Accelerometer data analysis for identification of seniors at risk of falling

Abstract

Background: Falls in the elderly are a public health problem due to its high incidence and severity of consequences and costs. Acceleration sensor-based equipment represent an option for the prevention and monitoring of elderly fallers. However, there is a gap in the literature concerning the analysis of accelerometer data of older persons. Database on gait of the elderly, as well as new approaches in the analysis of the accelerometer curves, could initiate discussions about the interpretation and use of these devices for health care and prevention. Objective: This paper attempts to test a triaxial accelerometer sensor potential to detect risks of falling, and to investigate the possible patterns that could track fall risks in elderly with the six-minute walk test. Method - Design: Exploratory sectional study. Selection: A convenience sample of 30 community-based volunteers. Groups: non-fallers (10), recurrent fallers (10) and fallers (10). Device: A triaxial accelerometer sensor coupled to a small development board, and a computer software to process the output data and the risk of falling into a patient. Procedures: 30 elderly volunteers will conduct the six-minute walk test using the device attached to the waist (center of mass). Statistical tests and signal processing techniques will be used to analyze differences between the acceleration curves of individuals belonging to the three groups. Expected Results: We expect to obtain data from a triaxial accelerometer sensor indicating differences between the patterns of fallers and non-fallers elderly, and it could contribute for the future development of an available equipment and forms of accelerometer data processing. In addition, a real pattern database of movement and fall risk of senescents individuals could assist researchers and professionals from Brazil and in the world, contributing for programs development, research and policies for the fall prevention.