Frequency-separated principal component analysis of cortical population activity

Area of Science:

• Neuroscience
• Computational Neuroscience
• Systems Neuroscience

Background:

• Neocortical activity exhibits coordinated, low-dimensional firing rate fluctuations across neurons.
• The precise role of these population-level fluctuations in sensory processing is not fully understood.

Purpose of the Study:

• To investigate the impact of neural population fluctuations on sensory processing in the primary visual cortex (V1).
• To develop and apply a novel method for characterizing neural fluctuations across different frequency bands.

Main Methods:

• Examined population activity of orientation-selective neurons in macaque V1 during visual stimulation and spontaneous activity.
• Introduced and utilized frequency-separated principal component analysis (FS-PCA) to analyze neural fluctuations.
• Characterized frequency components, eigenvalues, and variance, observing a power-law distribution.

Main Results:

• FS-PCA revealed a broad spectrum of neural activity components, predominantly low-frequency.
• Subpopulations of V1 neurons correlated positively or negatively with low-dimensional fluctuations.
• These subpopulations exhibited distinct activation patterns and noise correlations during stimulus viewing.

Conclusions:

• Slow, low-dimensional fluctuations in V1 population activity significantly shape individual neuron responses to oriented stimuli.
• These population dynamics may influence the transmission of sensory information to downstream visual areas.
• The study highlights the importance of population-level activity patterns in sensory coding.