Synaptic plasticity, the activity-dependent modification of neural connections, is a fundamental mechanism underlying learning, memory, and pathological brain states. Epilepsy, a chronic neurological disorder affecting 0.8% of the population, often proves drug-resistant, with half of treated patients experiencing persistent seizures and progressive cognitive decline. This study investigates how functional connectivity (FC), derived from EEG signals and modeled via graph theory, varies between basal, preictal, and postictal brain states to improve preictal state prediction. The research is grounded in the principle that synaptic plasticity modulates network strength based on use, favoring efficient, metabolically economical information transfer between frequently synchronized nodes. However, it is hypothesized that in epilepsy, this same mechanism may lead to pathological hypersynchronization. We explore the dynamics of cortical networks by analyzing functional connectivity graphs, testing the premise that preictal states are characterized by a measurable reorganization of network properties, reflecting aberrant plasticity processes. This work aims to provide a novel, network-based biomarker for seizure prediction by quantifying these critical transitions in brain dynamics.