Towards understanding on the development of wearable fall detection: an experimental approach

Falls are one of the main public health threats to the elderly. One approach to prevent the severe effect of falling is to detect those involving falling. Recent advancement in sensing technology offers the possibility of the objective. The effectiveness of fall detection was based on the detection accuracy to differentiate the occurrence of falls. In this regard, the dataset obtained constitutes the basis for the performance development of fall detection techniques. Hence, this present systematic review set out to assemble and analyze the dataset acquisition process on providing objective fall detection assessment in older adults. A systematic review was conducted by adopting Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA). The analysis of the datasets providing process is performed comprehensively, taking account of the multiple factors involved in the definition of the dataset’s preparation. The analysis comes out with few main themes: development of dataset and elderly involvement, type of sensor, sensor placement and emulated activity. Overall, these devices have the potential to provide an accurate fall detection assessment. This systematic review of the traces brings to light the lack of a common experimental benchmarking procedure.