Simulating EEG hyperscanning data using coupled Kuramoto oscillators following stochastic delay differential dynamics and using connectome data. Benchmarking information-theoretic measures against standard connectivity measures.
A. Brain Criticality Dynamics (criticality.py)
- Plot criticality dynamics of Kuramoto model of simulated source signals. Varying
$C_{\text{intra}}$ [0,1] in 50 steps, 20 iterations, and three noise conditions (none, medium, high).
B. Finding Best Cintra (best_cintra.py, mahalanobis_distance.py)
- Calculating the Mahalanobis distance between the MI Gaussian connectivity matrices of various single-brain simulations —
$C_{\text{intra}}$ [0.45,7], 25steps — and real resting-state EEG datasets (Gifford, Pérez).
- Simulating source neural dynamics (simulations.py)
- Calculate the phases of each oscillator (N=180) in a large Kuramoto model which follows stochastic delayed differential dynamics.
- Convert all the phases into simulated EEG data using forward gain model (N=64).
- Plotting extensive parameter space, varying inter-brain connectivity (
$C_{\text{inter}}$ ), biological noise (phase_noise, freq_std), and external/recording noise (amp_noise, sensor_noise).
-
Calculating Connectivity Measures (IB_analysis.py)
a. Calculate Standard Connectivity Measures: PLV, PLI, wPLI, CCorr, COH, iCOH, envCorr, and powCorr
b. Calculate Mutual Information: Histogram, Box Kernel, Gaussian, KSG, and Symbolic Estimators
c. Calculate Integrated Information Decomposition Measures: Time-Delayed Mutual Information, Transfer Entropy, Pure Information Transfer, Redundancy, and Synergy
-
Statistical Analysis (statistical_analysis.R)