Assessment of Post Stroke Functioning Using Machine Vision
Sonya Allin, Deva Ramanan
icon We present a system to automatical ly assess the functional performance of stroke survivors along axes defined by the Arm Motor Ability Test (AMAT). The upper body motion of seven stroke survivors was measured in a laboratory environment using a commercial motion capture device and a novel kinematic tracker of our design. Statistics generated by each individual were related to expert-determined assessments of functional health. Results indicate several kinematic targets that correlate with and predict opinions of health. These include recorded motion of the torso during the performance of tasks and flexion about the elbow on the hemiparetic (impaired) side. We show that both kinematic statistics can be cheaply and robustly measured with video cameras while stil l preserving their diagnostic value. Such cheap and robust measurements will ultimately facilitate assessment outside of clinics and in places where functioning is valuable to individuals, such as homes and workplaces.

Download: pdf

Text Reference

Sonya Allin and Deva Ramanan. Assessment of post stroke functioning using machine vision. In IAPR Conference on Machine Vision Applications (MVA). 2007.

BibTeX Reference

@inproceedings{AllinR_MVA_2007,
    AUTHOR = "Allin, Sonya and Ramanan, Deva",
    TAG = "people",
    TITLE = "Assessment of Post Stroke Functioning Using Machine Vision",
    BOOKTITLE = "IAPR Conference on Machine Vision Applications (MVA)",
    YEAR = "2007"
}