Living Brain for Targeted Epilepsy Surgery

The EPINOV project is one of 10 large-scale projects selected in the 3rd wave of the French scientific excellence call entitled « Recherche Hospitalo-Universitaire en santé » (RHU) operated by the National Research Agency (ANR). This work is supported by a public grant from the second “Investissements d’Avenir” program (reference: ANR-17-RHUS-0004).

The EPINOV project relies on a public-private partnership coordinated by an academic Institution, Aix-Marseille University (AMU) and composed of an industrial partner, Dassault Systèmes, two public health establishments, the Assistance-Publique Hôpitaux de Marseille (AP-HM) and the Hospices Civils de Lyon (HCL) and a public research institution, the French Institute of Health and Medical Research (Inserm).
In addition to the 5 main partners, EPINOV’s multi-centric clinical trial will bring together the main French health centers experts in epilepsy management and surgery located in Marseille, Lyon, Lille, Strasbourg, Paris, Grenoble, Bordeaux, Toulouse, Nancy and Rennes.

The research done in the context of the EPINOV project largely relies on access to key state-of-the-art technology, including: a high-performance computing cluster dedicated to neural modelling; The Virtual Brain software; a MEG platform; 3T and 7T MRI systems (the only 7T MRI system installed within a hospital in France); an epileptic patient unit expert in stereotactic EEG (sEEG) recordings at the hospital La Timone; and the HCL, ECRIN-certified data management center.

The EPINOV project aims to guide surgical strategies to improve epilepsy surgical prognosis using a novel approach of large–scale computerized brain modeling and simulation based on data of individual epileptic patients. Retrospective pilot studies have confirmed the feasibility of this approach and shown promising results in predictability of surgical results. This approach will be achieved by:

1. conducting a multicenter trial involving major French epilepsy surgery centers trained to use state-of-the-art medical neuroinformatic technology; and
2. developing a Virtual Brain-based simulation approach.

Phenomenologically, epilepsy is defined by the onset and offset of high-frequency neural discharges. In order to model and better describe these discharges, a novel mathematical model was developed that is able to predict details of epileptic seizure discharge properties. This model describes fast spiking, spike wave events, the dynamics of seizure onset, time course and offset and also postulates of the existence of a yet to be identified slow variable. A theoretical mathematical analysis of epileptic discharges allows researchers to classify and predict seizure types, highlighting the relevance of conducting theoretical work to better understand brain function and dysfunction.

With the aim of generating large-scale personalized brain network models of epileptic patients, the EPINOV partners have merged the TVB connectome-based approach with the mathematical model and data fitting strategies to give rise to the Virtual Epileptic Patient (VEP). VEP simulations, data fitting and mathematical analysis pipeline result in the prediction of the most likely seizure propagation patterns through the patient’s brain and allow the exploration of personalized brain intervention strategies. Personalized data obtained in a clinical trial setting will help improve the accuracy of localization of the epileptogenic network, thus guide optimal surgical strategy and consequently would lead to a better surgery outcome. The EPINOV project will then be able to evaluate the impact of the VEP technology on surgical strategies and surgical morbidity at the individual patient level. 

The EPINOV kick-off took place on January, 8-9th, gathering the EPINOV partners and ANR. Here are the principal investigators, work package leaders and academic leaders.