EPINOV — Improving EPilepsy surgery management and progNOsis using Virtual brain technology
Overview
EPINOV was a French national RHU (Recherche Hospitalo-Universitaire) clinical trial and research project, funded by ANR under PIA3 and running from January 2018 to November 2023. With a €5.3M budget, it was the first randomised controlled trial (NCT03643016) to test whether personalised whole-brain simulation could improve surgical outcomes in drug-resistant focal epilepsy. The trial enrolled 356 patients across 12 French epilepsy surgery centres. It was led by a team at the Institut de Neurosciences des Systèmes (INS, INSERM/AMU Marseille), with AP-HM, Hospices Civils de Lyon, and Dassault Systèmes as partners.
The Virtual Epileptic Patient (VEP)
EPINOV’s core scientific intervention was the Virtual Epileptic Patient (VEP): a personalised whole-brain network model constructed for each patient from their own SEEG (stereoelectroencephalography) data and structural MRI/DTI connectivity. Using the Virtual Brain (TVB) simulation framework developed within the Human Brain Project, each patient’s epileptogenic zone and seizure propagation dynamics were modelled pre-surgically. The hypothesis was that the VEP model could predict optimal resection targets more accurately than standard SEEG-based clinical assessment alone. The trial’s primary outcome was seizure-free status at 12 months post-surgery.
The SEEG-TVB pipeline driving the VEP models is the direct methodological descendant of TVB computational infrastructure developed under the Human Brain Project. EPINOV was operationally independent, with its own RCT protocol, funding, and clinical sites, while TVB provided its scientific foundation.
EBRAINS Dataset
Following trial completion, EPINOV published an open dataset of 30 Virtual Epilepsy Patient models on EBRAINS, providing the neuroscience community with fully documented, FAIR-compliant personalised brain models complete with the underlying SEEG data, connectivity matrices, and simulation parameters. This is one of the first open datasets of clinically derived computational brain models for epilepsy.
Timeline
- January 2018: Project launched, ANR/PIA3 funding confirmed
- 2018–2022: Patient recruitment and SEEG acquisition across 12 centres
- 2023: Trial closed; primary outcome analysis; VEP open dataset published on EBRAINS
- November 2023: Official project end
Connections
- fundedBy: ANR Open Science Policy
- registeredIn: ClinicalTrials.gov
- registeredIn: EBRAINS (VEP open dataset of 30 personalised models)
- relatedTo: Human Brain Project (TVB simulation framework underpinning the VEP methodology)
- relatedTo: Virtual Brain Twin (direct successor project)
Resources
- https://clinicaltrials.gov/study/NCT03643016 (RCT registration, NCT03643016)
- https://ins-amu.fr (INS, Institut de Neurosciences des Systèmes, lead institution)
- https://search.kg.ebrains.eu (EBRAINS Knowledge Graph, VEP dataset)
- https://doi.org/10.1016/j.neuroimage.2017.04.020 (Jirsa et al. 2017, NeuroImage, VEP methodology)