Cortical Activation for 3-D Shape Constructed from Different Depth Cues

Human observers are able to extract information from multiple depth cues in a visual scene. In this study, we investigated cortical responses to 3D shapes constructed from different depth cues with functional magnetic resonance imaging (fMRI). We used three different depth cues, binocular disparity, kinetic depth cue and perspective, and their combinations to study how brain processes depth information. We compared blood oxygenation level dependent (BOLD) activation to images of different depth cues with activation to images of no depth cue in health human observers. Our results showed that the middle occipital gyrus (MOG), intra-occipital sulcus (IOS), and right intra-parietal sulcus (IPS) were activated by 3D shapes defined by all depth cues tested. The ventral occipital cortex was only activated by monocular depth cues. Further localizer experiments showed that the MOG and IOS activation included part of V3B, the lateral occipital complex (LOC) and the kinetic occipital area (KO). The areas activated by monocular depth cues in the ventral occipital cortex included part of V2v, V3v and V4. When combined disparity and kinesthetic depth cues, the IPS, KO, LO and ventral occipital activation increased. When combined disparity and perspective cues, however, only KO and left IPS activation was augmented. The ventral occipital may involve in the processing of extracting depth information from local elements in the monocular cues rather than the representation of the depth perception. Our results imply that stereopsis is mainly a function of the dorsal stream while monocular depth cues require both ventral and dorsal streams of vision to process.