Spatiotemporal and sensory modality attention processing with domain-specific representations in frontoparietal areas

Area of Science:

• Neuroscience
• Cognitive Neuroscience
• Neuroimaging

Background:

• The frontoparietal network (FPN) is implicated in attention processing across spatial-temporal and sensory domains.
• It remains unclear if the FPN encodes generalizable representations across these attentional subdomains.

Purpose of the Study:

• To investigate whether the frontoparietal network (FPN) contains neural representations of attention that generalize across different sensory modalities and spatial-temporal domains.
• To identify domain-specific and generalizable attentional representations within the FPN and other brain regions.

Main Methods:

• Functional magnetic resonance imaging (fMRI) data were acquired from 20 participants.
• Multivariate activity patterns were decomposed to identify latent representations generalizing across experimental conditions organized into unimodal-bimodal and spatial-temporal models.
• Representation similarity analysis was employed to compare brain activity patterns.

Main Results:

• The FPN emerged as the primary network modulating attention across domains.
• Activation patterns within the FPN reorganized based on specific attentional demands, particularly across different sensory modalities, indicating distinct regional neural representations.
• Domain-specific areas outside the FPN, such as the dorsolateral prefrontal cortex, were also identified.
• Two types of activated brain regions were revealed: those processing detailed attention domain information and those processing general information.

Conclusions:

• The frontoparietal network (FPN) plays a key role in attention processing, with its internal representations adapting to specific attentional demands.
• Distinct neural mechanisms within the FPN support detailed versus general information processing during attention.
• Attentional processing involves both FPN-based and extrastriate brain regions, highlighting a distributed network.