Simulating bistable perception with interrupted ambiguous stimulus using self-oscillator dynamics with percept choice bifurcation.

A behavioral stochastic self-oscillator model is used for simulating interrupted ambiguous stimulus-induced percept reversals. The results provide further support for a dynamical systems foundation of cognitive and psychological problems as discussed in detail within the context of Gestalt psychology by Wagemans et al. (Concept Theor Found Psychol Bull 138(6):1218-1252, 2012), and for coordination dynamics of the brain (Kelso in Philos Trans R Soc B 367:906-918, 2012). Statistical evaluation of simulated reversal time series predicts a maximum of the percept reversal rate that conforms with a number of results in the literature. The macroscopic model is based on two inhibitorily coupled sets of three coupled nonlinear equations, one triplet for each percept. The derivation of our specific dynamics equations is based on a drastically simplified field theoretical approach using well-known phase synchronization for explaining brain dynamics on the macroscopic EEG level. The degree of coherence (contrast μ, 0 ≤ μ ≤ 1) of the superimposed fields required for onset of bistable dynamics is related to a phase synchronization index of EEG fields, and it is used in the present context as ambiguity control parameter. For quantitative agreement with the experimental data, the addition of a stochastic Langevin force term in the attention equation proved essential. Formal analysis leads to a quantification of well-known "cognitive inertia" and supports the interplay between percept choice (bifurcation) dynamics during stimulus onset and adaptive gain (attention fatigue) driven quasiperiodic percept reversals.