Is the Organization of the Visual System Distributed or Modular, or Both?

A new research report supports the idea that the brain is organized into subcomponents or modules, each dedicated to processing and representing a particular type of visual information (Downing, Jiang, Shuman, and Kanwisher, 2001). But other new reports support the idea that brain processing and representation are distributed, that is, any information is processed by many different parts of the brain and any brain region is likely to represent many classes of information (Haxby et al., 2001; Vaina et al., 2001). Cohen and Tong (2001) discuss the reports of Haxby et al. (2001) and of Downing et al. (2001) and consider both the modular and distributed theoretical positions and also other formulations that might encompass both kinds of findings.

Dowling et al. (2001) studied fMRI responses of the brains of 11 human subjects and identified a region in the right lateral occipitotemporal cortex that produced a significantly stronger response when subjects viewed photographs of the human body or body parts than when they viewed nonhuman mammals or inanimate objects. This region also responded more strongly to line drawings or stick figures representing the human body than to control stimuli. This extrastriate body area (EBA) did not overlap with areas previously found that respond preferentially to human faces (the fusiform face area) or to spatial locations (the parahippocampal place area); in some subjects the EBA did overlap with the visual motion area (MT/V5) or the lateral occipital complex that responds preferentially to shapes, but in all 11 subjects the EBA was distinct from the other areas. Although the EBA responds specifically to the body or most body parts, it does not respond strongly to faces.

Most reports of brain imaging studies show the peaks (and sometimes the valleys) of activation but do not show the entire topology of activity. Figure 1.7 (p. 12) shows examples of this concentration on peaks, as do many other figures in the text. In contrast, Haxby and colleagues (2001) studied the entire patterns of responses in ventral temporal cortex while subjects viewed black-and-white photographs of 8 categories of objects: human faces, cats, houses, chairs, scissors, shoes, bottles, and nonsense patterns. Each kind of stimulus produced a distinct pattern of activation across the ventral temporal cortex. Most importantly, even when the subregion that responded maximally to a given category (which is often the only region reported in publications) was excluded from analysis, the category being viewed could be identified accurately on the basis of the rest of the pattern. "Patterns of response that discriminated among all categories were found even within cortical regions that responded maximally to only one category" (Haxby, p. 2425). Therefore Haxby et al. conclude that "the representations of faces and objects in ventral temporal cortex are widely distributed and overlapping" (Haxby, p. 2425).

An even wider distribution of brain activity was investigated by Vaina et al. (2001) in a study of perception of biological motion. Biological motion is observed when a viewer sees a pattern of point-lights attached to the head and major joints of a human walker. Unless the walker locomotes, the pattern of lights remains obscure, and this is also true if the lights are translated in space without relative movement among them. In a control task, subjects reported only the overall direction of motion of a similar pattern of point-lights. Functional MRI recordings showed that brain activity specific to recognition of biological motion appeared in several regions: the lateral cerebellum, a region in lateral occipital cortex previously shown to be sensitive to moving contours, and in the both the dorsal and ventral extrastriate cortical regions. The latter two regions are sometimes called the "where" and "what" visual processing pathways [see Figure 10.26(a)]; it has been suggested that these processing pathways converge in the superior temporal gyrus. The investigators suggest that recognition of biological motion may require activation of both pathways as well as their confluence in the superior temporal gyrus. "This hypothesis is consistent with our findings in stroke patients with unilateral brain lesions involving at least one of these areas, who, although correctly reporting the direction of the point-light walker, fail on the biological motion task" (Vaina, p. 11656). The involvement of the lateral cerebellum may have been caused by the necessity of spatial attention (mediated by this cerebellar region as well as by other brain regions) in order to determine whether the moving dots portray a person walking.

These and other studies indicate that wide regions of the brain are involved in any task, which suggests we use caution in interpreting brain imaging studies. It is important to know which brain regions are most active during a particular task, but we should not conclude that these regions are working alone.

References:

Cohen, J.D. and Tong, F. (2001). The face of controversy. Science, 293, 2405-2407.

Downing, P.E., Jiang, Y., Shuman, M. and Kanwisher, N. (2001). A cortical area selective for visual processing of the human body. Science, 293, 2470-2473.

Haxby, J.V., Gobbini, M.I., Furey, M.L., Ishal, A., Schouten, J.L., and Pietrini, P. (2001). Distributed and overlapping representations of faces and objects in ventral temporal cortex. Science, 293, 2425-2430.

Vaina, L.M., Solomon, J., Chowdhury, S., Sinha, P. and Belliveau, J.W. (2001). Functional neuroanatomy of biological motion perception in humans. Proceedings of the National Academy of Sciences, U.S.A., 98, 11656-11661.