The Study Of Neural Mechanism On Visual Attention

Attention is a fundamental cognitive activity in human. Attention can improve the detection ability of behavior and shorten the reaction time. Since attentional deficit is related with many clinical disorders, such as schizophrenias, attention deficit hyperactivity disorder (ADHD), Alzheimer's disease and hypertension, to study neural mechanism of attention in normal human is the base of further clinical research. Using neuroimaging technology, researchs found that when human attented to a visual stimuli or event, there were more than thirty-two brain areas tend to activity. However, how do those active brain areas connect and communicate to represent the visual attention? Which areas' deficit or communicative disorder evokes the corresponding disease? The answers of these questions will be depending on the research of visual attention network. The research methods of attention network include psychology experiments, single-unit recordings, functional magnetic resonance imaging (fMRI), positron emission computerized topography (PET), electroencephalogram (EEG), attentional networks test (ANT), mathematical modeling and so on.This dissertation focused on the research of neural mechanism of visual attention. By the tool of event-related potential (ERP) and methods of modern signial processing, the deep, detailed and novel researchs on visual attention were accomplished. Our major works and achievements are listed as following:(1)A new method of spatio-temporal topographic mapping by correlation coefficient of K-means cluster was developed.The results of simulations and real data demonstrate the validity of the method in mapping correlated sources. The method successfully decomposed the ERPs collected in a visual attention experiment into three clusters located at left, right occipital and frontal.The estimated vectors of the contra-occipital area demonstrate that attention to the stimulus location produces increased amplitude of the P1 and N1 components over the contra-occipital scalp. The estimated vector in the frontal area displays that there are two large processing negativity waves around 100 ms and 250 ms when subjects are attentive, and there is a small negative wave around 140 ms when unattentive. Above findings indicated that frontal and parietal cortex involved in the early sense and perception processes in visual attention. To study the relationships among active areas, the global connectivity was estimated based on the phase synchronization cluster approach. The analyses of above real data related to human visual attention study showed that three connectivity sources could be detected and the difference of connectivity between attention and unattention experiments is quite distinct. The results support the results of our new mapping method of correlative sources.(2) A novel common spatial pattern (CSP) decomposition method and the standardized low resolution brain electromagnetic tomography (sLORETA) were applied to study the spatiotemporal dynamics of visual attention.The spatial patterns indicated that visual cortex, prefrontal cortex (PFC), anterior cingulate cortex (ACC) and posterior parietal cortex (PPC) were involved in the control of top-down attention. The temporal waveforms indicated that contralateral PFC and PPC were activated synchronously at about 150 ms after the stimulus onset, with early attention effects only occurring in PFC, and the PPC were activated early than that of PFC during 200-260 ms. The results imply that humans adopt different allocation strategies for resources in visual attention and un-attention situations. For attention case,visual cortex consumes the most resources and for non-attention situation, the ACC and PPC consume the most resources. We proposed a model of timing relationships among frontal, parietal and visual cortex based on those results.(3) The fronto-parietal network under the control of bottom-up and top-down visual attention was investigated by a visual target detection experiment.Reaction times and P300 latency in frontal areas were shorter under bottom-up than top-down control. Peak parietal P300 was generated under bottom-up control and medial fronto-central P300 was generated under top-down control.There was increased power in parietal areas for bottom-up control, and increased power in right and left frontal regions for top-down control within 200-650 msec and 2-24 Hz. The results provide evidence that control of bottom-up and top-down attention result from different cognitive processing in fronto-parietal network.