Imagery for shapes activates position-invariant representations in human visual cortex |
| |
Authors: | Stokes Mark Saraiva Ana Rohenkohl Gustavo Nobre Anna C |
| |
Affiliation: | Oxford University, Department of Experimental Psychology, Oxford, UK. mark.stokes@sjc.ox.ac.uk |
| |
Abstract: | In this study, we show that top-down control mechanisms engaged during visual imagery of simple shapes (letters X and O) can selectively activate position-invariant perceptual codes in visual areas specialised for shape processing, including lateral occipital complex (LOC). First, we used multivoxel pattern analysis (MVPA) to identify visual cortical areas that code for shape within a position-invariant reference frame. Next, we examined the similarity between these high-level visual codes and patterns elicited while participants imagined the corresponding stimulus at central fixation. Our results demonstrate that imagery engages object-centred codes in higher-level visual areas. More generally, our results also demonstrate that top-down control mechanisms are able to generate highly specific patterns of visual activity in the absence of corresponding sensory input. We argue that a general model of top-down control must account for dynamic modulation of functional connectivity between high-level control centres and overlapping neural codes in visual cortex. |
| |
Keywords: | |
本文献已被 PubMed 等数据库收录! |
|