Article

The aim of the study was to train and test support vector machines (SVM) and
self-organizing maps (SOM) to correctly classify gait patterns before, during and
after complete leg exhaustion by isokinetic leg exercises. Ground reaction forces
were derived for 18 gait cycles on 9 adult participants. Immediately before the
trials 7-12, participants were required to completely exhaust their calves with
the aid of additional weights (44.4+/-8.8kg). Data were analyzed using: (a) the
time courses directly and (b) only the deviations from each individual's
calculated average gait pattern. On an inter-individual level the person
recognition of the gait patterns was 100% realizable. Fatigue recognition was
also highly probable at 98.1%. Additionally, applied SOMs allowed an alternative
visualization of the development of fatigue in the gait patterns over the
progressive fatiguing exercise regimen.
CI - Copyright (c) 2010 Elsevier B.V. All rights reserved.

Langue : ANGLAIS