Towards automatic detection of falls using wireless sensors

Area of Science:

• Biomedical Engineering
• Sensor Networks
• Geriatric Care Technology

Background:

• Falls pose significant risks, especially for the elderly, necessitating rapid medical intervention.
• Existing fall detection methods often lack accuracy or real-time capabilities.
• Wireless sensor networks offer a promising avenue for unobtrusive and continuous monitoring.

Purpose of the Study:

• To develop and evaluate a wireless sensor network system for automatic fall detection.
• To enhance the accuracy and reliability of fall detection using a multi-sensor approach.
• To provide a system capable of real-time fall detection for prompt medical assistance.

Main Methods:

• A system integrating body-worn triaxial accelerometers and area-based motion detectors was designed.
• Data transmission utilized the IEEE 802.15.4 protocol with Carrier Sense Multiple Access-Collision Avoidance for channel reuse.
• A two-stage algorithm processed acceleration and motion data sequentially for fall detection.

Main Results:

• The first stage identified potential falls using normalized energy expenditure from acceleration data.
• The second stage confirmed falls by detecting the absence of motion.
• Simulated fall evaluations with 15 subjects demonstrated the system's high promise for real-time detection.

Conclusions:

• The proposed wireless sensor network system offers a reliable solution for automatic fall detection.
• The combination of accelerometers and motion detectors improves fall detection accuracy.
• This technology has the potential to significantly improve response times for fall-related injuries.