Signals from real-world falls

Research overview

The text below is the source article’s original abstract under a verified reuse license. cc by

Objective: measurement of real-world fall events by using body-worn sensor devices can improve the understanding of falls in older people and enable new technology to prevent, predict, and automatically recognize falls. However, relative to the required recording time, these events are rare and hence challenging to capture. Therefore, the FARSEEING (FAll Repository for the design of Smart and sElf-adapaive Environments prolonging INdependent livinG) consortium and associated partners established a meta-database of signals from real-world falls. Until the end of 2014, 397 falls were measured and reported. This includes falls data from several settings and disease groups, mainly geriatric rehabilitation, Parkinson’s disease, cerebellar and sensory ataxia. Seventy-five per cent of the falls were measured with a sampling rate of 100 Hz with devices including at least accelerometers and gyroscopes. To date more than 100 of these real-world falls have been validated and finally processed for data analyses. The observed signal patterns showed a high heterogeneity and differed considerably from those of simulated falls. Preliminary analyses of the available real-world falls data with two different fall-detection approaches using wavelets as well as temporal and mechanical thresholds considerably improved the detection performance. The FARSEEING consortium will continue to increase the number of measured real-world falls in the meta-database beyond the end of the project. External users can request data access on the FARSEEING website.

Related topics