Detecting stumbles with a single accelerometer

Abstract

Falls are a common problem in the elderly population, and their prediction has been a major interest for the medical field. The relationship between stumbles and falls has not been very well understood yet. A critical requirement in advancing the study of this relationship is the realization of a realistic and effective stumble detection system. In this paper, we present a system for the detection of stumbles during walking. Our system consists of a single low cost triaxial accelerometer that may be worn by patients and is convenient for a wide range of subjects. We formulate the problem as an anomaly detection and we validate our system with a large data set collected from 9 subjects. The data set contains a total of 100 stumbles and 45 minutes of walking. We compare 7 different placements for the accelerometer, and show that our system achieves a 99% detection rate, with a 0.2% false alarm rate using an accelerometer worn on the chest.

Detection of Real-World Trips in At-Fall Risk Community Dwelling Older Adults Using Wearable Sensors

“…These peaks indicate changes in velocity, particularly abrupt changes during the swing phase. This is consistent with a recent study identifying trips as deviations from a statistical model for normal walking patterns ( 13 ). In some cases trip identification was supported by data from the contralateral foot sensor (demonstrating increased velocity of the step taken after the trip) and from the lower back sensor demonstrating a sharp increase in lower back tilt in the pitch axis.…”

Multi-scale Conditional Random Fields for first-person activity recognition on elders and disabled patients

“…Previous approaches use thigh or ankle for gait studies [2][3][4], or wrist and joints to monitor the progress of diseases [5,6]. Moreover, occasional events such as falls [7] or stumbles [8] have also been studied with accelerometers. In alternative approaches movements from the entire body have also been analysed [9,10].…”

An automatic data mining method to detect abnormal human behaviour using physical activity measurements

“…Classification based [15,40], clustering based [41,42] and statistical based [43,44] techniques are the main approaches used to detect anomalies. Clustering and statistical based techniques can be unsupervised and there is no need for training and labelling stages.…”