SPARKLING: variable-density k-space filling curves for accelerated T2*-weighted MRI
Find out more about subscribing to add all events.
SPARKLING: variable-density k-space filling curves for accelerated T2*-weighted MRI
Tue, 27/11/2018 - 10:00
Location:
Speaker:
Philippe Ciuciu
Affiliation:
CEA/NeuroSpin & INRIA Saclay /Parietal
Synopsis:
Magnetic Resonance Imaging (MRI) is one of the most powerful and safest imaging techniques for examining the human body. High-resolution MRI is expected to aid in the understanding and diagnosis of many neurodegenerative pathologies involving submillimetric morphological alterations such as Alzheimer’s disease. Although ultra-high-magnetic systems can deliver a sufficient signal-to-noise ratio to increase spatial resolution, long scan times and motion sensitivity continue to impede the exploitation of HR-MRI.
In the last decade, the newly developed theory of compressed sensing (CS) offered a promising solution for reducing the MRI scan time. The application of CS to MRI commonly relies on simple sampling patterns in k-space, such as straight lines, spirals or slight variations of these elementary shapes, which do not take full advantage of the degrees of freedom offered by the hardware. In this talk, I will present a new method called SPARKLING that may overcome these limitations. SPARKLING stands for Spreading Projection Algorithm for Rapid K-space sampLING. It is a versatile optimization-driven technique that generates non-Cartesian sampling trajectories, which comply with key criteria for optimal sampling and are compatible with MR hardware constraints.
Combining sampling efficiency with compressed sensing and parallel imaging, I will show that SPARKLING patterns allowed us to achieve 20-fold reduction in MR scan compared to fully-sampled Cartesian acquisitions for T2*-weighted 2D in vivo human brain imaging at 7 Tesla. The proposed approach was also extended to 3D imaging for which preliminary ex-vivo baboon brain results at 0.6 mm isotropic resolution suggested the possibility to reach up to 70-fold acceleration (45s vs 1hour) for T2* and susceptibility weighted imaging.
Biography:
Philippe Ciuciu is Director of Research and head of the Compressed sensing group in the INRIA/CEA Parietal team at NeuroSpin (CEA Saclay, France). Since 2017, he has been leading the transversal research program at NeuroSpin entitled “Toward 500-micron functional MRI in the human brain”. Prior to that, Dr Ciuciu received the Eng. degree in Electrical Engineering from ESIEA Paris, France, and the M.Sc. degree in Signal Processing and Automatic Control from the University of Paris-Sud, both in 1996. Then, he got his PhD in signal processing from the same university in 2000 and his Habilitation degree (HDR), the highest academic diploma in France, from the same university in 2008. In 2012-13, Dr Ciuciu was visiting Professor at University of Toulouse in the applied maths department. His research interests range from the development of new compressed sensing solutions to speed up Magnetic Resonance Image (MRI) acquisition and improve image reconstructions to applications in cognitive neuroscience for understanding the neural basis of multiperceptual learning from complementary MEG and functional MRI (fMRI) data. He is also scientific advisor for pharmaceutical groups in the context of MRI studies for clinical trials. In this context, he is also developing blind deconvolution solutions of the pharmacologically-modulated fMRI signal. Dr Ciuciu has coauthored 45 peer-reviewed journal articles, 2 MRI-related patents licensed in the US, Japan and China, 4 book chapters and more than 120 pee-reviewed conference communications and abstracts. Dr Ciuciu is member of the ISMRM and EURASIP societies and IEEE senior member, currently involved in the BioImaging Signal Processing technical committee of the ISBI conference as well as vice chair of the Bio-Imaging Signal Analaytics EURASIP technical committee.