2 PhD Positions

100 %

Center for AI in Radiation Oncology (CAIRO) – Pediatric Digital Oncology

Duration: 3-4 years

Start date: As soon as possible, latest March 2026

The Center for AI in Radiation Oncology (CAIRO) is a newly established research group within the Department of Radiation Oncology at Inselspital and affiliated with the University of Bern. The Center focuses on developing data-driven and mechanistic models to improve diagnosis, treatment planning, and outcome prediction in radiation oncology.

Supported by a Swiss National Science Foundation Starting Grant, CAIRO investigates how artificial intelligence can support precision therapy in pediatric oncology, with a particular emphasis on Diffuse Midline Glioma (DMG) – a rare and difficult-to-treat brain tumor in children. Our interdisciplinary team combines expertise in data science, medical physics, computational biology, and clinical oncology, working closely with national and international partners to support these projects. We are currently inviting applications for two PhD positions focusing on complementary aspects of pediatric digital oncology.

Projects
Project 1: Predictive Modeling for Biomarker Identification and In-Silico Therapy Design in Pediatric Diffuse Midline Glioma

Background:
Recent advances in pediatric neuro-oncology have made new therapies, such as the DRD2 antagonist Modeyso (ONC201), available for children with DMG. However, individual responses vary substantially, and there is a need for computational tools to predict treatment outcomes and identify biomarkers of resistance and sensitivity.

Project Description:
This PhD project will develop and validate AI-based prediction models that link molecular tumor profiles to drug response, integrating multi-omics data (transcriptomics, methylation, CNV, mutational data) and preclinical screening results. The goal is to establish a clinically relevant framework for predicting treatment response and for in-silico identification of combination therapies in pediatric DMG. This project is suggested to compare two possible tracks to assess model generalization: foundation models and mechanistic learning to guide the selection of training regimes, included features and architecture design based on the current understanding of the drug's mechanism of action.

Methods And Data:

• Multi-omics data integration from pediatric and pan-cancer datasets including public and private data provided by our collaborators
• Deep learning and explainable AI approaches for drug response prediction
• Benchmarking of foundation models vs. meachnistic learning approaches for drug response modelling
• Model interpretation to identify molecular determinants of treatment sensitivity and resistance to recommend combination therapies

Candidate Profile:

• MSc in computer science, data science, bioinformatics, or a related field
• Experience in computational biology and the processing of different types of omics data
• Interest in translational applications of AI in pediatric oncology

Project 2: Personalized Radiotherapy for Pediatric Diffuse Midline Glioma

Background:
Radiotherapy remains the cornerstone of palliation and improved overall survival for DMG, yet its effect varies widely across patients. Predicting individual response before treatment and optimizing therapy delivery could improve both survival and quality of life.

Project Description:
This project aims to develop and test computational models that predict individual RT response based on pre-treatment imaging data and explore personalized adaptations of fractionation and dose distribution. Combining mechanistic modeling with AI-based image analysis, the project will contribute to a framework for individualized RT planning in pediatric brain tumors. Key challenges, such as limited data imply a dedicated training scheme and exploration of relevant transfer learning scenarios from adult GBM data available both publicly and at Inselspital.

Methods And Data:

• MRI-based prediction of RT response using MRI foundation models
• Mechanistic learning frameworks for tumor growth and radiosensitivity modeling
• Counterfactual simulations of alternative fractionation schemes
• Generative space-time modeling for anatomical prediction of tumor recurrence for a given radiotherapy treatment plan

Candidate Profile:

• MSc in (medical) physics, biomedical engineering, computer vision, or a related discipline
• Experience with medical image analysis and/or generative modeling
• Interest in radiotherapy modeling and quantitative imaging

*General Candidate Requirements*

• Master's degree in a relevant field (e.g. computational biology, computer science, data science, physics, biomedical engineering)
• Solid programming skills in Python; experience with machine learning frameworks (PyTorch, TensorFlow, MONAI)
• Motivation to work in a collaborative, interdisciplinary environment involving clinicians and data scientists
• Good communication skills and proficiency in English

We offer

• A fully funded 3.5 (project 1) and 4-year (project 2) PhD position within the Faculty of Medicine, and the Department of Digital Medicine, University of Bern
• Integration in a multidisciplinary team at the interface of medicine, physics, and computer science
• Access to unique pediatric datasets through international collaborations
• Support for conference participation and opportunities for teaching and mentoring
• Access to high-performance computing and clinical data environments within Inselspital and the University of Bern

Application Instructions
Please Submit In a Single PDF:

• Curriculum Vitae
• Cover letter outlining your motivation and preferred project (1–2 pages)
• Academic transcripts (Bachelor's and Master's)
• Contact information for two references or relevant letters of recommendation

Application Deadline:
30 November 2025

(applications will be reviewed on a rolling basis until the positions are filled)

Please send your application to

With The Subject Line:
"PhD Application – Pediatric Digital Oncology"

A note on LLM use. We understand these tools are widely used. However, please ensure your materials, especially your cover letter, reflect your own voice, experiences, and specific motivation for these projects. Generic or non-specific letters that do not address the fit with our pediatric-focused work will not be considered further.

Legal Notice

Apply now