This project has been fundedin 2016 by the CEA Fundamental Research Division (DRF) for 2 years to foster collaborations between CEA research scientists and engineers working on biomedical imaging (Magnetic Resonance Imaging or MRI) at NeuroSpin (Team leader: Philippe Ciuciu) and on astrophysics at the CosmotStat Lab (Team leader: Jean-Luc Starck) within the Astrophysics Department) on Compressed Sensing (or Compressive Sampling, see here for details about CS).
On the one hand, the NeuroSpin team has reached a scientific expertise in Compressed Sensing for MRI, especially in the design of optimal sparse sampling (or undersampling) schemes of the Fourier domain (also termed k-space in the MRI community). Because MRI data are collected sequentially along piecewise continuous or smoother k-space trajectories, the critical aspect lies in the ability of yielding trajectories that mimic a prescribed density, that fulfill hardware constraints and that cover the whole k-space sufficiently fast. This work is also the result of a fruitful collaboration with mathematicians from CNRS in Toulouse, noticeably Pierre Weiss (ITAV) & Jonas Kahn (IMT).
On the other hand, the CosmoStat team is focused on computational cosmology. They develop and apply new methods from statistics, signal processing, and compressed sensing to cosmology and other fields. The main CosmoStat research areas are:
- Statistics & Signal Processing: Develop new methods for analyzing astronomical data, and especially in cosmology (PLANCK, Euclid, etc.) where the needs of powerful statistical methods are very important.
- Cosmology: Analyze and interpret data. This includes CMB, weak lensing, and other fields such as galaxy clustering and BAOs.