Cambridge Residency Programme: Postdoctoral Researcher - Multimodal Deep Learning for Healthcare

Microsoft Research Health Futures conducts world-class research at the intersection of AI and healthcare, drawing on deep expertise in signal processing, biomedical imaging and deep learning. We are seeking a motivated postdoctoral researcher to join our vibrant team of researchers and engineers. We focus on the development of models that can process raw medical instrument signals and improve diagnostic measurements and biomarker extraction. As a postdoctoral researcher within our team, you will have the opportunity to:

• Deepen your expertise in biomedical image reconstruction and AI.
• Explore multi-modal models for digital signal processing.
• Conduct experimentation with world-class computational resources., The ideal candidate will have strong intellectual curiosity and passion to solve real-world problems in healthcare using machine learning. Responsibilities will include:
• Design, implement, evaluate and deploy new machine learning methods and models.
• Share your research with the broader research community through publication or open sourcing.
• Collaborate with clinical experts and across Microsoft Research.
Work as part of a team to advance our ambitious research agenda.


• PhD degree (or on-track to complete), or equivalent expertise in independent research, in areas such as computer science (e.g. machine learning, deep learning, signal processing), electrical engineering, computational biology, medicine.
• Prior experience with deep learning frameworks (e.g., PyTorch) and software engineering practices (e.g. git).
• Passion for healthcare and medicine.
• Ability to work and learn in a collaborative and diverse environment., Experience with digital signal processing and/or medical image reconstruction.
• Design and evaluation of large multimodal transformers.
• Biomarker discovery (broadly interpreted).
• Experience with real-world healthcare data
• Track record of publication in top conferences or journals, such as: NeurIPS, ICML, ICLR, CVPR, EMNLP, Nature (and affiliated, e.g. NPJ Digital Health), JAMA, Radiology AI, Lancet Digital Health etc.