Welcoming Our New Members in 2026

We are delighted to welcome new members to the AI for Digital Health (AI4DH) Lab as the new academic year begins. Our new DPhil students, collaborating researchers, and postdoctoral researchers bring diverse backgrounds and research interests, further strengthening the group’s work in AI for healthcare.

Joy Lai joins the group as a DPhil student and Clarendon Scholar, bringing a background in biomedical engineering from the University of Toronto. Her previous research focused on developing human-centred technologies to support people living with dementia and their caregivers. This includes integrating machine learning and behavioural anomaly detection into digital reminder systems, aiming to identify meaningful changes in daily routines, support independent living and reduce caregiver workload.

Zihan Zhang joins us as a DPhil student after completing a master’s degree at Peking University. Zihan’s research interests span AI for healthcare, recommender systems and data mining, with previous work on graph-based medication recommendation. This research explores how medical knowledge can be incorporated into recommendation models to improve their interpretability and support treatment decisions.

Luisa Kurth brings an interdisciplinary background in psychology and machine learning. She holds a bachelor’s degree in Psychology from the University of Mannheim and a master’s degree from the Oxford Internet Institute, and subsequently pursued further master’s studies at the University of Tübingen, focusing on deep learning, large language models and explainable AI. Her research interests include medical image analysis, individual variation and trustworthy AI. While primarily based in Oxford’s Parker Jones Neural Processing Lab, Luisa will work closely with AI4DH as a collaborating DPhil student.

Xu Yang joins AI4DH as a postdoctoral researcher. He holds a PhD from City University of Hong Kong and a bachelor’s degree from Sichuan University, bringing research experience in large language models, domain generalisation and few-shot learning. His recent work explores how training-data selection can improve the robustness and data efficiency of large multimodal models, addressing the challenge of developing models that perform reliably beyond their training data.

We also look forward to welcoming Tianyi Chen, who is expected to join the group as a DPhil student.

We warmly welcome all our new members and wish them a rewarding start to their time with the group. We look forward to the ideas and perspectives they will bring to our research and academic life.