Analyzing High-Order and Pairwise Interactions in SEEG networks.

Epilepsy is a neurological condition in which seizures arise from disrupted communication across distributed brain networks. Rather than originating from a single, isolated focus, seizures often involve broader circuits that shift and reorganise in characteristic ways. Understanding how this coordination changes as the brain moves from a baseline reference state into the ictal period is key for interpreting seizure generation and its impact on network organisation. This thesis examines how coordination evolves at multiple scales around seizure onset, using stereo-electroencephalography (SEEG) recordings from individuals with focal, drugresistant epilepsy. Pairwise synchronisation is quantified through the Phase-Locking Value (PLV), while higher-order, triadic interactions are assessed using Circular Omega Complexity (COC). To capture the diversity of coordination patterns, the analysis also incorporates Shannon entropy computed from the distributions of PLV edges and COC triads across standard frequency bands. By comparing baseline segments with ictal activity, the study provides a multiscale view of how network organisation shifts during seizures. The results show a consistent pattern: pairwise synchrony increases, higher-order structure weakens, and coordination strengths become more homogeneous across the epileptogenic zone. Together, these changes suggest that seizures push the network toward a more constrained, lower-dimensional mode of interaction. Overall, the findings reinforce the view of epilepsy as a disorder of network organisation and highlight the importance of higher-order coordination measures for understanding ictal dynamics.