3rd International ICST Conference on Body Area Networks

Research Article

Classifying Wheelchair Propulsion Patterns with a Wrist Mounted Accelerometer

Download799 downloads
  • @INPROCEEDINGS{10.4108/ICST.BODYNETS2008.2963,
        author={Brian French and Asim Smailagic and Dan Siewiorek and Vishnu Ambur and Divya Tyamagundlu},
        title={Classifying Wheelchair Propulsion Patterns with a Wrist Mounted Accelerometer},
        proceedings={3rd International ICST Conference on Body Area Networks},
        publisher={ICST},
        proceedings_a={BODYNETS},
        year={2010},
        month={5},
        keywords={eWatch wearable sensors machine learning manual wheelchair propulsion patterns},
        doi={10.4108/ICST.BODYNETS2008.2963}
    }
    
  • Brian French
    Asim Smailagic
    Dan Siewiorek
    Vishnu Ambur
    Divya Tyamagundlu
    Year: 2010
    Classifying Wheelchair Propulsion Patterns with a Wrist Mounted Accelerometer
    BODYNETS
    ICST
    DOI: 10.4108/ICST.BODYNETS2008.2963
Brian French1,*, Asim Smailagic2,*, Dan Siewiorek3,*, Vishnu Ambur4,*, Divya Tyamagundlu5,*
  • 1: ECE Department, Carnegie Mellon University, Pittsburgh PA 15213 +1 412 268 3372
  • 2: ICES, Carnegie Mellon University, Pittsburgh PA 15213 +1 412 268 7863
  • 3: HCI Institute, Carnegie Mellon University, Pittsburgh PA 15213 +1 412 268 2570
  • 4: Department of Biomedical Eng., Georgia Institute of Technology, Atlanta, Georgia 30322 +1 404 727 9874
  • 5: ECE Department, Carnegie Mellon University, Pittsburgh PA 15213 +1 412 596 7405
*Contact email: bfrench@ece.cmu.edu, asim@cs.cmu.edu, dps@cs.cmu.edu, gtg785q@gmail.com, divyatyam@cmu.edu

Abstract

In this paper, we describe a manual wheelchair propulsion classification system which recognizes different patterns using a wrist mounted accelerometer. Four distinct propulsion patterns have been identified in a limited user study. This study is the first attempt at classifying wheelchair propulsion patterns using low-fidelity, body-worn sensors. Data was collected using all four propulsion patterns on a variety of surface types. The results of two machine learning algorithms are compared. Accuracies of over 90% were achievable even with a simple classifier such as k-Nearest Neighbor (kNN). Being able to identify current propulsion patterns and provide real-time feedback to novice and expert wheelchair users is potentially useful in preventing future repetitive use injuries.