D-128 | A Multiscale Symbolic Approach to Decoding Delta and Ripple Oscillations Bands as Biomarkers for Epileptiform Discharges

Theoretical and Computational Neuroscience
Author: Mauro Granado | Email: granadomauro@gmail.com

Mauro Granado^1°, Federico Miceli^1°, Nataniel Martinez^2°,  Lijandy Jiménez Armas^3°, Fernando Montani^1°

^1° Instituto de Física La Plata – Universidad Nacional de La Plata – CONICET
^2° Instituto de Investigaciones Físicas De Mar De Plata – Universidad Nacional de Mar del Plata – CONICET
^3° Instituto de Matemática Aplicada del Litoral – Universidad Nacional del Litoral – CONICET

We apply a multiscale symbolic approach to analyze the complex dynamics of temporal lobe refractory epilepsy using high-resolution intracranial EEG (iEEG). Focusing on basal and preictal phases, we examine frequency bands up to 240 Hz, revealing key periodicities and time scales in neural dynamics. By band-pass filtering signals into delta, theta, alpha, beta, gamma, and high-frequency oscillations (HFO), we assess distinct nonlinear dynamics. Our method identifies critical time lag scales (τ) within these bands, crucial for studying refractory epilepsy. Metrics like permutation entropy (H), Fisher information (F), and complexity (C) uncover nonlinear patterns, revealing intrinsic τ that maximize complexity. Comparing basal and preictal signals, we identify significant differences in the delta and 200-220 Hz (HFO 6) bands. These differences in Fisher information before seizures underscore the importance of delta oscillations and HFO waves as potential biomarkers, offering new insights into focal epilepsy dynamics.