Stochastic phase resetting of two coupled phase oscillators stimulated at different times

A model of two coupled phase oscillators is presented, where the oscillators are subject to random forces and are stimulated at different times. Transient phase dynamics, synchronization, and desynchronization, which are stimulus locked (i.e., tightly time locked to a repetitively administered stimulus), are investigated. Complex coordinated responses, in terms of a noise-induced switching across trials between qualitatively different responses, may occur when the two oscillators are reset close to an unstable fixed point of their relative phases. This can be achieved with an appropriately chosen delay between the two stimuli. The switching of the responses shows up as a coordinated cross-trial (CT) response clustering of the oscillators, where the two oscillators produce two different pairs of responses. By varying noise amplitude and coupling strength we observe a stochastic resonance and a coupling-mediated resonance of the CT response clustering, respectively. The presented data analysis method makes it possible to detect such processes in numerical and experimental signals. Its time resolution is enormous, since it is only restricted by the time resolution of the preprocessing necessary for extracting the phases from experimental data. In contrast, standard data analysis tools applied across trials relative to stimulus onset, such as CT averaging (where an ensemble of poststimulus responses is simply averaged), CT standard deviation, and CT cross correlation, fail in detecting complex coordinated responses and lead to severe misinterpretations and artifacts. The consequences for the analysis of evoked responses in medicine and neuroscience are significant and are discussed in detail.