Stabilizing Fractional Dynamical Networks Suppresses Epileptic Seizures

TL;DR

This paper introduces a novel fractional dynamical network stabilization method that effectively suppresses epileptic seizures, based on intracranial EEG data analysis, showing significant amplitude reduction and state differentiation.

Contribution

The study presents a new fractional dynamical network approach for seizure suppression, demonstrating its effectiveness on real patient data and multi-scale analysis of epileptic states.

Findings

Suppressed 34 out of 35 spontaneous seizures.

Achieved an average 49% amplitude reduction.

Differentiated four epileptic brain states using fractal analysis.

Abstract

Medically uncontrolled epileptic seizures affect nearly 15 million people worldwide, resulting in enormous economic and psychological burdens. Treatment of medically refractory epilepsy is essential for patients to achieve remission, improve psychological functioning, and enhance social and vocational outcomes. Here, we show a state-of-the-art method that stabilizes fractional dynamical networks modeled from intracranial EEG data, effectively suppressing seizure activity in 34 out of 35 total spontaneous episodes from patients at the University of Pennsylvania and the Mayo Clinic. We perform a multi-scale analysis and show that the fractal behavior and stability properties of these data distinguish between four epileptic states: interictal, pre-ictal, ictal, and post-ictal. Furthermore, the simulated controlled signals exhibit substantial amplitude reduction ($49\%$ average). These findings highlight the potential of fractional dynamics to characterize seizure-related brain states and demonstrate its capability to suppress epileptic activity.