Abstract
PURPOSE: We tested the hypothesis that more neural resources in visual cortex are used in visual processing when subjects are involved in a task, compared to their retinally-driven responses. Using fMRI, we investigated whether there is any difference in activation in the visual areas between passive viewing and a discrimination task.
METHODS: The entire visual field was divided into five bars (horizontal or vertical), and two bars equally distant from the fixation point were presented in each trial. The bars consisted of small chromatic/achromatic squares, (black/white, L−M, and (L+M)-S, following the DKL space), that flickered at 5 Hz and lasted for 2.5 seconds. The bar locations were randomly selected from predetermined five locations with respect to the fixation. Subjects' task was either to fixate (‘blank’ and ‘no-go’ blocks), or to fixate and press a button (‘go’ blocks) to indicate which contrast of the flickering squares is greater between the two bars. There were five trials in each block and the three blocks were randomly presented five times in each session.
RESULTS: In the fixation task, as the bars moved from the fixation to the periphery, activations spread from foveal representation to the periphery in the visual cortex in a manner consistent with known retinotopy. For both vertical and horizontal bars, most of the visual areas were activated. In the contrast discrimination task, similar (sometimes enhanced) retinotopically-appropriate responses were observed. However, small but significant additional activations were observed in the areas surrounding the retinotopically-appropriate areas in V1, V2, V3A, and V4/V8. Strong additional activations were also shown in V4 for chromatic (L−M and S) bars than for luminance bars.
CONCLUSIONS: Discrimination task not only enhances retinotopically- appropriate response, consistent with enhanced response in single-cell research, but also involves additional neural resources in V1, V2, V3A, and V4.