Neural Activity Predicts Reaction in Primates Long Before a Behavioral Response

Area of Science:

• Neuroscience
• Neural Engineering
• Cognitive Science

Background:

• Predicting behavior from neural activity is crucial for real-time brain-computer interfaces (BCIs).
• Previous research focused on neural activity shortly before an action, leaving early predictive signals unclear.

Purpose of the Study:

• To investigate if neural activity occurring long before a decision can predict subsequent behavior.
• To determine the predictive power of different neural signals (LFPs vs. spikes) for behavioral outcomes.

Main Methods:

• Recorded extracellular neural activities, including local field potentials (LFPs) and neural spikes, from the visual cortex of rhesus monkeys.
• Analyzed neural signals occurring more than 2 seconds before behavioral responses.

Main Results:

• Both LFPs and neural spike rates recorded long before a response predicted behavioral performance in a spatially selective manner.
• Local field potentials (LFPs) demonstrated a stronger predictive capacity for behavior compared to neural spikes.
• LFP amplitude showed a positive correlation with behavioral reaction time, while spiking activity showed a negative correlation.

Conclusions:

• Neural activity, particularly LFPs, occurring significantly before an action can predict future behavior.
• LFPs are a promising signal for early behavior prediction in brain-computer interface applications.
• These findings suggest LFPs contribute to the speed of information processing in sensory neural circuits.