Related Experiment Videos
Detecting destabilizing wheelchair conditions for maintaining seated posture.
Anna Crawford1,2, Kiley Armstrong1,2, Kenneth Loparo3
1a Motion Study Laboratory, Louis Stokes Cleveland, Affairs Medical Center , Cleveland , OH , USA.
Disability and Rehabilitation. Assistive Technology
|April 4, 2017
Summary
This study developed an algorithm using trunk accelerations to detect destabilizing events for manual wheelchair users with spinal cord injuries (SCI), aiming to prevent falls. Classification accuracy reached 95.8% for trunk sensors, showing feasibility for protective interventions.
Area of Science:
- Biomechanics
- Rehabilitation Engineering
- Assistive Technology
Background:
- Spinal cord injuries (SCI) significantly impact quality of life and independence.
- Manual wheelchair users face risks of falls due to destabilizing conditions.
- Developing systems to enhance wheelchair user stability is crucial for safety and participation.
Purpose of the Study:
- To detect and classify potentially destabilizing conditions for manual wheelchair users with SCI.
- To dynamically increase stability and prevent falls.
- To assess the effectiveness of sensor placement for intervention triggers.
Main Methods:
- A volunteer with T11 paraplegia propelled a manual wheelchair over various terrains.
- Wireless inertial measurement units (IMUs) were attached to the wheelchair and user's sternum.
- Algorithms based on Mahalanobis distance classified destabilizing events using recorded acceleration and angular velocity data.
Main Results:
- The decision algorithm achieved high classification accuracies: 95.8% for chest sensors, 58.3% for wheelchair sensors, and 91.7% for both.
- Mahalanobis distance classification using trunk accelerations proved feasible for detecting destabilizing events.
- Integrating wheelchair-mounted sensor data reduced the false negative rate.
Conclusions:
- Trunk acceleration-based classification is a viable method for triggering protective interventions in wheelchair users.
- Combined sensor data enhances detection reliability, decreasing false negatives.
- This technology has implications for improving safety, independence, and participation in daily activities for individuals with SCI.