OJEDA AGUILAR, ALEJANDRA MANOELA
(2013)
TOWARDS MONITORING WHEELCHAIR PROPULSION IN NATURAL ENVIRONMENT USING WEARABLE SENSORS.
Master's Thesis, University of Pittsburgh.
(Unpublished)
Abstract
Due to lower limb paralysis, individuals with spinal cord injury (SCI) rely on their upper limbs for activities of daily living (ADLs) and wheelchair propulsion (WP). Previous research has found that specific biomechanical parameters of WP are associated with risk of UE pain and injury. However, the repetitiveness and quality of upper limb movements during WP are unclear. Recently, wearable sensors have been used to collect mobility characteristics of wheelchair users, but little research has looked into using them to monitor the quality of UE movements for WP in the natural environment. The purpose of this thesis was to develop and evaluate a WP monitoring device that can monitor wheelchair users’ activities, and propulsion parameters in the natural environment.
This thesis is organized into three studies. The first study aims to develop activity classifiers that can distinguish WP episodes from a range of ADLs. Two classifying models were built using a Machine Learning (ML) technique. The model that yielded the highest accuracy showed an overall accuracy of 88.0%. Time spent on each activity was estimated based on the classifiers, and compared with the video observation. Percentage of difference between the criterion and estimated time ranged from 2.2% to 11.6%.
The second study aims to estimate temporal parameters of WP, including the stroke number (SN) and push frequency (PF), using wearable sensors. The estimated SN and PF were compared with the criterion measures using the mean absolute errors (MAE) and mean absolute percentage of error (MAPE). Intraclass Correlation Coefficients were calculated to assess the agreement. The accelerometer placed on the upper arm yielded the highest accuracy with the MAPE of 8.0% for SN and 12.9% for PF.
The third study aims to estimate wheelchair propulsion forces. Propulsion forces were estimated from the accelerometer placed on the upper arm using a bagging regression technique. The estimated forces were compared with the criterion. Mean absolute errors (MAE), mean absolute percentage of error (MAPE), were calculated. The results showed an overall MAPE of 17.9%. Intraclass Correlation Coefficients and Bland-Altman plots were used to assess the agreement between the criterion and the estimated force.
Share
| Citation/Export: |
|
| Social Networking: |
|
Details
| Item Type: |
University of Pittsburgh ETD
|
| Status: |
Unpublished |
| Creators/Authors: |
| Creators | Email | Pitt Username | ORCID  |
|---|
| OJEDA AGUILAR, ALEJANDRA MANOELA | alo34@pitt.edu | ALO34 | |
|
| ETD Committee: |
|
| Date: |
24 May 2013 |
| Date Type: |
Publication |
| Defense Date: |
28 February 2013 |
| Approval Date: |
24 May 2013 |
| Submission Date: |
10 April 2013 |
| Access Restriction: |
No restriction; Release the ETD for access worldwide immediately. |
| Number of Pages: |
70 |
| Institution: |
University of Pittsburgh |
| Schools and Programs: |
School of Health and Rehabilitation Sciences > Rehabilitation Science and Technology |
| Degree: |
MS - Master of Science |
| Thesis Type: |
Master's Thesis |
| Refereed: |
Yes |
| Uncontrolled Keywords: |
Wheelchair Propulsion, Monitoring Devices, Machine Learning |
| Date Deposited: |
24 May 2013 14:27 |
| Last Modified: |
15 Nov 2016 14:11 |
| URI: |
http://d-scholarship.pitt.edu/id/eprint/18311 |
Metrics
Monthly Views for the past 3 years
Plum Analytics
Actions (login required)
 |
View Item |
![[feed]](/images/feed-icon-32x32.png){kind=icon}