Abstract
Recently, TMS has been used to investigate the timing of the contribution of visual areas to motion perception, but with widely conflicting results. For example TMS of human visual area V5/MT substantially disrupts global motion judgements. Taken together, previous reports suggest a very broad temporal window (∼ ±200 ms relative to stimulus onset) over which TMS impairs performance on global motion tasks. To resolve this issue we employed standard psychophysical techniques, in conjunction with TMS, to investigate the disruption profile for random-dot-kinematograms depicting either translational, radial or rotational global motion. We measured the influence of TMS over a broad temporal range (±200 ms), sampled at 13 ms intervals, to reveal the fine temporal structure of the disruption profile. Prior to applying TMS we established baseline thresholds for the three types of global motion: specifically the stimulus coherence level was manipulated by constraining a fixed proportion of the dots (“signal” dots) to move coherently along either a translational, radial or rotational (circular) trajectory and the remainder (“noise” dots) to move in random directions. Four observers judged the global direction (either up vs. down, expansion vs. contraction or clockwise vs. anticlockwise, respectively) and the coherence level producing 75% correct performance was taken as the threshold. Adopting each observer's individual coherence level at threshold we then examined how single-pulse TMS (2T) delivered to V5/MT modulated the percentage of correct judgments. Results showed that TMS produced a peak performance deficit of ∼15%. More importantly the disruption profile revealed two distinct epochs (separated by 60 to 80 ms) during which global direction judgments were reliably disrupted, although there were individual differences regarding the onset of the time windows of disruption. The bimodal nature of the disruption profiles, is consistent with feed-forward and feed-back pathways between visual areas mediating global motion processing.
PM and TL are supported by the Wellcome Trust and BBSRC, respectively.