Accelerometer-based analysis of gait and prediction of fall risk in older adults

Falls are the most common cause of injury in older adults with two-thirds of individuals over the age of 65 falling at least once a year. It is well known that falls represent a significant challenge to preserving quality of life as we age, but current clinical methods of screening for fall risk remain insufficient to prevent falls. This thesis summarizes the development of a modern approach to fall risk analysis and fall prevention through the use of hip-mounted triaxial accelerometers to passively monitor gait quality in free-living environments and predict risk of future falls. Data from over 4000 individuals enrolled in the Women Health Initiative’s Objective Physical Activity and Cardiovascular Health study were used for the development of an activity recognition pipeline for extraction of free-living walking bouts measured by accelerometers. A variety of measures of gait were computed from walking bout data and used as input to train statistical models which analyze gait to predict fall risk and future falls. Results suggest that hip-mounted accelerometers are able to capture free-living gait patterns which can be used to predict measures of fall risk and physical function such as the Short Physical Performance Battery. However, these same measures of gait prove to be insufficient for direct prediction future falls.

Name thesis:degree_name

M.S.

Level thesis:degree_level

Thesis

Discipline thesis:degree_discipline

Bioengineering

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2019

Author dc:creator

• Quicksall, Zachary Stokes

Contributors dc:contributor

• Schatz, Bruce R.

• falls
• fall risk
• accelerometer
• older adults
• SPPB
• prediction
• OPACH
• machine learning

Statement dc:rights

• Copyright 2018 Zachary S. Quicksall

Language dc:language

en

Handle dc:identifier

http://hdl.handle.net/2142/102899

OAI identifier oai:identifier

oai:www.ideals.illinois.edu:2142/102899