Master's Thesis

Application of diffusion MRI to investigate structural brain differences between Mild Cognitive Impaired and Cognitively Unimpaired subjects

Sofia Silvestre de Oliveira — 2026

Key information

Authors:

Sofia Silvestre de Oliveira (Sofia Silvestre de Oliveira)

Supervisors:

Sandra Raquel Ramos Tecelão; Rita Homem de Gouveia Costanzo Nunes (Rita Homem de Gouveia Costanzo Nunes)

Published in

July 6, 2026

Abstract

Alzheimer’s disease (AD) begins years before dementia is clinically apparent, yet imaging biomarkers that detect tissue injury and predict cognitive decline remain elusive. Diffusion MRI is sensitive to microstructural change, but conventional diffusion tensor imaging (DTI) metrics are confounded by extracellular free water (FW), particularly in atrophic regions. Prior FW-corrected work has relied on single-shell acquisitions, for which the FW model is ill-posed, and has focused on white matter, leaving the extent, specificity, and prognostic value of FW abnormalities unclear. We applied multi-shell FW-DTI in 146 biomarker-stratified participants (CU−, n=47; CU+, n=23; MCI−, n=19; MCI+, n=57), analysing white matter, subcortical, and cortical grey matter with linear mixed-effects models, and assessing prognosis with Cox regression and Kaplan–Meier analyses. Conventional DTI showed apparent abnormalities, but these were driven by free water: once the FW compartment was modelled, the corrected diffusivities diverged from—and in limbic regions reversed—the standard estimates. The abnormalities were carried by elevated FW and were largely confined to MCI+, remaining minimal in both CU+ and MCI− individuals, indicating specificity to amyloid-positive disease rather than MCI alone. The strongest white matter effects occurred in the fornix, and the largest FW increases in the hippocampus and caudate. Baseline hippocampal FW predicted subsequent cognitive progression independently of amyloid status and APOE ε4 (HR = 2.18 per standard deviation, 95% CI 1.09–4.37; subcortical model C-index = 0.832). Multi-shell FW-DTI may capture AD-related microstructural change across tissue compartments and provide prognostic information beyond established molecular biomarkers.

Publication details

Authors in the community:

Supervisors of this institution:

Fields of Science and Technology (FOS)

industrial-biotechnology - Industrial Biotechnology

Publication language (ISO code)

eng - English

Rights type:

Embargoed access

Date available:

May 27, 2027

Institution name

Instituto Superior Técnico