A novel extended Granger Causal Model approach demonstrates brain hemispheric differences during face recognition...

Area of Science:

• Neuroscience
• Computational Biology
• Systems Biology

Background:

• Dynamic Causal Modeling (DCM) and Granger Causal Modeling (GCM) are distinct methods for analyzing time-series data in biological systems.
• These methods are widely used in neuroimaging and applicable to gene, protein, and metabolic pathway temporal changes.
• Existing approaches are typically used independently, limiting comprehensive causal inference.

Purpose of the Study:

• To introduce a novel approach extending GCM, integrating features of DCM's bilinear approximation.
• To validate the extended GCM using simulated data and in vivo electrophysiological recordings.
• To investigate learning-induced changes in neural connectivity and oscillatory activity.

Main Methods:

• Development of an extended Granger Causal Model (GCM) incorporating bilinear approximation features.
• Extensive testing of the extended GCM on simulated datasets in both time and frequency domains.
• Application to in vivo multi-electrode array recordings of local field potentials in sheep.

Main Results:

• The extended GCM demonstrated efficacy in simulated and real-world data.
• Face discrimination learning induced significant changes in inter- and intra-hemispheric connectivity.
• Learning altered the hemispheric predominance of theta and gamma frequency oscillations in the inferotemporal cortex.

Conclusions:

• The novel extended GCM provides a unified framework for causal inference in biological time-series data.
• First evidence is presented for learning-dependent connectivity modifications within and between sheep inferotemporal cortex hemispheres.
• The findings highlight the model's utility in understanding neural plasticity related to learning and memory.