Extracting global and local dynamics from the stochastics of rhythmic forearm movements

The authors examined the dynamics governing rhythmic forearm movements that 9 participants performed under a variety of task constraints by using a generic, unbiased analysis technique for extracting the drift coefficients of Fokker-Planck equations from stochastic data. From those coefficients, they reconstructed and analyzed vector fields and phase portraits to identify characteristic, task-dependent kinematic and dynamical features. They first directly estimated the parameters of weakly nonlinear self-sustaining oscillators from the extracted drift coefficients. The estimated parameters that the authors had selected instinctively and then particularized by using averaging methods largely confirmed previously derived limit-cycle models. Next, they ventured beyond limit-cycle models to examine global and local dynamical features that those models cannot adequately address, particularly task-dependent changes in flow strength and curvature and distinct dynamical features associated with flexion and extension. The authors argue that those features should be focal points of researchers' future modeling efforts to formulate a more adequate and encompassing account of the dynamics of rhythmic movement.