Modeling, Inverse Problems and Machine Learning in Cryogenic Microscopy using Three-Dimensional Tomography

Cryo-electron tomography (cryo-ET) is an imaging technique that visualizes biological molecules and intracellular structures in their original 3D environment at nanometer resolution. Recovering the volume from observations faces a number of problems, such as high noise, dimension of the data, ill-possedness of the inverse problem or complex physical phenomenon to be modeled. In this thesis, we will develop reconstruction algorithms to improve the resolution of cryoET volumes. 

More info at: https://sites.google.com/view/debarnot/open-positions

The thesis will be carried out at the CREATIS laboratory, a recognized multidisciplinary laboratory with extensive expertise in medical imaging, at INSA Lyon. The diversity of researchers present in the laboratory (skills in physics, mathematics, computer science, etc.) makes it an ideal environment for an interdisciplinary thesis, where the tools developed can find applications in di!erent imaging modalities.

The successful candidate will develop skills in signal processing, machine learning, computer science, and applied mathematics. An understanding of the physics of the acquisition system and of the biological issues addressed by cryo-tomography will be important, yet no previous knowledge in physics or biology is required.