Performance Baseline of Phase Transfer Entropy Methods for Detecting Animal Brain Area Interactions.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Systems Neuroscience

Background:

• Phase transfer entropy (TEθ) methods show promise in animal sensory-spatial associative learning.
• The comparative advantages and limitations of TEθ methods are not well-defined, hindering their application.
• Establishing a performance baseline is crucial for selecting appropriate TEθ methods.

Purpose of the Study:

• To establish a performance baseline for phase transfer entropy (TEθ) methods.
• To identify the most suitable TEθ method for analyzing directional coupling in neural data.
• To investigate neural interactions between the hippocampus (Hp) and nidopallium caudolaterale (NCL) in pigeons during visual-spatial associative learning.

Main Methods:

• Four TEθ methods were applied to simulated neural signals from a neural mass model.
• The methods were tested on ferret neural data with known interaction properties to assess accuracy, stability, and computational complexity.
• The optimal TEθ method was selected and applied to local field potential data from pigeons.

Main Results:

• A performance baseline table was generated, indicating the most suitable TEθ method for various scenarios.
• The chosen TEθ method revealed a preferential information flow from the hippocampus (Hp) to the nidopallium caudolaterale (NCL) in pigeons.
• This interaction was specifically detected within the θ frequency band (4-12 Hz) during visual-spatial associative learning.

Conclusions:

• The study provides a performance baseline for TEθ methods, aiding in method selection for brain connectivity analysis.
• The findings highlight a specific directional information flow between Hp and NCL in avian visual-spatial associative learning.
• This research offers a valuable reference for utilizing TEθ methods to detect inter-areal brain interactions.