Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

Analysis of the Effects of Pre-processing and Dual-Tasking on Gait Accelerometry Signals

Millecamps, Alexandre (2014) Analysis of the Effects of Pre-processing and Dual-Tasking on Gait Accelerometry Signals. Master's Thesis, University of Pittsburgh. (Unpublished)

Primary Text

Download (1MB) | Preview


One-third of older adults over 65 years of age fall each year. Falls are the main cause of injury and death among this population. Understanding the causes of falling is therefore a necessity for gerontologists. Gait accelerometry is an important approach for the quantitative assessment of human walking. It is an inexpensive, portable and reliable method to measure trunk accelerations. The latter may give indications on balance control even though there is no agreed measure of it. Accelerometry requires the measured accelerations to be preprocessed, but previous studies have not studied thoroughly its effects on signal features. We therefore constituted a set of features in the time, frequency and time-frequency domains and we evaluated the impact of tilt correction and wavelet denoising - two pre-processing operations - on these features.

Signals used in this thesis were collected on 35 participants aged 65-year-old and over: 14 were healthy controls (HC), 10 had Parkinson's disease (PD) and 11 had peripheral neuropathy (PN). They walked on a treadmill at preferred speed. We first applied tilt correction and wavelet denoising separately, then we applied operations one after another. Denoising had nearly no effect on features compared to the raw accelerations. Tilt correction led to better discrimination between groups. Joint pre-processing operations showed trends that were similar to the tilt correction alone.

Older adults also face increasing difficulties to perform an activity during walking and this threatens their stability. Thus, during another trial, the same groups of subjects were asked to press a button at hearing a tone. We observed the impact of dual-tasking on the features. Several features such as Lempel-Ziv complexity, bandwidth of accelerations and harmonic ratios remained unaffected by dual-task walking while the remaining features were affected. We also examined the impact of dual-tasking on each group. Task differences were almost the same for every group and revealed lower harmonic ratios during dual-task walking.


Social Networking:
Share |


Item Type: University of Pittsburgh ETD
Status: Unpublished
CreatorsEmailPitt UsernameORCID
Millecamps, Alexandrealm257@pitt.eduALM257
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Committee ChairSejdić, Ervinesejdic@pitt.eduESEJDIC
Committee MemberChaparro, Luis F.lfch@pitt.eduLFCH
Committee MemberLi, Ching-Chungccl@pitt.eduCCL
Thesis AdvisorSejdić, Ervinesejdic@pitt.eduESEJDIC
Date: 29 January 2014
Date Type: Publication
Defense Date: 22 November 2013
Approval Date: 29 January 2014
Submission Date: 25 November 2013
Access Restriction: 5 year -- Restrict access to University of Pittsburgh for a period of 5 years.
Number of Pages: 72
Institution: University of Pittsburgh
Schools and Programs: Swanson School of Engineering > Electrical Engineering
Degree: MS - Master of Science
Thesis Type: Master's Thesis
Refereed: Yes
Uncontrolled Keywords: Gait accelerometry, pre-processing effects, dual-task walking, signal features
Date Deposited: 29 Jan 2014 15:11
Last Modified: 29 Jan 2019 06:15


Monthly Views for the past 3 years

Plum Analytics

Actions (login required)

View Item View Item