One of our scientific objectives is to study new methods for inverse problems related to various tomographic imaging modalities. They are based on classical variational approaches, on deep learning approaches, and are tightly linked with the physics of image formation and simulations (Axis II). Among the various projects, we can cite activities on spectral photon counting CT, ion imaging, blood velocity field estimation, region-of-interest image reconstruction, low dose tomographic reconstruction, X-ray phase contrast imaging, compressive optical imaging, Compton camera imaging, etc.