PhD defense of Samuel St-Jean

On Thursday 9 January, Samuel St-Jean will defend his PhD thesis entitled: “Restoration and characterization of diffusion MRI data.”

Abstract

Diffusion magnetic resonance imaging (dMRI) is a non-invasive technique used to investigate microstructural properties of tissues. However, the properties of the measured dMRI signal and its quality depend on the scanner hardware used to acquire the data. One important trade-off arises between the spatial resolution, the image quality and the acquisition time; a higher spatial resolution requires longer acquisition time due to the increased image size, while also suffering from a decrease in the image quality due to a lower measured signal. This complicates the comparability of datasets acquired in different centers if these subtle differences are not taken into account. To facilitate the analysis of dMRI datasets, this thesis presents new algorithms and techniques which can be applied to any already acquired datasets and enhance their quality.

The new ideas presented therein include a new algorithm exploiting spatial and angular redundancy of dMRI datasets to counterbalance the loss of image quality caused by increasing the spatial resolution. A realignment algorithm is also introduced to ensure accurate spatial correspondence of major fiber pathways between subjects before their analysis. An automated algorithm to estimate the scanner specific noise distribution to analyze datasets from three different centers is also developed. The last chapter presents a new algorithm to facilitate comparison of datasets acquired with different scanning protocols or from multiple centers. These new developments bring us one step closer to understanding the inner workings of the brain by enhancing the dMRI analysis pipeline.