Non-contact radar assessment of One-Legged Stand Test for fall risk in aging

BACKGROUND: Falls are a leading cause of morbidity and mortality in older adults, requiring better longitudinal assessment tools. While established clinical methods such as the One-Legged Stand Test (OLST) are effective, they rely on supervised, contact- or video-based systems and are typically performed only annually, missing changes in fall risk between visits. OBJECTIVE: We evaluated a privacy-preserving Frequency-Modulated Continuous Wave (FMCW) radar system's ability to autonomously proctor the OLST and assess postural instability during the test, compared with simultaneously collected gold-standard measurements of force plates and motion capture (MOCAP). METHODS: In a cross-sectional study of 32 healthy adults (15 young ≤32 y, 17 older ≥64 y), synchronized radar, force plate, and motion capture data were collected during short (4 s) and long (20 s) OLST trials. A convolutional neural network-long short-term memory-attention (CNN-LSTM-Attention) model was trained on radar-derived range-Doppler maps to classify OLST phases, with performance evaluated against force plate and MOCAP ground truth. All de-identified data and visualization code are publicly available on PhysioNet. RESULTS: The model detected Foot-Up (FU) and Foot-Down (FD) events with 94.0% and 90.7% accuracy, respectively, improving to 96.9% and 94.5% after logical post-processing. In long trials, radar-derived Doppler wobble metrics during the stability phase strongly correlated with OLST duration (R²=0.58, p<0.001), and aligned with similar trends in force plate (COP ellipse area, R²=0.72, p<0.001) and MOCAP (trunk pitch-period IQR, R²=0.65, p<0.001) metrics. In short trials, radar-derived features separated the three study-defined cohorts, with significant pairwise differences (p<0.05) and large effect sizes (d=1.13-2.72). CONCLUSION: FMCW radar accurately detects OLST transitions and captures wobble signatures linked to neuromuscular control and fall risk. These findings establish proof-of-concept for non-contact radar assessment of balance in older adults and support further clinical validation for at-home monitoring.