David Maxen

I work as part of the Oxford Molecular Diagnostics Centre, where I research deep learning approaches for early multi-cancer detection. The main idea is that cancerous cells have unique epigenetic changes, and that when these cells die (e.g. via apoptosis or necrosis) small fragments of their DNA enter the bloodstream. Hence by taking a blood sample, there is a chance that we will be able to detect these fragments, and diagnose cancer at an early stage. Furthermore, we can also try to predict where the tumour is in the body. This technique is an example of a liquid biopsy, and there have been very promising clinical trials to test these in the NHS, for example the NHS Galleri trial and the SYMPLIFY study run by Oxford Cancer. One key area for improvement for these tests is to lower the false positive rate – we don’t want to misdiagnose someone with cancer. Under the supervisor of Dr Dimitris Vavoulis and Prof. Anna Schuh I am working on improving these tests through the use of deep learning techniques.

How could your research ultimately benefit patients?

Being able to diagnose multiple cancers at an earlier stage with a cheap, non-invasive techniques could allow better triage of patients, help facilitate the development of new therapeutics which can target early stage cancers, and potentially allow for early stage treatment.

About David 

I have a BSc and MASt in Mathematics from the University of Warwick, where I focused on probability, statistics, combinatorics and data science. I also undertook a research internship at the University of Lancaster on their STOR-i programme, where I was investigating how to reconstruct evolutionary (phylogenetic) trees from DNA sequences. This was a great introduction to bioinformatics and helped me decide that I wanted to study for a PhD in a similar but more healthcare focused area.

Outside of my research, I usually spend my time teaching, rowing, playing squash and beekeeping.