My research mostly focuses on developing a new mathematical framework to analyze, understand, and explain functional connectivity networks both in human brains and artificial systems.
More specifically, I am interested in defining new tools to detect the presence of a pathology in the human brain and provide new information and insights on the way the pathology affects brain connectivity. I mainly work in resting-state functional connectivity networks and I have applied my proposal to Parkinsonian Patients with different symptoms or just-diagnosed and Comatose Patients.
Concerning artificial systems, I am interested in the interpretability of the mechanisms of artificial neural network information processing, particularly in the presence of adversarial attacks or catastrophic forgetting phenomena.
I worked under the supervision of Sophie Achard and Michel Dojat, and given the multidisciplinarity of my research subject, I collaborated with multiple teams with different research interests.
My Ph.D. research was partially funded by the MIAI - Multidisciplinary Institute of Artificial Intelligence in Grenoble in the context of the research axis Towards Robust and Understandable Neuromorphic Systems which aims to design bio-inspired algorithms to improve the behavior of deep neural networks.
I conducted my research in two teams and research labs: the Grenoble Institute of Neuroscience and the Laboratory Jean Kuntzmann, respectively in the Functional Neuroimaging and Brain Perfusion team and the Statify team located at Inria - Rhone Alpes research center.