One-dimensional stochastic models for spontaneous and burst blink

The present study proposed one-dimensional stochastic diffusion models for spontaneous eyeblink and analyzed and discussed the blink burst during low vigilance. The models presumed the interblink interval distribution to be first-passage-time probability densities of the Ornstein-Uhlenbeck process. To estimate the parameters, the behaviour of upper eyelids was optically measured when the vigilance of the subjects was high and low, and transformed into point series data. The observed data was in good agreement with the model output and the threshold of a virtual blink generator changed with vigilance. It became evident that the Ornstein-Uhlenbeck first-passage-time models are valid spontaneous blink models and that the burst patterns occur because of a shift of the threshold.