Postdoctoral Fellow in Generative Space-Time Modeling for Pediatric Oncology

Entry April 2025 or upon agreement Temporary for 2 years

The new Center for AI in Radiation Oncology (CAIRO) within the Inselspital and affiliated with the University of Bern will investigate data-driven solutions for radiation oncology applications in the context of outcome predictions, treatment personalization, and multi-modal learning. At the core of our research is the collaboration across disciplines spanning expertise in medicine, biology, computer, and data science. As part of the new group and with the support of the SNSF Starting Grant we seek a motivated postdoctoral fellow to join this growing team and contribute to interdisciplinary research partnerships. The anticipated start date is spring 2025. Be part of a newly formed team that will start in the spring of 2025! We are seeking a dedicated and talented postdoctoral researcher to contribute to an exciting project at the intersection of medical imaging, computer vision, and oncology. The focus of this position is to develop and refine generative space-time models for predicting tumor growth and response to radiotherapy (RT) in pediatric patients with diffuse midline glioma (DMG). This role offers an opportunity to address critical challenges in understanding and managing a rare and fatal brain tumor while advancing methods for personalized treatment. By leveraging state-of-the-art machine learning approaches and clinical data, this project aims to improve patient care and provide a foundation for future research into pediatric precision oncology.

• Investigate and implement Denoising Diffusion Implicit Models (DDIM) for generating anatomical tumor images based on MRI data.
• Develop a space-time generative model to predict tumor growth trajectories from synthetic longitudinal MRI data.
• Integrate mechanistic mathematical models to guide generative models toward clinically relevant tumor growth predictions.
• Train models using a unique dataset of longitudinal multi-contrast MRIs and clinical data from DMG patients.
• Evaluate model performance using Response Assessment in Pediatric Neuro Oncology (RAPNO) criteria, tumor segmentations, and growth probability maps.
• Collaborate with multidisciplinary teams, including pediatric oncology experts and machine learning researchers.
• Communicate with clinical experts regarding the requirements for data preparation and feature extraction.
• Prepare manuscripts, and presentations to disseminate findings.
• Be part of an engaging and collaborative team, including supervision of junior researchers.

• PhD in computer science, biomedical engineering, or a related field with a strong focus on machine learning or computer vision.
• Expertise in generative modeling (e.g., diffusion models) and experience working with medical imaging data.
• Some background in mechanistic mathematical modeling is a plus
• Strong programming skills, particularly in Python and deep learning frameworks (e.g., PyTorch, TensorFlow).
• Experience in software development including collaborative coding, version control, and use of compute clusters.
• Interest or experience in healthcare applications of AI, particularly in oncology or radiology.
• Excellent problem-solving skills and the ability to work both independently and collaboratively in a multidisciplinary environment.
• You ideally have a background in biomedical projects with experience in interdisciplinary collaboration.
• Motivated to work as part of a team and strive towards scientific excellence in your field.
• Proficient in English in writing and speaking.

We offer a 2-year Postdoc position at the Faculty of Medicine of the University of Bern, 80-100% that includes:

• Opportunity to work on a multidisciplinary project combining medical imaging, machine learning, and clinical oncology.
• Access to high-quality datasets and collaboration with leading international DMG treatment centers.
• Competitive salary and funding for professional development opportunities.
• Opportunities to engage with different communities bridging data science and biomedical research leading to high-impact publications
• You will be part of a highly motivated, multidisciplinary and collaborative team
• We encourage the attendance of relevant (inter-) national conferences to increase your visibility and present the project outcomes
• You can be involved in the supervision of junior researchers and optional teaching in the lab
• Access to state-of-the-art computational resources and collaborative research networks
• Opportunities for professional development and career advancement