Previous experiments (Kress 2014) have suggested that brain features known as perivascular spaces play an important role in the clearance of metabolic waste from the brain tissue. They extend along arterioles, capillaries and venules, communicating freely with perineuronal and other spaces between glial cells (i.e. the “brain cleaners”) and fiber tracts, and contain cerebrospinal fluid. Perivascular spaces (PVS), when enlarged, can be seen in MRI and have attracted the attention of the clinical community as they have been found associated with ageing, vascular risk factors and with a myriad of inflammatory and neurodegenerative diseases (Francis et al. 2019). However, the interaction between PVS and venules and the haemodynamic characteristics of these clearance processes are not known, making difficult (if not impossible) the prognosis and contributing to inefficient treatment strategies.
We have developed computational methods to segment venules, perivascular spaces, main venous drainage pathways and cerebrospinal fluid-filled spaces within the intracranial volume. However, several factors hamper their accuracy and limit their applicability, having these been identified as: motion artefacts, poor tissue-structure contrast, biological tissue inhomogeneities, presence of lesions and features with similar intensity and size characteristics, scanning protocol variations and the lack of ground truth. Therefore, a robust unsupervised or semi-supervised approach to segment and characterise the venules and venous pathways consistently and accurately for multicentre studies and personalised medicine is needed.
This project aims to develop a robust state-of-the-art computational approach to segment, model and characterise venules and components of the brain drainage pathways and investigate their association with PVS, lesion progression, and brain tissue loss in patients with small vessel and Alzheimer’s diseases.
We hypothesise that using state-of-the-art machine learning techniques it will be possible to overcome the limitations of the current methods that segment these venous pathways. Also, that it will be possible to estimate the inflammatory vs. vascular contributions in the pathological features seen in brain MRI if we use an integrated approach that combines the results from the detailed segmentation of venules, perivascular spaces, brain lesions, and main venous drainage pathways and cerebrospinal fluid-filled spaces within the intracranial volume, with textural tissue characteristics, retinal vasculature measurements and relevant clinical information.
This project will use data from well-characterised mild stroke patients with long term outcomes, and Alzheimer’s disease patients from a publicly available database (http://adni.loni.usc.edu/
) from where the necessary data and priors of the segmentations are available. Tissue properties including mineral deposition and extraction of the diffusion characteristics in tissues will be used. The methods to apply involve variants of Hessian filters, combined with the machine learning methods we have evaluated in the past: UResNet, Generative Adversarial Networks, and with the descriptors we have explored and developed in the last five years (see publications of principal supervisor in https://www.research.ed.ac.uk/portal/en/persons/maria-valdes-hernandez(f22f22d9-52bb-4883-bf94-52aa23a691e1).html
). All data necessary for this project is already available and has been generated as part of ongoing projects.
This MRC programme is joint between the Universities of Edinburgh and Glasgow. You will be registered at the host institution of the primary supervisor detailed in your project selection.
All applications should be made via the University of Edinburgh, irrespective of project location. For those applying to a University of Glasgow project, your application along with any supporting documents will be shared with University of Glasgow. http://www.ed.ac.uk/studying/postgraduate/degrees/index.php?r=site/view&id=919
Please note, you must apply to one of the projects and you must contact the primary supervisor prior to making your application. Additional information on the application process is available from the link above.
For more information about Precision Medicine visit: http://www.ed.ac.uk/usher/precision-medicine
Kress BT, Iliff JJ, Xia M, Wang M, Wei HS, Zeppenfeld D, Xie L, Kang H, Xu Q, Liew JA, Plog BA, Ding F, Deane R, Nedergaard M.Impairment of perivascular clearance pathways in the aging brain. Ann Neurol. 2014 Dec;76(6):845-61.
Francis F, Ballerini L, Wardlaw JM. Perivascular spaces and their associations with risk factors, clinical disorders and neuroimaging features: A systematic review and meta-analysis. Int J Stroke 2019; 14(4): 359-371.
Muñoz Maniega et al. Spatial Gradient of Microstructural Changes in Normal-Appearing White Matter in Tracts Affected by White Matter Hyperintensities in Older Age. Front Neurol 2019 https://doi.org/10.3389/fneur.2019.00784
Ballerini et al. Perivascular Spaces Segmentation in Brain MRI Using Optimal 3D Filtering. Scientific Reports (2018) 8:2132