Uni-Logo
You are here: Home BFNT Research A4: Adaptive interception of epileptic events based on predictive network dynamics
Document Actions

A4: Adaptive interception of epileptic events based on predictive network dynamics

Ulrich EgertR


R
= Biomicrotechnology


Scientific background

The response of neuronal networks to stimulation depends on the recent network dynamics and its expected evolution in time. Despite the multitude of studies exploring basic neuronal mechanisms associated with epileptic seizures, predicting when, how or why a seizure occurs in humans remains unreliable. If it were possible to identify signatures in the network activity indicative of epileptic dynamics, therapeutic intervention could be improved.
Although computational models aid the understanding of network dynamics and their relation to mass signals, the loop between the detection of upcoming seizures and successful intervention remains unclosed. In prior work we identified a reliable, coherence-based measure within a suitable timeframe of the seizure initiation process in a mouse model for mesial temporal lobe epilepsy (MTLE), developed network simulations of epileptiform EEG, machine-learning algorithms for spike time prediction as well as seizure detection and epilepsy type identification in human EEG recordings, and characterized inhomogeneous epileptogenic networks in mice that are similar to human MTLE network structures.


Objectives

We aim (I) to understand both, the network dynamics and structures involved in seizure generation, (II) to develop control systems for adaptive stimulus configuration, and (III) to subsequently utilize these to interfere with upcoming epileptiform network dynamics.

Personal tools