Information representation in an oscillating neural field model modulated by working memory signals

Area of Science:

• Computational neuroscience
• Neural dynamics
• Information theory

Background:

• Working memory (WM) influences sensory processing.
• Neural activity exhibits oscillatory dynamics.
• Stimulus information can be encoded in neural firing rates or spike timing relative to oscillations.

Purpose of the Study:

• To investigate how working memory (WM) signals modulate neural activity to represent stimulus information.
• To compare the efficacy of rate coding versus oscillatory phase coding for stimulus discrimination under WM modulation.

Main Methods:

• Developed a neural field model with oscillatory dynamics near an instability.
• Incorporated WM-like input to modulate oscillation frequency and amplitude.
• Defined and analyzed spike rate and oscillatory phase codes using information theory.

Main Results:

• WM input modulates both spike rate and oscillatory phase.
• Phase code discrimination is enhanced by WM input, while rate code discrimination is reduced.
• Phase code performance is significantly superior (two orders of magnitude) to rate code performance.

Conclusions:

• Oscillatory phase coding offers a powerful mechanism for representing sensory information, enhanced by working memory.
• Working memory signals can improve sensory representations in oscillatory activity beyond rate-based encoding.
• Findings have implications for understanding prefrontal cortex-sensory area interactions in cognition.