A jerk threshold-based Involuntary Lateral Movement algorithm

Algorithms for automatic fall detection are often studied in the field of ambulatory human health supervision. These algorithms are developed to generate hospital emergency alarms. In the present paper, involuntary lateral movements (ILM) are presented. ILM are a premature sign of health deterioration. Therefore, this algorithm embedded in a sensor device could be used for continuous health monitoring in ambulatory situations. Several studies show that human bodies try to minimize acceleration body movements, so they are based on minimum jerk. The proposed algorithm is based on a jerk threshold detection. In this work it is supposed that ILM will produce important jerk values above other daily movements, so that they can be distinguished using a threshold.