I am a Machine Learning Research Scientist at Oxford Nanopore Technologies, where my primary focus is on enhancing the speed, affordability, and accuracy of genetic data analysis. Prior to this role, I completed a DPhil/PhD in computer vision within the Visual Geometry Group at the University of Oxford as part of AIMS CDT, under the guidance of Professor Andrew Zisserman and Dr. Timor Kadir. My research predominantely focused on leveraging modern advances in deep learning to analyse medical images and I maintain collaborative ties with the research group.
I am particularly interested in the application of machine learning to real-world problems in science and healthcare. During my PhD, I developed SpineNet, a source-available tool for analysing spinal MRI scans for a range of common spinal disorders, with comparative accuracy to expert radiologists. This has featured in several large scale studies into back pain and has also been adopted by major pharmaceutical companies to support clinical trials. I am also interested in data-efficient learning and novel forms of supervison, such as learning about radiological images from associated free-text reports . More information is available in my CV here.
Outside of my research, I am into sports, both real (running, touch rugby) and debatably-real (chess).
Applying for a PhD? If you're thinking of applying for a PhD, and you have questions regarding Oxford/AIMS/VGG/Computer Vision research in general, please don't hesitate to get in contact even if you feel like your background isn't suitable (I definitely felt that way when I applied). I'll try to respond as soon as possible although it may take a day or two.
05/04/2023: Vision-Language Modelling For Radiological Imaging and Reports In The Low Data Regime" has been awarded an Oral presentation at MIDL2023!
19/09/2022: I am attending MICCAI 2022 in Singapore, please get in touch if you'd like to talk! Poster: M007
07/08/2022: I am attending Oxford Machine Learning Summer School (healthcare track).
25/07/2022: SpineNet's source code is now publicly available. See the repo here
16/06/2022: "Context-Aware Transformers For Spinal Cancer Detection and Radiological Grading" has been accepted to MICCAI 2022. Preprint Link
05/05/2022: New paper provisionally accepted to MICCAI 2022! More News Soon
03/05/2022: SpineNetV2 Technical Paper Released. See the preprint here
23/10/2021: "3D Spinal Column Segmentation with Single Plane 2D-Projected Annotations" accepted to MedNeurips 2021!
17/09/2021: Presented 'Self-Supervised Multi-Modal Alignment for Whole Body Medical Imaging' at MICCAI 2021. See the talk on youtube here and the project page.
16/08/2021: Released DICOMcat, an simple pip-installable, open-source tool for quickly viewing DICOM files in the terminal. View on Github.
14/07/2021: Our new paper, 'Self-Supervised Multi-Modal Alignment for Whole Body Medical Imaging' has been accepted to MICCAI 2021! See arXiv (code coming soon)
22/04/2021: SpineNet Version 2 is now live! This software takes clinical spinal MRI scans as input and performs automatic vertebrae detection and labelling as well as radiological grading for a range of spinal diseases with accuracy comparable to clinical radiologists. Have a look at the online demo here.
06/07/2020: Our paper, 'A Convolutional Approach to Vertebrae Detection and Labelling in Whole Spine MRI ' was accepted to MICCAI 2020! See the arxiv version here
28/04/2020: Some work by our collaboration on automated scoliosis detection has been covered on the Department of Engineering Science website here.
12/04/2020: I've redesigned my website
03/04/2020: The first paper of my PhD, "The Ladder Algorithm: Finding Repetitive Structures in Medical Images by Induction" has been published at ISBI 2020. Have a look on arxiv .
A quick post detailing how I made this new website.