Shahrum Nedjati-Gilani

Awarded VEIV EngD

Diffusion MRI provides an insight into the microstructural architecture of tissue by observing the restricted and hindered displacement of water molecules undergoing Brownian motion. Diffusion-Tensor MRI (DT-MRI) is the most common diffusion MRI technique and is often used for mapping fibre orientations. However, DT-MRI is only capable of recovering a single fibre orientation in each voxel, and cannot resolve the orientations of crossing-fibres. A new generation of reconstruction algorithms such as PAS-MRI, q-ball and spherical deconvolution has appeared that can resolve crossing fibre orientations; however, these more complex algorithms are generally less successful in correctly identifying the orientation of single fibres and are more computationally demanding than DT-MRI. Therefore, it would be useful to know the expected number of fibre populations present in each voxel in order to choose a suitable reconstruction algorithm. The aim of the project is to find and evaluate methods for mapping the number of fibre orientations in each voxel of a 3D diffusion MRI acquisition in a fast and accurate manner.


Primary Supervisor: Daniel Alexander

Industry Sponsor: Philips Medical Systems

Research Area: Medical: Diffusion MRI