The extended transentropy function as a useful quantifier of human motion variability

A new concept is presented for quantifying the variability of iterated human motions and other recurrent biological phenomena. The concept is based on the idea of interpreting the variations occurring in repeatedly executed stereotyped motions as particular realizations of a regular stochastic process with superimposed but infrequently occurring chaotic excursions. Under these premises, the extended transentropy of such a process turns out to be a dimensionless and appropriate quantifier of motion variability. Extended transentropy functions can be used to monitor variability changes during motion execution. The practical applicability and usefulness of the new method is demonstrated by means of an example from gymnastics (the kip on the horizontal bar). Major areas of future applications are expected to be gait analysis, all sports movements, and the objective evaluation of the success rates of certain medical treatment procedures for specific movement disorders.