A shared, flexible neural map architecture reflects capacity limits in both visual short-term memory and enumeration

Area of Science:

• Cognitive Neuroscience
• Neuroimaging
• Computational Neuroscience

Background:

• Human cognition exhibits significant capacity limitations, restricting the number of items processed simultaneously.
• The neural basis of these capacity limits across different cognitive tasks remains a subject of debate.
• Investigating shared neural mechanisms can elucidate the flexibility of cognitive systems.

Purpose of the Study:

• To determine if capacity limitations in visual short-term memory (vSTM) and enumeration tasks rely on a common neural system.
• To identify the specific brain regions involved in shared capacity limitations.
• To model the neural mechanisms underlying these limitations.

Main Methods:

• Functional magnetic resonance imaging (fMRI) was used to measure brain activity in adults performing vSTM and enumeration tasks.
• Multivariate pattern analysis (MVPA) was employed to decode item quantities from brain activity.
• A computational model simulating a saliency map architecture was developed to interpret the findings.

Main Results:

• Task-specific capacity limits for vSTM (2-3 items) and enumeration (3-4 items) were neurally reflected in the posterior parietal cortex (PPC).
• An identical set of voxels in the PPC showed activity patterns that varied according to task-specific capacity limits.
• MVPA successfully decoded item numbers in enumeration across a wider range than in vSTM.
• A saliency map model with mutual inhibition successfully simulated the observed capacity limitations.

Conclusions:

• The posterior parietal cortex (PPC) houses a common, flexible neural system responsible for capacity limitations across distinct cognitive tasks.
• This system's flexibility allows it to adapt to varying precision requirements, such as high precision for vSTM and lower precision for enumeration.
• A saliency map architecture provides a plausible computational framework for understanding these shared capacity limitations.