Veerubhotla, Akhila
(2019)
How Well Can Wearable Devices Measure Physical Activity in Manual Wheelchair Users with Spinal Cord Injury.
Doctoral Dissertation, University of Pittsburgh.
(Unpublished)
Abstract
Manual wheelchair users (MWUs) with spinal cord injury (SCI) are at the lower end of the physical activity (PA) spectrum. Physical inactivity combined with physiological changes after SCI puts them at an increased risk for secondary health conditions such as cardiometabolic and cardiorespiratory diseases. Increased PA and decreased sedentary time have been shown to contribute to the primary and secondary prevention of chronic health conditions. The objective measurement of habitual PA is an important part of health promoting efforts to address physical inactivity. While wearable devices are believed to be best suited for objective measurement of PA in everyday life and are widely used in the ambulatory population, most commercially available devices do not cater to people with SCI. Currently there are limited validated tools that can provide comprehensive PA measures relevant to this population. This dissertation developed and evaluated custom algorithms for assessing activity intensities and wheelchair propulsion characteristics using off-the-shelf wearable devices in people with SCI. The data collection included a wide range of activities of daily living and exercises from lab-based structured protocols to a one-week community-based trial. Both traditional statistical modeling and machine learning techniques were used to develop the custom algorithms, and a rigorous procedure was followed to evaluate these algorithms, which includes not only iterative cross-validation, but also out-of-sample validation where a separate data set from the training data was used to evaluate the algorithm performance. The developed algorithms along with wearable devices from this dissertation could help researchers fill a number of voids in health and PA research targeting people with SCI and other mobility impairments.
Share
| Citation/Export: |
|
| Social Networking: |
|
Details
| Item Type: |
University of Pittsburgh ETD
|
| Status: |
Unpublished |
| Creators/Authors: |
|
| ETD Committee: |
|
| Date: |
16 September 2019 |
| Date Type: |
Publication |
| Defense Date: |
22 July 2019 |
| Approval Date: |
16 September 2019 |
| Submission Date: |
13 September 2019 |
| Access Restriction: |
5 year -- Restrict access to University of Pittsburgh for a period of 5 years. |
| Number of Pages: |
119 |
| Institution: |
University of Pittsburgh |
| Schools and Programs: |
School of Health and Rehabilitation Sciences > Rehabilitation Science and Technology |
| Degree: |
PhD - Doctor of Philosophy |
| Thesis Type: |
Doctoral Dissertation |
| Refereed: |
Yes |
| Uncontrolled Keywords: |
machine learning, signal processing, algorithm, manual wheelchair users, activity monitoring, physical activity, exercise, measurement |
| Date Deposited: |
26 Aug 2020 12:43 |
| Last Modified: |
26 Aug 2020 12:43 |
| URI: |
http://d-scholarship.pitt.edu/id/eprint/37635 |
Metrics
Monthly Views for the past 3 years
Plum Analytics
Actions (login required)
 |
View Item |
![[img]](https://d-scholarship.pitt.edu/style/images/fileicons/application_pdf.png)
![[feed]](/images/feed-icon-32x32.png){kind=icon}