Hybrid System for Fall Detection & Fall Prevention

Abstract

<p>In the elderly population, falls are amongst the most frequent and dangerous causes of accidental injuries. It is estimated that more than 1 in 3 persons over the age of 65 are victims to falls each year[4][6]. Proceeding an accidental fall victims are often injured or immobilized. In the elderly population over 72 years of age, approximately 47% of the injured victims are unable to regain upright stance[9]. Without assistance, fallers often remain on the ground. Long-lie is the term used to describe fall-induced immobilization of victims for extended durations of time[7]. In 2004 it was estimated that 20% of fall-induced hospitalizations were associated with Long-lie[11]. In the same statistical report, Long-lie following a fall-induced injury is closely linked to cases of mortality amongst the elderly[11]. An automated system that can reduce duration of fall- induced Long-lie or even frequency of falls will therefore be greatly beneficial. The objective of this project is to engineer a fall-detection system that will i) accurately detect a fall, ii) prevent falls if possible by alerting the user of unstable physical orientations and iii) send out a distress signal to a handheld device or other communication devices wirelessly. The device is to be a compact unit worn around the waist and will communicate wirelessly via Bluetooth.</p>