Year: 2013
Type: article
Abstract: People with disabilities who rely on manual wheelchairs as their primary means of mobility face daily challenges such as mobility limitations and environmental barriers when engaging in regular physical activity. Therefore, our research addressed the need for a valid and reliable physical activity monitor to assess and quantify physical activities among manual wheelchair users (MWUs) in free-living environments. Providing an accurate estimate of physical activity (PA) levels in MWUs can assist researchers and clinicians to quantify day-to-day PA levels, leading to recommendations for a healthier lifestyle. In the first stage we developed and evaluated new classification and EE estimation models for MWUs with spinal cord injury (N=45) using SenseWear, an off-the-shelf activity monitor, designed for the general population without disabilities. The results suggested that SenseWear can be used by researchers and clinicians to detect and estimate the EE for four activities tested in our study. The second phase of our research project developed an activity monitor especially designed for MWUs. Previous research in community participation of MWUs and the studies discussed above found that wheelchair mobility characteristics are necessary to study PA patterns in MWUs. This requirement led us to develop and evaluate a Physical Activity Monitor System (PAMS) composed of two components: a gyroscope based wheel rotation monitor (G-WRM for tracking wheelchair mobility and an accelerometer that quantifies upper arm movement. We tested PAMS in 45 MWUs with SCI in the structured (laboratory) and semi-structured environments (National Veterans Wheelchair Gamers 2012). In addition, we also tested a subsection of this population (N=20) a second time, in their home environments. The PAs were classified as resting, armergometry, other sedentary activities, activities involving some wheelchair movement, propulsion, basketball and caretaker pushing. The EE estimation results (error: -9.8%) and the classification results (accuracy: 89.3%) indicate that PAMS can reliably track wheelchair-based activities in laboratory and home environments. Furthermore, we used participatory action design to evaluate the usability of PAMS in six MWUs with SCI. The usability study indicated that users were very satisfied with PAMS and the information provided by the smartphone to the users about their PA levels.
Author: Shivayogi V. Hiremath
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W1030124