Classifying Spatial Navigation Performance in Healthy Adults Using Multimodal Behavioral Analysis of the Walking Corsi Test
Abstract
Cognitive navigation and spatial memory are essential for interacting with the environment, and their accurate assessment is increasingly important in cognitive research. This study presents a novel, non-invasive framework for evaluating spatial navigation abilities in healthy adults using the Walking Corsi Test (WalCT). Thirty-five participants performed the task while their movements were simultaneously recorded using camera-based pose estimation and IMU-based motion capture systems. A rich set of biomechanical features was extracted from these multimodal data, and participants also completed validated self-report questionnaires assessing spatial abilities. To classify individual navigation performance, two feature selection strategies were implemented: mutual information filtering followed by recursive feature elimination with cross-validation (MI-RFECV), and RFECV alone.
Type
Publication
IEEE Access
Status
Peer-reviewed
Open access
Authors
Mobina Zibandehpoor
(she/her)
M.Sc. Researcher
Hi, I’m Mobina, an M.Sc. mechatronics researcher passionate about non-invasive assessment of human cognitive performance. My work sits at the intersection of computer vision, biosignal analysis, and machine learning, developing multimodal frameworks that combine vision-based pose estimation, IMU motion capture, and eye tracking for interpretable behavioral analysis. I enjoy designing algorithms and experimental systems that connect mechatronic instrumentation with real-world cognitive assessment.
Please reach out if you’re interested in collaborating on Human-Computer Interaction, biosignal analysis, or AI-driven behavioral assessment!😜