Beyond the Clinic: Enhancing Fall Risk Prediction through the BE-FIT Built Environment Framework

Introduction: Traditional fall risk assessments are predominantly conducted in controlled clinical settings, often failing to capture the complex interaction between an individual’s gait and real-world environmental stressors. Current predictive models largely ignore extrinsic factors, such as surface irregularities or visual distractions. The ongoing BE-FIT project aims to bridge this gap by validating a novel, ecological fall risk prediction score that integrates intrinsic biomechanical markers with extrinsic built environment features.

Methods: This ongoing study analyzes data from a growing cohort of community-dwelling older adults (current N=20). Participants undergo a standardized walking protocol across diverse urban terrains (e.g., flat pavement, slopes, uneven surfaces). To capture the multifaceted nature of fall risk, we employ a multimodal sensor fusion approach:

1. Biomechanics: Wearable Inertial Measurement Units (IMUs) record gait kinematics and spatiotemporal parameters.
2. Context & Gaze: Synchronized scene cameras and eye-tracking goggles capture visual attention and environmental obstacles, while GPS provides geospatial localization. Data is processed through a custom pipeline to synchronize streams and extract key features.

Expected Outcomes: We hypothesize that the "Combined BE-FIT Model" will demonstrate significantly higher sensitivity and specificity than physiological models alone. Specifically, we expect that environmental features—such as changes in gradient or high-traffic density—will correlate with quantifiable markers of instability (e.g., increased gait variability), which are visualized via multi-domain spider plots to characterize individual risk profiles.

Conclusion: This study highlights the necessity of incorporating environmental context into fall risk stratification. Upon completion, the BE-FIT prediction score aims to deliver a robust, ecologically valid screening tool, providing actionable data for both clinical interventions and evidence-based urban planning to create age-friendly environments.