Describing gait as a sequence of states.

Traditionally, gait analysis has been based on normalizing the stride time to a percentage and then averaging several strides measured under the same conditions. This procedure relies on the questionable assumptions that gait is a cyclic movement with superimposed noise and that there is no variability in the timing of activation or in the angles within the stride so no rescaling occurs during the percentage conversion. However, there is a fluctuation in the timings at which the peak values occur. A typical hallmark of this time-rescaling is the increase of the joint angle standard deviation when the angular velocity increases. The goal of this paper is to present a description of gait to avoid averaging without distorting the original curves. In addition, it allows the analysis of the fluctuation between consecutive strides. In this method, it is assumed that gait is quasi-periodic. The key point is the representation of gait by a state vector that evolves in time. This state vector can be used to calculate the instantaneous period and provides a measure of the time fluctuations between strides. The sequence of states method describes a quasi-periodic movement like gait with a continuous estimate of cycle time and provides measure of the deviations between cycles.