Robust Coding of Eye Position in Posterior Parietal Cortex despite Context-Dependent Tuning

Area of Science:

• Neuroscience
• Cognitive Neuroscience
• Systems Neuroscience

Background:

• Neurons in the posterior parietal cortex (PPC) process sensory information, body posture, and action plans.
• Nuisance variables, like eye position, can corrupt relevant information due to context-dependent neural activity.
• Understanding how the PPC represents gaze direction across contexts is vital for spatial perception and brain-computer interfaces.

Purpose of the Study:

• To investigate the robustness of eye position representation in the macaque PPC across various behavioral and sensory contexts.
• To determine if the coding of eye position is invariant to task context in the PPC.
• To assess the reliability of PPC population codes for sensorimotor transformations.

Main Methods:

• Utilized demixed principal component analysis (dPCA) on neural spike counts from male macaque PPC.
• Defined contexts based on different stages of a visually guided reaching task.
• Employed context-naive decoders to estimate eye position across contexts.

Main Results:

• Eye position was robustly represented in the PPC population response across different task contexts.
• The subspace encoding eye position was orthogonal to the subspace encoding task context.
• Context-naive decoders accurately estimated eye position, with errors comparable to context-optimized decoders.

Conclusions:

• Population codes in the PPC effectively shield eye position signals from contextual crosstalk.
• The PPC provides reliable egocentric spatial information invariant to behavioral and sensory contexts.
• These findings support the role of PPC in robust sensorimotor transformations and spatial cognition.