Constanze Hofstoetter ¹ , Farshad Moradi ² , Christof Koch ²

Abstract

We used fMRI to investigate orientation selective adaptation (OSA) in visual cortex and its correlation with visual awareness. Psychophysical studies have demonstrated that OSA does not require visual awareness, and hence it presumably originates in early visual areas. Seven participants viewed high-contrast grating patterns in a 3T scanner (EPI, TR=2s, TE=30ms). The visual display was divided into 7 regions (co-centric rings covering the upper visual field). Corresponding retinotopic maps for each ring were determined using separate localizers (2 runs) for each participant in the occipital lobe. Each ring was composed of a grating (of either 45 or 135 deg stripes, random width 0.1-2 deg) which flickered at 2.5 Hz and flipped its orientation every 30-50s (10 runs*6 flips/ring). Participants were asked to monitor the rings and report those that flipped. On average, they missed the change in orientation in more than half of the trials due to crowding and flickering that masked changes in orientation, in particular for the peripheral rings. When correctly reported, the change in orientation was accompanied by a small but highly significant increase in BOLD signal (0.15+-0.02%, p<1.2e-8). This increase may reflect the transient associated with the change in orientation, or it may signal that a subpopulation of neurons that are not adapted is being stimulated. However, when the observers failed to notice the change in orientation, there was no increase in activity (0.007+-0.02%). In the few trials that the location of the change was misperceived, there was a significant increase in the activity at the reported location. These findings indicate that the increase in BOLD activity in early visual cortex can reflect the consciously perceived change, a signal presumably originating in higher stages. These results imply that attention and awareness are important confounding factors in fMRI adaptation studies.