Effective Connectivity in Brain Networks During Episodic Encoding

Figshare hosts a computational study by Hrishikesh Pable on neural self-organization during episodic encoding. The research analyzes source-localized EEG data to estimate deep recurrent effective connectivity among the Salience, Default Mode, and Task-Positive Networks. It investigates state-dependent interactions during rest and encoding phases for geometric shapes and words, incorporating individual differences in mental visual imagery.

Use Cases

• Analyze state-dependent bidirectional connectivity patterns between the Salience Network (SN), Default Mode Network (DMN), and Task-Positive Network (TPN) during rest versus encoding.
• Model the mediating influence of Vividness of Visual Imagery Questionnaire (VVIQ) scores on connectivity strength from SN to TPN for different stimulus types.
• Compare inferred effective connectivity for geometric shapes versus word encoding to understand stimulus-specific neural dynamics.
• Study the reversal in DMN-to-SN influence across cognitive states (rest vs. encoding) using a biologically-informed recurrent neural network framework.

Strengths

• Dataset is derived from a biologically-informed computational framework incorporating neurobiological constraints like Dale's Law and separate excitatory/inhibitory populations.
• Analysis focuses on the large-scale triple-network framework (SN, DMN, TPN), a major model in cognitive neuroscience.
• Integrates individual difference data via the Vividness of Visual Imagery Questionnaire (VVIQ).

Limitations

• Dataset is very small (15.2 KB), containing a research document rather than raw or processed tabular data for direct analysis.
• No column definitions, sample data, or row counts are provided, limiting immediate utility for machine learning.
• The primary content is a methodological description and findings report, not a structured dataset for model training.

Provenance

Source

Hrishikesh Pable via figshare.

Collection Method

Computational framework integrating sLORETA source-localized EEG with a biologically-informed recurrent neural network.

Freshness

Last updated March 23, 2026.