An individual-specific fall detection model based on the statistical process control chart

Abstract

Falls remain one of the leading causes of fatal and nonfatal injuries in many countries. Fall detection is an important method to protect fallers by minimizing injury severity. There are some common limitations in existing fall detection models. In particular, the fall indicators and detection thresholds were arbitrarily predetermined without any theoretical and/or experimental basis, and most fall detection models cannot address inter-individual differences. This study presents a novel pre-impact fall detection model based on the statistical process control chart that is able to address the existing limitations. The fall indicators in this model were selected based on experimental findings. The fall detection model is individual-specific, since it is constructed using individual historical movement data. The fall detection model demonstrates a high accuracy with up to 94.7% sensitivity and 99.2% specificity. In addition, this model can also provide sufficient time for triggering fall protection device in the pre-impact phase, thus efficient in preventing fall injuries.