Welcome to my academic website!

I am a PhD student in Machine Learning at University College London, where I research robust deep learning methods for medical imaging under the supervision of Prof. John Ashburner and Prof. Cathy J. Price. My work focuses on developing machine learning models that can generalise across diverse real-world settings, with particular application to stroke segmentation in clinical MRI and CT data.

Current Work

Currently, I work as a Founding Computer Vision Scientist at Prospectral, where I lead machine learning and AI integration and research. I’m part of the founding team building cutting-edge AI for spectral imaging applications.

Previously, I was a Computer Vision Researcher at Tractive, an A16Z-backed startup focused on 3D generative AI. My background includes research experience as an intern at Schlumberger Cambridge Research and undergraduate studies in Chemical Physics at the University of Bristol.

Research Interests

Research Focus

My work aims to develop robust and generalisable machine learning models that can operate reliably across diverse real-world settings. Much of my research focuses on medical image analysis, particularly brain pathology segmentation, with applications in clinical diagnosis and treatment planning.

My PhD research has focused on developing novel approaches to domain generalisation and synthetic data generation for 3D medical image segmentation. This includes creating convolutional attention models, implementing self-supervised learning techniques, and developing physically-constrained synthetic data frameworks that enhance the performance and reliability of deep learning models in real-world clinical scenarios. My work bridges the gap between theoretical machine learning advances and practical clinical deployment.

Recent Publications

My latest publications include “Synthetic Data for Robust Stroke Segmentation” published in the Journal of Machine Learning for Biomedical Imaging (MELBA), and “DeepISLES: a clinically validated ischemic stroke segmentation model from the ISLES’22 challenge” published in Nature Communications. I also have two new papers at MICCAI 2025 on physics-constrained synthetic data generation and sequence-invariant contrastive learning.

Feel free to explore my publications and CV to learn more about my academic journey and research contributions.