Cortical areas related to depth perception of natural images: a LORETA analysis.

Fischmeister, F. P. S., Pripfl, J., Tschida, U., Leodolter, U., and Bauer, H.
Biological Psychology Unit, Department of Psychology, University of Vienna, Austria


Normal binocular vision implies fusion of the two retinal inputs and processing of monocular and binocular cues to construct a vivid percept of depth. This seemly effortless task involves a series of neural processes that are yet not well known or localized. The aim of the present DC-EEG study was to determine which cortical areas get activated while perceiving natural images that switch randomly between scrambled, monoscopic and stereoscopic.


DC-EEG was recorded from thirty male subjects. Eye-movement related changes were eliminated using a regression approach. Artefact free single-trials were averaged per condition and further supplied to LORETA (Pascual-Marqui et al. 1999). For the statistical analysis we used the randomisation-permutation approach illustrated by Nichols and Holmes (2001).


Significant higher current density values for all stereoscopic conditions were found in the dorsal visual pathway of both hemispheres. This could be explained through the additional information provided by the horizontal disparity information induced in the stereo condition (Taira et al. 2001). Other activated regions involve the left medial frontal cortex and the anterior cingulate of both hemispheres. In these regions an activation increase can be seen for the stereo condition when switching from scrambled images suggesting a strong involvement of these areas in stereopsis (Gulyas et al. 1994). Neither did we find significant differences in the temporalcortex nor a strong right hemisphere dominance which were described by many studies. These findings suggest that the different aspects of stereoscopic natural images are processed by a number of distinct cortical areas.