PhD candidate 'Multimodal and Agentic AI for Precision Prostate Cancer Care'

Job description

This position is embedded within the PROMETHEUS project: PROstate cancer Multimodal Explainable Transferable Holistic Expert for Universal Stratification.

PROMETHEUS aims to develop a comprehensive AI "super-agent" that integrates multimodal health data, including MRI examinations, PSMA PET/CT imaging, histopathology results, and longitudinal electronic health records, to support clinical decision-making in prostate cancer management.

Research Objectives

• Predictive diagnosis: Should this patient undergo a prostate biopsy, and what is the likelihood that an identified lesion represents aggressive cancer?
• Personalised treatment selection: How is the patient's disease likely to evolve under active surveillance, surgery, or multimodal treatment?
• Evidence-based decision support: How can AI recommendations be accompanied by relevant evidence, uncertainty estimates, and transparent explanations aligned with current clinical guidelines?

Preferred starting date: 1 November 2026

Place of work

Our group offers an exceptional environment for clinically grounded medical AI research.

• We maintain the world's largest international research cohort of prostate MRI examinations, comprising more than 22,000 examinations from 46 cities across 22 countries. This cohort has been established in collaboration with a growing scientific consortium of academic and community hospitals, researchers, and clinicians.
• Our previous research includes landmark studies in prostate AI (PI-CAI) and pancreas AI (PANORAMA), published in The Lancet Oncology. These studies focused on developing and rigorously benchmarking state-of-the-art AI systems against more than 60 international radiologists from 45 centres across 20 countries.
• PROMETHEUS brings together six leading European institutions with complementary expertise in medical imaging, prostate cancer management, longitudinal modelling, and medical AI.

You will be embedded within the Minimally Invasive Image-Guided Intervention Center (MAGIC) and the Diagnostic Image Analysis Group (DIAG) at Radboud University Medical Center in Nijmegen.

MAGIC develops and translates image-guided technologies for minimally invasive cancer diagnosis and treatment into clinical practice and is internationally recognised for its expertise in prostate imaging and intervention.

DIAG is a world-renowned medical AI research group comprising more than 70 researchers. The group combines expertise in radiology, pathology, computer science, and biomedical engineering. Its work spans methodological AI research, large-scale clinical validation, and the translation of AI innovations into medical imaging systems used in clinical practice.

Learn here what it's like to do a PhD at Radboudumc

Profile

We welcome candidates with an MSc or equivalent expertise in artificial intelligence, medicine, computer or data science, biomedical engineering, technical medicine, or a related field.

You should have strong programming skills (Python, PyTorch), deep learning knowledge (multimodal learning, longitudinal modelling, agentic AI, uncertainty estimation), and an interest in clinically-grounded oncology research.

We particularly encourage applications from individuals across all groups traditionally underrepresented in the STEM fields.

Good to know

• You will be appointed for four years as a PhD candidate with the standard salary and secondary conditions for PhD candidates in the Netherlands. The salary depends on education. Scale 10 will be offered to MSc medical graduates. Those holding degrees in other disciplines will be offered scale 10A.
• The research should result in publications, leading up to a PhD thesis.
• This position requires a Certificate of Conduct (VOG).

We are recruiting for this position ourselves. Unsolicited marketing is not appreciated, but do feel free to share the vacancy in your network!