University of Twente Student Theses

Login

A novel approach to quantify and analyse brain imaging features on MRI

Coes, M.G. (2017) A novel approach to quantify and analyse brain imaging features on MRI.

[img]
Preview
PDF
3MB
Abstract:Atherosclerosis is a progressive inflammatory artery disease responsible for about 50% of deaths in the western world, mainly due to heart disease and stroke. Brain abnormalities that can be seen in patients with arterial disease (atherosclerosis) are heterogeneous and are the result of different underlying etiologies. The three main groups of brain abnormality etiologies are neurodegenerative disease, large vessel disease and small vessel disease. The main neuroimaging feature of neurodegenerative disease are localized (hippocampal, temporal, frontal and parietal) and global atrophy while large vessel disease (LVD) leads to cortical infarcts. Neuroimaging features of cerebral small vessel disease (CSVD) include: recent small subcortical infarcts, lacunar infarcts, white matter hyperintensities (WMH), dilated perivascular spaces, cerebral microbleeds and possibly even brain atrophy. CSVD results from a complex mix of genetic and cardiovascular risk factors. There are several types of CSVD based on etiology (most common are arteriolosclerosis, cerebral amyloid angiopathy and genetic SVD). These different types of CSVD may lead to different manifestations of imaging features of CSVD. In this thesis we investigated two novel approaches to the analysis of imaging features in patients with arterial disease. In the first chapter, we hypothesized that the shape of white matter hyperintensities (WMH) may be indicative of the underlying etiology. We assessed possible shape descriptors to study the shape of WMH lesions and to provide additional information of the WMH. We identified eccentricity, fractal dimension, convexity and solidity as plausible shape descriptors. In the second chapter, we hypothesized that different underlying etiologies may lead to different imaging manifestations. We used cluster analysis to combine magnetic resonance imaging (MRI) features to identify different imaging phenotypes. We identified three distinct groups with different WMH severities and cardiovascular risk factors. These three groups were divided further revealing seven distinct brain imaging patterns. Three of these patterns are associated with localized arterial disease, two patterns may be caused by different types of SVD and one pattern shows signs of LVD. The final group does not show signs of LVD or SVD and might have a neurodegenerative etiology.
Item Type:Essay (Master)
Faculty:TNW: Science and Technology
Subject:44 medicine
Programme:Technical Medicine MSc (60033)
Link to this item:http://purl.utwente.nl/essays/73349
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page