Building blocks for time-to-contact estimation by the brain

The ability to estimate impending collisions is a vital requirement for any organism. Behavioral studies in several animal species have demonstrated sensitivity to time-to-contact. The proven capability to estimate time-to-contact in humans and animals must have its origin in brain processes that extract or compute correlates of time-to-contact from the visual input. Unfortunately, not many studies have looked specifically at neuronal sensitivity for time-tocontact. Specific investigations of this question in the primate brain (including humans) are still lacking. However, a great deal is known about the neuronal selectivity for the building blocks of time-to-contact mechanisms: visual motion, looming, binocular disparity, motion in depth, etc. This chapter presents a discussion of the properties of neurons that encode these cues and hopefully will be helpful for homing in on the biological foundations of time-to-contact. Since we are able to precisely estimate time-to-contact, our brains must contain mechanisms that conduct the required sensory analysis. Candidates for neural time-to-contact mechanisms have been described in several animal species (Pigeon: (Frost and Sun, 2002), Locust: (Rind and Simmons, 1997; Gabbiani et al., 1999), Gerbil: (Shankar and Ellard, 2000)). The topic has not yet been specifically addressed, however, for the brains of higher mammals and especially primates. But a number of the sensory cues that are used to derive time-to-contact measurements