occurrences of the gait events during a gait cycle were selected
as gait features. In fact, human gait is characterized by complex
dynamics containing both non-stationarity and non-linearity,
and all the characteristic information about the change of gait
function resides in the intrinsic interaction between gait
variables. It is required that the gait features obtained by the
gait feature extraction technique must maintain both time and
frequency content of gait data. Unfortunately, the gait features
obtained by the traditional methods can not offer important
characteristic information associated with detection of the
change of gait function. This motivates the application of the
advanced analysis technique in gait data for the capture of the
time-frequency information about human gait function change
[3,4