Abstract
Functional magnetic resonance imaging (fMRI) activity in the visual cortex is often thought to be a correlate of processes that give rise to perceptual awareness. Indeed, there is a great deal of support for the idea that the fMRI BOLD signal increases with the increasing visibility of a stimulus. Yet, processes that give rise to perceptual awareness do not always produce an increase in stimulus visibility. Here we tested whether the fMRI signal in the visual cortex is correlated with stimulus visibility. To test this, we presented a visual illusion called metacontrast masking, which occurs when a briefly presented stimulus in one location is followed by another brief stimulus at an adjacent location. Due to the second stimulus, the visibility of the first stimulus is greatly reduced and can even appear invisible. In an experiment, we presented a metacontrast masked stimulus, and measured the retinotopic fMRI activation corresponding to the masked object. Because the stimulus was masked, its visibility was dramatically reduced. Despite the fact that the masked stimulus was nearly invisible, the fMRI BOLD signal actually increased in the retinotopic region corresponding to the masked stimulus. The results show that the fMRI BOLD signal does not necessarily reflect stimulus visibility or perceptual magnitude. They also show that fMRI activity in the visual cortex does not simply reflect the physical or retinal image. Rather, the BOLD signal must also reflect processes that reduce perceptual awareness of a stimulus.
Supported by grants from NIH to D.W. and CIHR to M.A.G.