Prospective DPhil applicants
I am always looking for motivated DPhil/PhD or Masters students to join my group.
I am happy to be named as a potential supervisor for applications in areas relevant to my group's research. There is no need to contact me simply to ask whether I am accepting DPhil students or whether you may name me as a potential supervisor. Please check that you meet the entry requirements and read the information already available online.
If you have a specific research-related question about the fit of your proposed research with my group's work, you are welcome to get in touch informally. Your message should explain briefly your relevant research background, the specific research question or direction you would like to pursue, and how this relates to our recent or ongoing work. Please include the word DPhil in the subject line.
A viable research proposal also needs appropriate data. Our main clinical collaborators specialise in orthopaedics, which is where we have access to much of our clinical data. We also study publicly available datasets, specified in our publications.
Proposals in other clinical fields are gladly received, but please identify the datasets you expect to use, or explain a realistic route by which suitable data could be made available. If proposing a clinical problem, please also consider what clinical expertise would be required and where this would come from.
Applicants coming from a different research area are welcome, but should explain how their existing technical background prepares them to work in medical image analysis and how the proposed research connects with the group's work.
Research areas
- Segmentation from 2D images or 3D volumes
- Small-structure localisation from weak annotations (scribbles, dots)
- 3D reconstruction of anatomical structures
- Anatomical landmark detection
- Validation and evaluation measures, particularly in the presence of multiple annotators
- Fairness and bias detection in healthcare
- Failure detection in image classification
- Multimodality data integration for more accurate diagnosis
- Trust and explainability in clinical applications