Pierre-Antoine Comby

Post-doctoral researcher in AI for Medical Imaging (mostly MRI reconstruction)

I develop computational methods that push anatomical and functional MRI toward higher spatial and temporal resolution, with one goal: making ultra-high-field fMRI a practical tool for neuroscientists. My work combines MR physics, inverse problems and deep learning. It spans non-Cartesian acquisition, accelerated reconstruction and realistic simulation, and it ships as open-source Python software used well beyond my own team, such as MRI-NUFFT and SNAKE.

I am currently a Post-Doctoral fellow at Neurospin (CEA), working on innovative functional MRI acquisition and reconstruction methods using deep learning. I foster reproducible science and contribute to open-source software.

I got a PhD on "Physics and Medical Imaging" from Universite Paris-Saclay, after my work at the interface between Deep learning, Inverse Problems and Neuroscience in the MIND and METRIC Team at Inria and Neurospin, at CEA Saclay (France), supervised by Philippe Ciuciu and Alexandre Vignaud.

• My Resumé
• My thesis manuscript and the recording of my thesis defense
• pierre-antoine.comby@(cea|inria|ens-paris-saclay).fr