Functional Neuroimaging and Cognitive Behavioral Testing

Human Perception

Our research is concerned with visual attention, object perception and memory. We explore the nature of the limits to human perception, the information-processing that results in the perception of objects and events, and the nature of the representations that underlie both conscious experience and implicit memory, shown in perceptual priming. We mainly use behavioral methods, but we are interested in relating our findings to the brain. We study patients with brain damage, and collaborate in studies using brain imaging or evoked responses.

Visual attention, search, and the "binding problem”: What kinds of information are available without focusing attention when we are presented with multi-element arrays and what kinds require focused attention? What variables control the deployment of spatial attention? Physiological findings suggest that the visual system sets up multiple specialized "maps" coding different aspects of the scene; how then do we combine information about the separate features of objects. Experimental tasks include visual search, and divided attention paradigms in which irrelevant stimuli evoke competing responses. Recent neuroimaging findings using fMRI have also begun to define specific areas within the prefrontal cortex that appear to be activated when subjects successful bind information in memory as compared to comparison conditions that control for factors such as the mental effort or number of items that must be maintained. One student is working on cross-modal binding, looking at interactions between visual and auditory stimuli and the brain basis of these interactions. Treisman is also collaborating with Lynn Robertson at UC Berkeley in studies of patients with neglect due to parietal lesions linking their major deficits in spatial localization to the binding errors that they also make.

The encoding of statistical properties, testing the hypothesis that global attention distributed over sets of elements results in the representation of statistical descriptors like the mean, range and variance of sets of similar items. They will look at fMRI activation as a function of attentional spread, and at statistical averaging on two dimensions at once. They showed earlier that participants could generate the mean size for two different subsets of intermingled sizes (defined by different colors).

Object perception: The Treisman lab uses novel shapes to probe the nature of the representations that are formed in the absence of any matching representations or prior knowledge. One project explores generalization from specific learned orientations of novel objects to its three-dimensional shape. They are also interested in the role of attention in forming object representations, testing for example whether attention is necessary for the perception of occlusion relations, identity and meaning.

Conscious awareness: The lab explores some conditions under which information is taken in without being consciously accessible. For example, in a phenomenon known as the "attentional blink", subjects search for two targets in a stream of visual stimuli presented successively at high rates, and appear to be blind to the second of two targets when they have just detected the first. Awareness can also be blocked in "repetition blindness": a repeated stimulus in a rapid sequential stream is much less likely to be seen than a nonrepeated one.

Visual memory: Studies of visual working memory have explored the role of binding in the short term retention of objects and of locations, and compared the effects of load defined by number of items, number of features, or by the precision of specification for a single item. Again, the brain basis of visual memory is tested using fMRI and EEG measures. Priming studies in the Treisman lab have shown that detailed representations of unattended shapes are formed and can last for weeks without any awareness or explicit memory of them. They are studying the nature of these representations, the conditions under which they are formed, and their relation to explicit memories formed with attention. They also study perceptual learning through the effects of repeated exposure and practice in visual tasks.

Contact:
Anne Treisman

Return to Approaches