Temporal Information of Directed Causal Connectivity in Multi-Trial ERP Data using Partial Granger Causality.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Signal Processing

Background:

• Partial Granger causality (PGC) is used to analyze neural connectivity, mitigating confounding variables.
• The temporal evolution and application of PGC to nonlinear models and event-related potentials (ERPs) remain underexplored.

Purpose of the Study:

• To evaluate the accuracy and robustness of time-domain PGC in nonlinear neural circuit models and ERP data.
• To investigate the temporal dynamics of causal connectivity using PGC.

Main Methods:

• Applied time-domain PGC to simulated nonlinear neural circuit models.
• Validated PGC against conditional Granger causality and partial directed coherence.
• Analyzed ERP data from an auditory oddball paradigm using PGC with a sliding time window.

Main Results:

• PGC demonstrated superior accuracy and robustness in detecting connectivity patterns in nonlinear circuits compared to other methods.
• PGC revealed faster convergence to stable connectivity configurations.
• Analysis of auditory ERPs identified significant causal influences in temporal, frontal, parietal, and cingulate areas, with stable connectivity emerging around 250–300 ms.

Conclusions:

• Time-domain PGC is a promising tool for accurately deciphering directed functional connectivity in nonlinear systems and ERP data.
• PGC offers early detection of causal connectivity dynamics and can reduce computational time for neural circuit modeling.