Emergence of extreme and super-extreme events in coupled neural ensembles via field-mediated interactions

Extreme events are a prominent phenomenon in nonlinear and complex systems, with far-reaching implications across disciplines due to their potential to trigger catastrophic transitions and drastically alter system dynamics. In neuroscience, one of the most critical manifestations of extreme events is epileptic seizures, whose unpredictable onset poses significant challenges for diagnosis and treatment. Understanding the underlying mechanisms of extreme event generation is thus crucial for improving predictability and control in neural systems. In this work, we investigate how subtle coupling mechanisms, specifically, interaction through a common external field, can induce the emergence of extreme and even super-extreme events in small ensembles of excitable neurons. Using a two-parameter bifurcation analysis, we systematically explore the routes to the birth and collapse of such events in a network of FitzHugh–Nagumo neurons coupled via both chemical and memristive synapses. We demonstrate that extreme events arise through the formation of a spiral chaotic attractor via an Afraimovich–Shilnikov homoclinic bifurcation, following the classical Shilnikov scenario. This transition is further validated through experimental simulations on the Simulink platform, confirming the robustness and physical plausibility of the observed dynamics. Our findings highlight the critical role of nonconventional coupling in triggering pathological-like bursts in neural networks, offering new insights into the dynamical origins of seizure-like events and potential strategies for their anticipation.