Lay summary

The anisotropy of diffusion in white matter can be exploited for mapping the structural neuronal connectivity of the brain, and structures invisible with other imaging modalities can be highlighted. The study of this connectivity is of course of major importance in a fundamental neuroscience perspective, for developing our understanding of the brain, but also in a clinical perspective, with particular applications for the understanding of stroke, schizophrenia, or Parkinson’s disease. As a consequence, our ability to achieve high angular resolution diffusion magnetic resonance (MR) imaging represents an important challenge for neuroscience.

The state-of-the-art diffusion spectrum imaging modality, which relies on Cartesian signal sampling, is known to provide good imaging quality but is significantly too time-consuming to be of real interest in a clinical perspective. Accelerated acquisitions, relying on a smaller number of sampling points, are thus required.

The primary aim of this project is to define advanced acquisition strategies for accelerated high angular resolution diffusion MR imaging. Since the signal of interest, identifying the fiber directions, lives on the sphere, we will consider multiple spherical shell sampling rather then Cartesian sampling, as suggested by recent approaches. The originality of our approach resides in the fact the our sampling strategies will be driven by the conditions of a new sampling theorem on the sphere, with the aim of avoiding as much as possible interpolation and aliasing issues that may hamper the imaging quality.

At the reconstruction level, in each voxel of the brain, the imaging problem for diffusion will be formulated in terms of denoising, deconvolution or even compressive sampling problems for the recovery of a sparse signal on the sphere, where the sparsity stems from the small number of fiber directions of interest.

The main applicant of this project has recognized expertise in signal processing on the sphere, sparsity applications, and compressive sampling, while the other applicants have an extensive expertise in diffusion MR imaging and its applications to structural connectivity analysis. Their collaboration therefore sets ideal grounds for the expected breakthroughs towards accelerated high angular resolution diffusion MR imaging.