CONSTRAINTS ON A NONLINEAR MODEL FOR SENSORIMOTOR SYNCHRONIZATION

M. L. BAVASSI; R. LAJE

BUZIOS

Congreso; I Congreso IBRO/LARC de Neurociencias. NEUROLATAM.; 2008

Animals developed multiple timing mechanisms on scales that range more than 10 orders of magnitude, from microseconds to days. We focus on timing in the range of tens to hundreds of milliseconds, where many forms of sensory and motorprocessing occur –most notably speech and music recognition and production, and fine motor coordination. Sensorimotor synchronization, a robust and spontaneous behavior specific to humans, occurs within this range. Finger tappingexperiments are usually performed to probe the mechanism in charge of keeping average synchrony. In these experiments, the subject is instructed to keep synchrony by tapping along with a periodic stimulus. A perturbation in the stimulus (a stepchange in period, for instance) may appear. The time difference between each stimulus and its corresponding response is recorded. In order to either keep average synchrony with a periodic stimulus or return to average synchrony after aperturbation, an error correction mechanism based on past performance should actively drive the subject’s response.We first review the literature on finger tapping time series (both isochronous and perturbations) in search of dynamical constraints on a model for the error correction mechanism. We found several qualitative features like correction overshootingthat, although usually disregarded, impose strong constraints on both qualitative and quantitative aspects of a model. We interpret and arrange these published experimental findings into a dynamical framework, and develop an informedmathematical model for the error correction mechanism. We put forward several qualitative predictions, like the amount of overshoot in response to a positive step change should increase smoothly with the step size.