Abstract
Objective: The present study was designed to determine which cortical areas are involved in the analysis of image shading for the determination of 3D shape.
Methods: We investigated cortical activation using functional Magnetic Resonance Imaging (fMRI) with a 1.5 T MR scanner. The whole brains of eight subjects were scanned. Subjects were required to maintain a fixation on the central fixation point throughout the entire experiment. Four experimental conditions (3D Shading, 2D Shading, 2D Uniform gray and Fixation) were presented in a block design.
The 3D stimuli (3D Shading condition) depicted roughly spherical objects with random patterns of ridges and valleys that were shaded with a reflectance model containing both diffuse and specular components. A set of matched control stimuli was created, which appeared perceptually as 2D luminance patterns (2D Shading condition), whose image luminance histogram and amplitude spectra were matched to those of the 3D surfaces. An additional 2D control was also included in which the outlines of the 3D objects were filled with a uniform gray luminance (2D Uniform gray condition). Each stimulus subtended approximately 9 degrees and was presented with a duration of 1.4 sec.
Results and Conclusions: The subtraction of 3D-2D shading patterns revealed significant activation in lateral occipital cortex (LOC), early visual cortex (areas V1/V2), and in intraparietal sulcus (IPS). These areas of activation are similar to those revealed in previous experiments to be involved in the processing of 3D shape from motion [Orban at al., 1999; Peuskens at al., 2002], texture or binocular disparity. These results suggest that there is considerable convergence in the neural processing of different depth cues. In these areas of activation 2D shading stimuli provide stronger activation than 2D uniform gray stimuli, therefore the cortical regions were less activated in the subtraction 3D Shading-2D Shading than 3D Shading-2D Uniform gray.