CORTICAL DYNAMICS SHAPING HIPPOCAMPAL RIPPLE EMERGENCE AND FEATURES IN THE HUMAN BRAIN

Hippocampal sharp-wave ripples (SPW-Rs) are brief high-frequency oscillations that support memory consolidation but are also closely related to pathological events such as interictal epileptiform discharges (IEDs). SPW-Rs are coordinated across widespread cortical and subcortical regions, and it has been hypothesized that the engagement of distinct brain areas biases ripple content. Yet, the contribution of distributed cortical activity to ripple generation in humans and its relationship to pathological IEDs remain poorly understood. Here, we investigated the network mechanisms underlying hippocampal SPW-R generation using intracranial EEG recordings from five epilepsy patients implanted with high-density depth electrodes (128 channels) targeting the hippocampus, entorhinal cortex, lateral temporal cortex, inferior and superior parietal cortex, and prefrontal regions, including orbitofrontal and dorsolateral prefrontal cortex. We characterized SPW-R and IED morphology using low-dimensional embeddings derived from waveform features, applying Uniform Manifold Approximation and Projection (UMAP) to capture variability across events. The event-specific UMAP embedding space (SPW-Rs and IEDs) was then related to large-scale cortical dynamics by quantifying frequency-resolved LFP
components across temporal, parietal, and frontal regions preceding and following each event. We then applied multivariate cortical decoding approaches to predict SPW-R morphological features from distributed cortical dynamics and to evaluate their relationship to transitions toward pathological IEDs. This work will provide a systems-level framework to study SPW-Rs generation in the human brain, bridging physiological memory-related dynamics and pathological network events, and offering new insights into how large-scale cortical states bias hippocampal activity toward adaptive or maladaptive outcomes.