Primates, Visual Attention in

views updated

PRIMATES, VISUAL ATTENTION IN

The neural mechanisms for visual attention intervene between sensation and action. Neither strictly sensory nor motor, they constitute the neural filters that select certain aspects of visual information for greater attention and influence over action. Visual attention includes mechanisms for arousal, alerting, vigilance, and selective attention for spatial location or visual features. Of these, the ones that most directly affect visual processing are selective attention to a location in space or to a visual feature of an object. These latter mechanisms help route selected visual inputs to visual-processing structures in the brain and also to motor control systems, usually at the expense of nonselected inputs. Behavioral examinations of visual processing, functional-imaging studies in humans, and single-unit studies in the nonhuman primate indicate that selection mechanisms are in play. These experiments aim to understand the influence of attention on the processing of information and the neural signals that underlie selective attention.

Behavioral Experiments

Behavioral experiments show that individuals cannot and do not process all the information present in complex visual scenes. It is not possible, for example, to recognize within the same instant more than one or two objects in a typical complex scene. When subjects are asked to report different attributes of simultaneously presented images, they nearly always perform worse than when asked to report the attributes of the objects separately. In certain conditions, major changes in scenes go undetected because the capacity of the brain's limited capacity to code and process complex information. But, when subjects are told to attend to only one of two simultaneously presented objects, or to a particular location in a scene, their performance returns to the levels expected for only one object. Selective attention provides a mechanism for choosing which information is noticed and which is ignored, thus ensuring that unimportant information does not crowd out important information. Moreover, since visual acuity is poor outside of the center of gaze, the eyes must be moved to scan and process visual scenes. Selective attention provides a method for selecting single targets for the neural system that controls the eyes (oculomotor system).

Control of Attention

The selective filters that underlie selective attention operate in space and on features of objects, selecting locations and features of visual stimuli for heightened processing. The neurobiology underlying this process lends itself to separation into two parts: the mechanisms that control attention and the mechanisms that are, in turn, influenced by attention. In some cases, attention is controlled by features intrinsic to a stimulus. Some objects attract attention to themselves, an automatic process that occurs because of separation of figures from their background resulting from differences in color, contrast, motion ("popout"). Yet, even after automatic figure-ground separation, a typical visual scene will still contain many different figures of equal salience that are nevertheless of different behavioral relevance. In this case, attention can be actively wielded to filter information in the scene; a neural signal that reflects this active process is a "top-down" signal that controls attention.

Studies using functional imaging in human subjects have identified several target structures that may provide the signals that control selective visual attention in humans. The control system for spatially directed attention, identified by functional imaging, is both widely distributed and closely associated with the oculomotor system; it involves the posterior parietal cortex, parts of the frontal cortex, the pulvinar nucleus of the thalamus, and possibly the superior colliculus. The close association with the oculomotor system accords with the principle that an important role for selective attention is the direction of the eyes to the behaviorally relevant stimuli in complicated visual scenes. Neurophysiology in nonhuman primates performing attention tasks indicates a neural mechanism for the operation of attention in some of these structures. In these structures, a given neuron responds to stimuli in a limited region of space (i.e., its receptive field), and this response is enhanced whenever the animal directs its attention into the region. This response enhancement may provide the "top-down" signal that potentiates the processing of visual stimuli that appear in that location. When these signals are absent because of damage or dysfunction within any of these structures, particularly if the damage is uni-lateral, subjects suffer a variety of attentional disorders. One such attentional disorder is extinction, a difficulty in detecting or perceiving objects within the affected portion of the visual field when a competing object is located in intact portions of the field.

Influence of Attention

The neural signals derived from these brain areas influence the processing of visual stimuli in the sensory areas that process visual information. Functional imaging studies show that attention influences the processing of visual information in all visual areas, from the earliest visual area, primary visual cortex to the last visual areas, in both the ventral and dorsal visual processing streams. In these imaging studies, the instruction to attend to a stimulus can modulate the response to visual stimuli, altering the response to a visual stimulus by 10 to 30 percent. This effect is larger in later stages of visual processing, raising the possibility that the smaller modulations in primary visual cortex result from feedback from higher areas. The results in human imaging replicate those in neurophysiological recordings of behaving primates, although the influence of attention on responses in early visual areas in primates has not been demonstrated in single-unit studies. But extrastriate visual cortical areas that process information about object features such as color and shape (such as Visual Area 4), receive attentional control signals, and act as the local attentional filters. The properties of the neurons' responses to visual stimuli and the effect of attention reflect both behavior in the absence of attention and the influence of attention on that behavior. When more than one stimulus is placed in the receptive field of a single neuron, the neuron's activity reflects an averaging of the response to both of the stimuli.

The presence of more than one stimulus in the receptive field interferes with that neuron's ability to carry information about the stimulus. Behaviorally, this interference is mirrored in poor performance with simultaneously presented objects. But when an animal is instructed to attend to a single stimulus within the receptive field of the neuron, the response of the neuron will be determined almost solely by the color, shape, or orientation of the attended stimulus. In presence of the top-down signal wielded by attention, neurons communicate little information about the features of ignored stimuli, and the interference is filtered out of the signal. Thus, the neurons that process the visual information are influenced by the control signals for attention to act as filters of information that select relevant information and discard irrelevant information, providing a mechanism for sorting out the complexity of the visual world.

Bibliography

Kanwisher, N., and Wojciulik, E. (2000). Visual attention: Insights from brain imaging. Nature Reviews Neuroscience 1, 91-100.

Kastner, S., and Ungerleider, L. G. (2000). Mechanisms of visual attention in the human cortex. Annual Review of Neuroscience 23, 315-341.

Olson, C. R. (2001). Object-based vision and attention in primates. Current Opinion in Neurobiology 11, 171-179.

Rensink, R. A. (2002). Change detection. Annual Review of Psychology 53, 245-277.

Reynolds, J. H., and Desimone, R. (1999). The role of neural mechanisms of attention in solving the binding problem. Neuron 24, 19-29, 111-25.

RobertDesimone

Revised byBharathiJagadeesh