Retinotopic Mapping with fMRI

The primate visual system contains a multitude of areas that are topographically organized. In other words, the spatial structure of an image is preserved in each area. Several decades of research have mapped out the visual areas of the monkey brain with neurophysiological techniques (for a review, see Van Essen 2004). Recently, visual areas in the human brain have been delineated using functional MRI techniques ( Kastner et al. 1998). With monkey fMRI we can delineate the monkey visual areas in the same manner as performed with the human. This allows us to both compare and confirm previous results from neurophysiology, as well as perform across-species comparisons between the human and the monkey visual systems.

Investigations of Object Category Representations with fMRI

It has been suggested with fMRI that the human ventral temporal cortex is organized in a category-specific manner with highly selective regions of cortex embedded within a distributed system representing object forms (for a review, see Kanwisher 2004. For example, images of spatial environments such as houses and indoor scenes tend to activate more medial regions in the fusiform, parahippocampal, and lingual gyri; while images of faces tend to activate more lateral regions including the fusiform gyrus, inferior occipital gyrus, and the posterior superior temporal sulcus.

On the other hand, it is not clear from single-unit recording studies whether the monkey ventral pathway is organized in a similar manner. We know from several groups, including the first reports from Charlie Gross here at Princeton, that there are cells highly or exclusively responsive to images of faces that tend to be clustered along the fundus and banks of the superior temporal sulcus and inferior temporal gyrus. Our fMRI studies using awake, fixating monkeys have revealed face-selective regions in the monkey temporal cortex as shown here. These regions of activity correspond nicely with known locations of face cells. Current studies are aimed at whether any of these activated regions in the monkey correspond to face areas found in the human and whether other categories of visual stimuli can activate the temporal cortex similarly as in the human as well.

Diffusion Tensor Imaging of Monkeys

Most information regarding the connectivity of the monkey cortex has been obtained using invasive tracing studies. In comparison, the white matter fiber tracts of the human cortex has recently begun to be mapped using a non-invasive technique known as diffusion tensor imaging (DTI). The degree of correspondence between DTI and the tracing studies is unknown. We are currently collaborating with several laboratories (FMRIB at Oxford University, CSBMB at Princeton University) to assess this correspondence using both human and monkey DTI. Future studies will use this technique for across-species comparisons of connectivity.

How brain areas interact to direct attention

Attention can increase the activity of neurons representing important elements of our environment, or decrease the activity of neurons representing irrelevant information. These attention effects result from feedback from higher brain areas or bottom-up mechanisms. An outstanding question is how the brain co-ordinates these interactions between areas. To address this question, we simultaneously record neural activity in visual cortex and thalamus of monkeys performing cognitive tasks. Evidence suggests that the degree of synchrony between activities in interacting brain areas plays a major role in determining attention effects.

Overview

Methods

Back