Innovative AI-Driven Radar System Detects Subtle Changes in Patient Mobility

Researchers at the University of Waterloo have developed a cutting-edge radar and artificial intelligence (AI) system designed to monitor the walking patterns of individuals in busy healthcare environments such as hospitals and long-term care facilities. This technology offers a non-invasive way to identify early signs of health decline by analyzing gait and mobility through radar signals. The compact device, roughly the size of a deck of cards and mounted on a wall, emits low-power radio waves that bounce off moving persons, creating heatmaps that visualize movement. Advanced AI algorithms process these heatmaps to isolate individual movements, filter out noise, and calculate walking speeds with high precision.

The significance of monitoring walking speed stems from its role as a 'functional vital sign.' Even minor reductions in gait speed can signal emerging health issues, including frailty or neurological decline. By continuously tracking these subtle changes, the system can alert healthcare providers early, enabling timely intervention.

The system was tested with older adults who had been on 14 days of strict bed rest, simulating recovery from serious illness or long flight-induced deconditioning, as part of a collaboration with the Canadian Space Agency. The results demonstrated the system's ability to detect nuanced variations in gait, emphasizing its potential for early diagnosis and ongoing health monitoring.

Building on prior work in fall detection, the radar-based system operates effectively regardless of lighting conditions and preserves patient privacy, as it does not require cameras or wearable sensors. This innovative approach opens up new possibilities for non-contact health surveillance.

Hajar Abedi, the project's lead scientist and now chief scientist at GoldSentinel, envisions this technology as a silent 'invisible safety net' that continuously monitors residents’ mobility and behavioral changes. The goal is to provide caregivers with early alerts long before crises occur, enhancing patient safety and improving healthcare outcomes.

Published in Scientific Reports, this research highlights a significant step forward in non-invasive health monitoring techniques, with promising applications for aging populations and individuals with mobility impairments.

For more details, visit: Scientific Reports

*Source: https://medicalxpress.com/news/2025-10-ai-radar-tracks-subtle-health.html