PATTERN RECOGNITION USING ANGLE-ANGLE DIAGRAMS FOR GAIT ASSESSMENT.

Jane Macfarlane, Max Donath

Research output: Contribution to conferencePaperpeer-review

Abstract

An Euclidean distance measure is proposed for measuring deviation from normal on the gait angle-angle diagram. The measures will have significance in tracking rehabilitation progress and in evaluating the efficacy of procedures impacting gait disabilities. The features for angle-angle diagram analysis are based on a 'deviation from normal' measure. Given a sampling of normal curves, a mean curve is determined. A boundary around the mean curve is defined based on the maximum deviation any of the normal curves make from the mean. If the patient's gait under consideration deviates outside the boundary area at any point in time, it is considered an abnormal angle-angle point, and the Euclidean distance between the point and the boundary is computed as a function of time. Furthermore, as one traverses the loop, each deviation is added to a running sum to give a final total deviation from normal measure.

Original languageEnglish (US)
Pages633-634
Number of pages2
StatePublished - Dec 1 1984

Fingerprint Dive into the research topics of 'PATTERN RECOGNITION USING ANGLE-ANGLE DIAGRAMS FOR GAIT ASSESSMENT.'. Together they form a unique fingerprint.

Cite this