EBRAINS Powers Virtual Brain Twins to Personalise Schizophrenia Treatment

From Lab to Clinic

The research workflow, which has been developed as part of an international collaboration that involves 18 partner institutions from across Europe, begins with detailed computational models of brain regions, which are built based on decades of neuroscientific research. EBRAINS researchers simulate these brain models using EBRAINS’ advanced tools like NEST, Arbor, and The Virtual Brain (TVB), running on Europe’s most powerful supercomputers. “We’re talking about simulations that take hours, even days, to capture the complexity of brain dynamics,” explains Wouter Klijn from the Julich Supercomputing Centre.

The researchers use AI-driven techniques to reduce these massive simulations into faster, clinically practical “neural mass models.” These reduced models retain the essential dynamics of a brain region’s activity and, when embedded into the EBRAINS tool TVB, can be matched to fMRI data. The models are then personalised by integrating the data from individual patients’ anatomical and functional MRI scans. The simulation of these personalised digital brain twins holds the potential to help clinicians to better understand how a specific patient’s brain works and develop personalised treatment plans based on this new understanding.

EBRAINS Makes It Possible

“This isn’t just about having powerful tools – it’s about having them all work together,” says EBRAINS CSO and project leader Viktor Jirsa from the Institut de Neurosciences des Systèmes (INS) in Marseille. This is made possible by EBRAINS’ integrated infrastructure. 

The integration of EBRAINS tools is critical at every step: the Knowledge Graph provides standardised data; simulation platforms enable multiscale modelling; AI and upscaling methods bridge the gap between research and clinic; and the EBRAINS Software Delivery (ESD) system ensures seamless interoperability among more than 70 tools. This level of integration, where outputs from one tool seamlessly feed into others, is impossible with standalone tools. “Without the EBRAINS platform, we’d simply not be able to run this workflow,” answers Klijn, when asked whether the individual tools could also be used in isolation. 

EBRAINS’ unified environment also ensures FAIR data principles, GDPR compliance, and access to supercomputing resources, all of which are essential for translating complex neuroscience into clinical practice. The result is a workflow that not only accelerates discovery but also fosters collaboration across disciplines, from basic science to patient care.

The workflow’s success hinges on the integration. Standalone tools, no matter how advanced, cannot match EBRAINS’ ability to connect data, simulations, AI, and clinical applications in a single and FAIR environment.

Clinical Trial

To test the potential of the virtual brain twins for the treatment of schizophrenia, the EBRAINS research team will work closely together with medical doctors: A clinical trial in hospitals in Marseille, France, and Munich, Germany, is set to begin in early 2027. The trial will investigate the extent to which these virtual brain twins can help doctors select the most effective treatment methods.

If successful, the workflow could extend beyond schizophrenia to other brain disorders, from dementia to ADHD. Using virtual brain twins for personalised medicine has already shown potential for the treatment of epilepsy. The results of a 400-patient clinical trial carried out in France are currently being analysed. 

The researchers stress that to them, this is just the beginning. “EBRAINS stands for a future where neuroscience and medicine work hand in hand – and where neuromedicine can be faster, more data-driven and more personalised than ever before,” says Jirsa.