Use this URL to cite or link to this record in EThOS:
Title: Parcellation of the human cerebral cortex using diffusion MRI
Author: Ganepola, Tharindu
ISNI:       0000 0004 7230 4396
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Histological methods have long been used to segment the cerebral cortex into structurally distinct cortical areas that have served as a basis for research into brain structure and function and remain in use today. There is great interest in adapting and extending these methods to be able to use non-invasive imaging, so that tighter structure-function relationships can be measured in living subjects. Whilst diffusion neuroimaging methods have been widely applied to white matter, the reduced anisotropy in the thin, complexly folded grey matter of the cortex has so far limited its study. In vivo parcellation pipelines have instead focussed on T1 and T2 weighted MRI. Recent advances in imaging hardware have reignited interest in grey matter diffusion MRI as a viable candidate for characterising architectonic domains. This Thesis explores the capabilities of dMRI as a measure of cortical microstructure using in vivo datasets from healthy adult participants. A cortical parcellation pipeline was developed in which both unsupervised and supervised algorithms were explored. Results were presented at both the group level and single subject level across the entire cortical sheet. The diffusion-based feature space characterised the known variation in cellular composition and fibre density relative to the local cortical surface normal. Thus they remain invariant to the confounding orientation changes associated with cortical folding, which usually inhibit studies of cortical microstructure. The features were compared to the alternative T1w/T2w myelin mapping methods to demonstrate that the diffusion MRI signal provides a complementary mode of contrast. A series of classification experiments were used to determine the most effective methods for utilising diffusion in grey matter applications. Several additional methods from the dMRI literature were compared to highlight the benefit of higher-order tissue representations. Similarly, classification tasks were used to corroborate the benefits of sampling multiple b-values in cortical studies. The experimental chapters provide strong evidence in favour of the future use of diffusion MRI as a measure of the varying microstructure that defines cortical areas.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available