2018 DRF Impulsion Days
Upcoming -> COSMIC presentation@DRF impulsion days (2nd year, Oct, 2 2018)
After two years of existence, the scientific achievements of the COSMIC project will be presented to the whole DRF community.
2nd year COSMIC evaluation
COSMIC evaluation (Sep, 11 2018)
After a 2-year life, the COSMIC project has been very well evaluated and will continue for at least 3 years. (The slides presented during the defense are available here).
Welcome to Dr Abderrahim Halimi (April, 20th 2018)
Dr Abderrahim Halimi, now currently Assistant Professor at Heriot-Watt University (Edinburgh, Scotland), is visting us for 6 weeks at NeuroSpin (April, 20- June, 2) to develop a collaboration on inverse problem solving in medical imaging (MRI, fMRI).
2017 DRF Impulsion Days
COSMIC presentation@DRF impulsion days (1st year, Oct 2017)
After one year of existence, the scientific achievements of the COSMIC project have been presented to the whole DRF community (The slides of my talk are available here).
Congrats to Sam!
On the 1st of Dec. 2017, Samuel Farrens has been hired on a permanent CEA research scientist position with part time dedicated to the COSMIC project. His main affiliation is CosmoStat and Sam will spend time with us at NeuroSpin on a weekly basis. Congrats Sam!
Invitation to the DEDALE Workshop (Sep., 2017)
We are glad to attend the Dedale (EU FET-OPEN project led by J.-L. Starck) meeting (Sep. 4-6 in Nice) and present some recent work on compressed sensing for Magnetic Resonance Imaging, both at the acquisition and reconstruction levels.
Invitation to the Special Session "Compressed sensing and inversion" at the 2017 Gretsi conference
We have been invited to contribute to the special session led by Jérôme Idier (IRCCyN/CNRS) and Cédric Herzet (INRIA Rennes) on the Compressed Sensing topic. We will highlight the recent results we obtained at 7 Tesla in CS-MRI based on the novel SPARKLING (Segmented Projection Algorithm for Random K-space samplING say shortly: SPARse K-space samplING) trajectories for ex-vivo T2* imaging.
Dynamic MRI image reconstruction using adaptive regularization methods
Abstract: Dynamic MRI image reconstruction is an inherently under-determined inverse problem because the object is changing as the data is collected. (There is no such thing as ``fully sampled data'' in dynamic MRI). The ill-posed nature of dynamic MRI requires some form of regularization (signal models) to distinguish among candidate solutions. Traditional k-space data sharing methods (like keyhole imaging) use implicit signal models, whereas modern regularized methods use explicit signal models. Typical regularization methods are based on simple mathematic signal models such as wavelets. This talk will focus on newer methods that are adaptive, where the signal model is learned either from training data or concurrently with reconstruction of the dynamic image sequence.
This is joint work with Sai Ravishankar.