Search

Cerebral vasculature plays an important role in maintaining brain function and homeostasis. In ageing, a number of genetic and environmental risk factors can compromise brain vascular health and contribute to multiple neurological disorders. White matter (WM) is more vulnerable in cerebrovascular ageing compared with grey matter. With the development of automated and reproducible neuroimaging techniques, there is an opportunity to detect many ageing-related or diseases-related WM changes at an early stage with greater sensitivity. The overall objective of this PhD project is, therefore, to develop reliable neuroimaging biomarkers for characterising WM integrity and examine their relationships with cerebrovascular burden. Four separate studies were carried out to investigate this topic in depth. In the first study, a new diffusion-weighted imaging (DWI) measure, Difference in Distribution Functions (DDF), was developed to overcome the limitations of existing DWI measures in characterising the white matter (WM) microstructural integrity. DDF showed a stronger correlation with age and cognition than other DWI measures investigated. The second study extended the first study by employing longitudinal datasets to examine the effects of various risk factors on WM microstructural integrity over time. Results showed that older age is the primary risk factor for decline in WM integrity in a healthy general population sample. In the third study, a new imaging metric, WM brain age, was introduced to comprehensively assess the health of WM and cerebrovascular disease burdens for each individual. WM brain age was calculated by using a three-dimensional convolutional neural network (3D-CNN) deep learning model. Findings showed that WM brain age is sensitive to most vascular risk factors and cognitive domains related to vascular dysfunction. In the fourth study, the impacts of high blood pressure on grey matter (GM) and WM, respectively, were examined using 3D-CNN to construct GM and WM ages. Findings revealed that hypertension is associated with both GM and WM impairment, and that WM integrity is more vulnerable to hypertension. In conclusion, by developing novel neuroimaging biomarkers for the changes in cerebrovascular ageing, this thesis provides novel contributions to the existing literature. These findings have the potential to better quantify the vascular burden in future research work and in clinical practice.

This thesis focuses on the development and applications of magnetic resonance electrical properties tomography (MREPT), which is an emerging imaging modality to noninvasively obtain the electrical properties of tissues, such as conductivity and permittivity. Chapter 2 describes the general information about human research ethics, MRI scanner, MR sequence and the method of phase-based MREPT implemented in this thesis. Chapter 3 examines the repeatability of phase-based MREPT in the brain conductivity measurement using balanced fast field echo (bFFE) and turbo spin echo (TSE) sequences, and investigate the effects of compressed SENSE, whole-head B_1 shimming and video watching during scan on the measurement precision. Chapter 4 investigates the conductivity signal in response to short-duration visual stimulus, compares the signal and functional activation pathway with that of BOLD, and tests the consistency of functional conductivity imaging (funCI) with visual stimulation across participants. Chapter 5 extends the use of functional conductivity imaging to somatosensory stimulation and trigeminal nerve stimulation to evaluate the consistency of functional conductivity activation across different types of stimuli. In addition, visual adaptation experiment is performed to test if the repetition suppression effect can be observed using funCI. Chapter 6 explores if resting state conductivity networks can be reliably constructed using resting state funCI, evaluates the consistency of persistent homology architectures, and compares the links between nodes in the whole brain. Chapter 7 investigates the feasibility of prostate conductivity imaging using MREPT, and distinctive features in the conductivity distribution between healthy participants and participants with suspected abnormalities.

