Trustworthy Generative AI in Healthcare

About the project

About the Project For more information and to apply, please visit the institution website. One fully funded PhD studentship, open to UK and international applicants, is available in trustworthy generative AI for healthcare. The project sits in the broad area of statistical machine learning, with a focus on multimodal large language models, uncertainty-aware AI, and 3D medical imaging. You will develop and evaluate generative and vision language models that fuse medical images, video, and tabular clinical data for disease detection, prediction, and decision making. A central theme is trustworthiness: calibrated uncertainty, robustness under distribution shift, and evaluation frameworks that hold up in high-stakes clinical settings. Applications are drawn from oral and wider healthcare, including radiographic interpretation and clinical record analysis, working alongside clinicians in the Faculty of Dentistry, Oral & Craniofacial Sciences. Prior domain knowledge of dentistry or medicine is not required. The project is based in the Translational AI Research (TAIR) Lab at the Centre for Oral, Clinical & Translational Sciences in the Faculty of Dentistry, Oral & Craniofacial Sciences at King’s College London, with wider access to resources across King’s, including those in the Department of Informatics in the Faculty of Natural, Mathematical & Engineering Sciences. Through this project, the successful student will develop or enhance skills in statistical machine learning foundations and applications, training and evaluation of multimodal and generative models, uncertainty quantification, research dissemination, and interdisciplinary research and communication.

Research areas