Meet the expert of eBRAIN-Health
Dr. Bertrand Thirion
Dr. Bertrand Thirion is a senior researcher in the MIND team at the National Institute for Research in Digital Science and Technology (INRIA) in France, where he develops statistical and machine learning methods for brain imaging, with a focus on functional neuroimaging. Alongside his research, he has made substantial contributions to open-source software, helping ensure that advanced analytical tools are widely accessible to the scientific community.
Role within eBRAIN-Health
Within eBRAIN-Health, Dr. Thirion and his team contribute primarily to Work Package 8, “Big Data Analytics”, where they focus on developing methods that make AI models interpretable. His work aims to ensure that machine learning systems used in neuroscience are not only high-performing but also transparent, enabling researchers to better understand how predictions are generated.
Recent Results and Key Milestones
A major milestone for the team was the publication of a new method at the International Conference on Machine Learning (ICML 2025). This approach provides a novel way to compute variable importance in machine learning models for treatment effect estimation, offering researchers clearer insights into the factors that drive model predictions.
Hurdles and How They Were Overcome
One of the main challenges has been the limited availability of public medical datasets, which constrains model training and evaluation. To address this, the team has relied on publicly available resources such as ADNI and OASIS, while noting that health data accessibility remains a major impediment to the development of AI-powered European medical research. Continued efforts to clarify access frameworks and improve technical data accessibility would further support progress in this field.
Contribution to the eBRAIN-Health Research Platform
Dr. Thirion’s work plays a central role in strengthening the interpretability layer of the eBRAIN-Health research platform. By delivering transparent and unbiased AI methods, his contributions support researchers in understanding the models they create. This work promotes more trustworthy and responsible AI for brain health and helps advance the platform toward its goal of modelling and simulating complex neurobiological phenomena.