Genetic, epidemiological and post-mortem studies implicate inflammation in the pathogenesis of schizophrenia. While inflammation can regulate neurogenesis, and is associated with brain structural and functional abnormalities, research addressing the link between inflammation and neurobiology in schizophrenia is limited. In the subependymal zone (SEZ), the brain’s largest niche for neurogenesis, I discovered vast differences across the whole transcriptome in schizophrenia compared to controls (post-mortem cohort n=20-21/group). CD163, a macrophage marker, had significantly increased expression in schizophrenia. Overall, genes in immune pathways were significantly upregulated, whereas genes in neurogenesis promoting pathways were significantly downregulated in schizophrenia along with reduced markers for all stages of neurogenesis. In a larger cohort, including bipolar disorder (n=29-32/group), macrophage cell density was increased specifically in schizophrenia. Cluster analysis of inflammation genes defined heightened inflammation subgroups comprising 37% of schizophrenia cases, 32% of bipolar disorder cases and only 10% of controls. The high inflammation schizophrenia subgroup had increased peripheral immune cell marker mRNAs, increased macrophage density, and altered markers for neurogenesis in the neurogenic niche suggesting increased stem cell quiescence and reduced neuronal progenitor cells and newborn neurons. These findings were replicated in an independent cohort (n=20-22). An analysis comparing the transcriptomes of high and low inflammation schizophrenia subgroups revealed that heightened inflammation alters the SEZ microenvironment by increasing angiogenesis and extracellular matrix stiffness, suggesting a fibrotic state. These changes likely increase blood-brain-barrier permeability and immune cell infiltration, and alter neuronal differentiation and migration. Allowing for translation potential of this research, I identified that peripheral inflammation marker, C-reactive protein (CRP), was increased in schizophrenia in a living cohort (n=1143). Increased CRP was associated with poorer attention ability and reduced cortical thickness in the caudal middle frontal, pars opercularis and posterior cingulate cortices in schizophrenia (MRI subset n=293). This thesis elucidated the molecular and cellular contribution of inflammation to aberrant neurogenesis in schizophrenia. Future manipulation of the identified cells and processes may enable the restoration of neurogenesis in schizophrenia and allow us to harness its potential for regenerative treatments. Further, inflammation biomarkers that have cognitive and brain structural relevance could identify schizophrenia patients who may benefit most from adjunctive treatment of inflammation.

Cerebrovascular disease (CVD) is a leading cause of morbidity and mortality, with manifold clinical consequences including vascular dementia (VaD), the second most common dementia subtype. CVD also worsens the risk and expression of Alzheimer’s disease (AD). Diagnosing VaD requires determining whether a patient’s cognitive deficits can be explained by the current CVD burden. However, CVD is markedly pleomorphic and its full extent has been difficult to determine. Traditional markers of CVD such as white matter hyperintensities (WMH) are inconsistently associated with clinical outcomes. There is an unmet need to better quantify total CVD burden and relate it to cognition and dementia. CVD indices have been published, which combine information from several neuroimaging markers, in the hope to better capture the variability seen on neuroimaging and lead to more robust associations with cognition. In this thesis, the existing literature on neuroimaging and neuropathological indices of CVD was systematically reviewed. The contributions of two common but under-researched MRI CVD markers, cerebral microbleeds (CMB) and dilated perivascular spaces (PVS), were then assessed. The second chapter examines the association of CMB with both cross sectional and longitudinal impairment, finding associations with executive function and visuospatial function respectively. There was no existing reliable rating scale for quantification of PVS and the development of this is the theme of the third chapter. The fourth chapter examines the associations of dilated PVS with longitudinal cognitive impairment and incident dementia. I found that individuals with the most severe PVS had declines in Global Cognition and their presence triples the risk of developing dementia over 8 years. The fifth chapter describes the development of an MRI-based composite CVD index and its validation in two independent cohorts. Peak skeletonised mean diffusivity, a DTI measure, and WMH volume contributed most to the variability seen in Global Cognition. The Index explained 9% of the proportion of the variance seen in the development sample and 5 and 13 % in the validation cohorts respectively. It performed better than the most widely used CVD index in all cohorts examined. This work broadens the range of vascular pathologies used in the determination of the total CVD burden. The resulting composite index, while needing further refinement, represents a promising step-forward in the assessment of such burden in an individual, which could potentially help clinicians and researchers in determining the vascular contributions to dementia.