthcare of Atlanta
Nealen Laxpati, MD, PhD – Emory University School of Medicine
Joshua Chern, MD – Childrens Healthcare of Atlanta
Timothy Gershon, MD – Emory University
Kartik Reddy, MD – Emory University
Donald Bearden, PhD, ABPP-CN – University of Tennessee Health Science Center and Le Bonheur Children's Hospital
Daniel Drane, PhD – Emory University School of Medicine
Rationale: Consistent disagreement between functional outcomes and neuropsychological performance among epilepsy patients suggests that traditional cognitive assessments cannot fully evaluate function. Recent technological advancements have enabled the creation of sophisticated cognitive assessments that simulate real-world scenarios to better understand and evaluate human memory functions. Leveraging these technologies, we previously developed a gamified assessment to study critical aspects of memory across multiple modalities that are at risk in adults with epilepsy. We present preliminary data on a recently developed pediatric version of this multimodal task.
Methods: Using the Unity Game Engine, we developed a stand-alone application to administer a multimodal memory assessment, the Emory Pediatric Multimodal Learning Test (EPMLT). The task consisted of two blocks: a learning block and delayed recall block, where subjects view professionally developed videos of individuals engaged in various activities each in a unique setting. Subjects must recall traditional verbal material, but also multifaceted unfolding visual scenes depicting complex stories that unfold over time (visual narrative). This task uses a virtual delivery platform, allowing for any recall delay, and using machine learning for automated transcription, scoring, and interpretation. We have collected data at the immediate, 30-min and 1-week intervals for 10 pediatric epilepsy patients, 40 adult epilepsy patients, and 15 healthy controls. Performance between subgroups was assessed with a mixed design ANOVA, and within group differences were assessed with paired t-tests.
Results: Healthy controls perform better than children and adults with epilepsy across all facets of the EPMLT (p< .001).
Behavior