University of Leeds Featured PhD Programmes
University of Kent Featured PhD Programmes
John Innes Centre Featured PhD Programmes
Catalysis Hub Featured PhD Programmes
University of Reading Featured PhD Programmes

MRC DTP 4 Year PhD Programme: Use of Machine Learning and Computer Vision to detect Cerebral Microbleeds in SWI MRI

This project is no longer listed on and may not be available.

Click here to search for PhD studentship opportunities
  • Full or part time
    Dr J Zhang
    Dr A Doney
    Prof S McKenna
  • Application Deadline
    No more applications being accepted
  • Competition Funded PhD Project (European/UK Students Only)
    Competition Funded PhD Project (European/UK Students Only)

About This PhD Project

Project Description

Small areas of bleeding in the brain, known as cerebral microbleeds (CMB), are emerging as important features of an aging brain. Not only are they a marker for unhealthy blood vessels associated with development of dementia, but they also indicate an increased risk of major bleeding in the brain. This is particularly a concern in the common situation where doctors need to use medicines that stop clots from forming, and therefore increase risk of bleeding, to prevent heart attacks and ischaemic strokes. Although CMB are common they are not checked for routinely in conventional commonly used brain scan techniques and can only be detected with a specialised modality of Magnetic Resonance Imaging (MRI) of the brain known as Susceptibility Weighted Imaging (SWI).

In this project we would like to develop automated image processing techniques to count the number and location of CMB’s in the brain from MRI SWI images. This project is part of a larger programme of ground-breaking work being conducted in a partnership between the Computer Vision and Image Processing Group and clinicians at Ninewells Hospital exploiting advanced image processing techniques to improve risk assessment, the diagnosis and stratification of dementias. Existing work on this topic includes the use of mixtures of Gaussians [2], deep learning [3], and other machine learning classifiers [1]. One starting point would be to investigate the use of a 3D ’blob’ detector [4] to automatically generate candidate CMB regions, followed by a machine learning classifier to detect the true positives.

Datasets and annotations are already in place. This is an excellent chance for a PhD candidate to develop AI algorithms for automated microbleed detection.

[1] Fazlollahi et al. “Efficient machine learning framework for computer-aided detection of cerebral microbleeds using the Radon transform" ISBI 2014
[2] Seghier et al. “Microbleed Detection Using Automated Segmentation (MIDAS): A New Method Applicable to Standard Clinical MR Images”, PLOS ONE, 2011
[3] Q. Dou et al., "Automatic Detection of Cerebral Microbleeds From MR Images via 3D Convolutional Neural Networks," in IEEE Transactions on Medical Imaging, vol. 35, no. 5, pp. 1182-1195, May 2016.

FindAPhD. Copyright 2005-2019
All rights reserved.