Clinical gait analysis by neural networks: issues and experiences

Abstract

Clinical gait analysis is an area aiming at the provision of support for diagnoses and therapy considerations, the development of bio-feedback systems to train patients, and the recognition of effects of multiple diseases and still active compensation. The data recorded with ground reaction force m e~~u r e m e n t platforms is a converrient starting point for gait analysis. We argue in favor of using the raw data from such force platforms and apply artificial neural networks for gait malfunction identification. In this paper we discuss OUT latest results in this line of research by using a supervised learning rule. The employed classification approach is learning vector quantization which proved to be highly robust in the training process yielding a remarkably high recognition accuracy of gait patterns.

Methods for person identification on a pressure-sensitive floor: Experiments with multiple classifiers and reject option

“…x ball End point of ball strike (when amplitude is below x min ) 19 y ball Amplitude value in end point of ball strike (ampl. in x ball ) 20 length heel Length of the heel impact (x heel , x min ) 21 length ball Length of the ball of the foot impact (x min , x ball ) 22 shape heel ((y max1 À y min )/(x min À x heel )) 23 shape ball ((y max2 À y min )/(x ball À x mid )) (e.g., in Euclidean sense) from the training set, and assigns unknown examples to the class of majority vote of labels.…”

“…A lot of work has also been done in medical research domains, including [19], where pattern recognition methods were used to classify different gaitrelated injuries based on GRF sensor measurements.…”

Gait classification in children with cerebral palsy by Bayesian approach

“…The process, however, is often subjective because it mainly relies on the interpretation of complicated scatter plots. Recently, several supervised learning methods have been exploited in automatic analysis of gait patterns, and gait pathology related to CP in particular [1,3,[6][7][8]. The general goal of these research is the application of intelligent classification systems which may not only enable clinicians to differentiate gait patterns into clinically significant categories in the process of their decision-making, but would also facilitate standardization of gait management and communication across professional boundaries.…”

A Survey on Visual Surveillance of Object Motion and Behaviors

“…Unfavorable factors, such as illumination variance, shadows and shaking branches, bring many difficulties to the acquirement and updating of background images. There are many algorithms for resolving these problems including temporal average of an image sequence [15], [82], adaptive Gaussian estimation [70], and parameter estimation based on pixel processes [79], [80], etc.…”