I will present some works related to computational neuroscience where we defined the notion of asynchronous evaluation in the context of visual attention modeling. By considering those models as being a generalization of cellular automata, we'll investigate to what extent this notion of asynchronicity can be translated into the cellular automata theory.