Temporal properties of centre-surround interactions in human motion perception

Abstract

A recent electrophysiological study revealed that the response to the antagonistic surround of macaque area MT receptive fields is delayed (15 - 20 ms) relative to the centre (Perge et al, 2002 Investigative Ophthalmology & Visual Science 43 E-Abstract 3931). We investigated whether this phenomenon, ie the delay between centre and surround, can be demonstrated in a human psychophysical experiment. The stimulus was a foveally presented circular patch of a moving random-dot pattern (centre) surrounded by a moving random-dot pattern annulus (surround). Stimuli in both centre and surround consisted of two-frame, single-step motion, repeated 6 times at 100 ms time intervals. The motion steps for the surround were presented earlier, simultaneously, or later than the centre (from -60 ms in 10 ms steps to +60 ms relative to the centre). The surround effect was quantified by measuring the strength of motion induction. Observers performed a direction-discrimination task for motion in the centre, with the surround moving in the same direction, or in the opposite direction. Performance was measured as a function of coherence level in the centre, always with maximum coherence in the surround. In accordance with earlier findings, motion induction was stronger at higher stimulus speed. Using this paradigm, we found the strongest surround effect at simultaneous centre - surround presentation. These results are in contrast to the electrophysiological results in area MT. This might indicate that short temporal delays between individual centre - surround receptive fields in the visual system are cancelled out at the level of perception.