Comparison of low-complexity fall detection algorithms for body attached accelerometers

• Published in: Gait & posture Vol. 28; no. 2; pp. 285 - 291
• Main Authors: Kangas, Maarit, Konttila, Antti, Lindgren, Per, Winblad, Ilkka, Jämsä, Timo
• Format: Journal Article
• Language: English
• Published: England Elsevier B.V 01.08.2008
• Subjects: Accidental Falls; Adult; Algorithms; Elderly; Embedded System; Female; Hip fracture; Humans; Inbyggda system; Independent living; Male; Middle Aged; Monitoring, Physiologic - instrumentation; Movement analysis; Orthopedics; Sensitivity and Specificity; Movement analysis; Hip fracture; Elderly; Independent living
• ISSN: 0966-6362; 1879-2219; 1879-2219
• DOI: 10.1016/j.gaitpost.2008.01.003

The elderly population is growing rapidly. Fall related injuries are a central problem for this population. Elderly people desire to live at home, and thus, new technologies, such as automated fall detectors, are needed to support their independence and security. The aim of this study was to evaluate different low-complexity fall detection algorithms, using triaxial accelerometers attached at the waist, wrist, and head. The fall data were obtained from standardized types of intentional falls (forward, backward, and lateral) in three middle-aged subjects. Data from activities of daily living were used as reference. Three different detection algorithms with increasing complexity were investigated using two or more of the following phases of a fall event: beginning of the fall, falling velocity, fall impact, and posture after the fall. The results indicated that fall detection using a triaxial accelerometer worn at the waist or head is efficient, even with quite simple threshold-based algorithms, with a sensitivity of 97–98% and specificity of 100%. The most sensitive acceleration parameters in these algorithms appeared to be the resultant signal with no high-pass filtering, and the calculated vertical acceleration. In this study, the wrist did not appear to be an applicable site for fall detection. Since a head worn device includes limitations concerning usability and acceptance, a waist worn accelerometer, using an algorithm that recognizes the impact and the posture after the fall, might be optimal for fall detection.