Abstract:The visual environment is complicated, and humans and other animals accordingly prioritise some sources of information over others through the deployment of spatial attention. We presume that attention has the ultimate purpose of guiding the abstraction of information from perceptual experience in the development of concepts and categories. However, neuroscientific investigation has focussed closely on identification of the systems and algorithms that support attentional control, or that instantiate the effect of attention on sensation and perception. Much less is known about how attention impacts the acquisition and activation of high-level information in the brain. Here, we use machine learning of EEG and concurrently-recorded EEG/MRI to temporally and anatomically characterise the neural network that abstracts from attended perceptual information to activate and construct semantic and conceptual representations. We find that the trial-wise amplitude of N2pc - an ERP component closely linked to selective attention - predicts the rapid emergence of information about semantic categories in EEG. Similar analysis of EEG/MRI shows that N2pc predicts MRI-derived category information in a network including VMPFC, posterior parietal cortex, and anterior insula. These brain areas appear critically involved in the attention-mediated translation of perceptual information to concepts, semantics, and action plans.

What problem does this paper attempt to address?

This paper explores how the human brain abstracts concepts and category information from perceptual information through attention mechanisms. The study used machine learning to analyze EEG (electroencephalogram) and simultaneously recorded EEG/MRI data to temporally and anatomically characterize the neural networks that transform attention-guided perceptual information into semantic and conceptual representations. The research found that the amplitude of the ERP (event-related potential) component N2pc, closely related to selective attention, can predict the rapid emergence of semantic category information in visual stimuli. Further analysis revealed that N2pc is associated with MRI category information in brain regions including the anterior cingulate cortex, posterior parietal cortex, and anterior insular cortex, which play a crucial role in attention-mediated information transformation. Through experiments, the researchers discovered that in high-density EEG experiments, the categorization of target object category information starts around 215 milliseconds after the stimulus, while in simultaneously recorded EEG/MRI experiments, this time point is approximately 240 milliseconds. Importantly, they found a correlation between the quality of attention and the N2pc amplitude and category classification accuracy, indicating that stronger attentional selection leads to enhanced expression of target object category information in neural signals. In summary, this paper aims to fill the gap in our understanding of how the brain utilizes attention to gather and construct conceptual information and reveals the crucial role of specific brain regions (such as the anterior temporal lobe, prefrontal cortex, and posterior parietal cortex) in this process.

Attentional propagation of conceptual information in the human brain