Exploration and implementation of a pre-impact fall recognition method based on an inertial body sensor network

Area of Science:

• Biomedical Engineering
• Gerontology
• Wearable Technology

Background:

• Falls in the elderly cause significant health and economic burdens.
• Early fall detection and self-protective measures can mitigate fall-related injuries.
• The aging global population necessitates advanced fall prevention strategies.

Purpose of the Study:

• To develop and implement a pre-impact fall recognition and alarm system.
• To analyze fall dynamics using an inertial body sensor network.
• To identify optimal sensor placement and thresholds for fall prediction.

Main Methods:

• Utilized an inertial body sensor network (nine MTx sensor modules) to capture kinematic data.
• Subjects performed daily living and simulated fall activities (forward, sideway, backward).
• Analyzed body segmental kinematic features, pre-impact lead time, and postural stability angles.

Main Results:

• The chest was identified as the optimal sensor location for both pre-impact alarm and post-fall detection.
• Established acceleration thresholds: 7 m/s/s for pre-impact alarm and 20 m/s/s for post-fall detection.
• Determined critical postural stability angles and pre-impact lead times for different fall directions, with backward falls being the most challenging to avoid.

Conclusions:

• A chest-mounted inertial sensor system can effectively provide early fall prediction and detection.
• Specific acceleration thresholds and analysis of postural stability are key for an accurate alarm system.
• Backward falls present the highest risk due to short lead times and small stability angles, requiring further research for effective prevention.