Inter-trial Gait Variability Reduction Using Continous Curve Registration

Timing in peak gait values shifts slightly between gait trials. When gait data are averaged, some of the standard deviation can be associated to this inter-trial variability unless normalization is carried out beforehand. The objective of this study was to determine how continuous curve registration, an alignment technique, can reduce inter-subject variability in gait data without altering the original curve characteristics. Gait data was obtained by means of a four-camera high-speed video system synchronized to a force plate. Fifty-nine gait trials data were collected from twenty young and healthy subjects. Curve registration was applied to hip angular displacement, net moment and power curves generated in the sagittal plane. Following registration, the peak values increased by an average of 1.2% (0.11±0.26°) for angular displacement, and for powers by 11.2% (0.11±0.09 W/kg). First and second derivatives of the unregistered and registered curves did not displayed important differences and harmonic content of the signals was practically unaffected. Continuous curve registration would thus be an appropriate technique for application prior to any statistical analysis using able-bodied gait patterns.