Postdoctoral Position (m/f/d) Machine Learning for the Acceleration of Simulations

• Das Universitätsklinikum Schleswig-Holstein (UKSH) verbindet internationale Spitzenforschung mit interdisziplinärer Krankenversorgung. Wir sind einziger Maximalversorger und größter Arbeitgeber des Landes. Unsere rund 17.500 Mitarbeitenden stellen eine höchst individuelle Versorgung sicher - unverzichtbar für die Menschen in Schleswig-Holstein.
• VIVID Research Group
• Department of Radiology and Neuroradiology
• Project: STRIVE - Simulation-based TRIals for Vascular trEatment

• About the project
• Cardiovascular diseases remain the leading cause of death worldwide, yet current clinical approval for disease treatment, especially for novel implants, relies on population-based studies that overlook individual patient variability. In silico trials — computer-based simulations of clinical scenarios — offer a powerful alternative: by creating digital twins of patients and implants, treatment outcomes can be tested rapidly and safely, without invasive procedures or large-scale clinical studies.

STRIVE aims to develop a regulatory-aligned computational framework for personalized vascular interventions, with an initial focus on intracranial aneurysms. The project combines advanced simulation methods, experimental validation, and close collaboration with clinicians and industry partners, ultimately enabling faster implant development, safer preoperative planning, and a meaningful reduction in the number of animal and human trials. As part of a young interdisciplinary team, you will contribute directly to building this framework and shaping the medicine of tomorrow.

• Start in our team
• We are looking for professional and competent support to start as soon as possible, limited for 4 years.

• What we offer:
• •The salary will be based on the German E13 TV-L scale (100%), if terms and conditions under collective bargaining law are fulfilled
• •A full-time employment, currently 38.5 hours/week; a part-time employment may be possible within the framework of certain working time models
• •Flexible working hours to accommodate individual needs
• •Interdisciplinary research at the intersection of medicine, physics, engineering, and computer science
• •A vibrant, international biomedical research group embedded within a university hospital setting

• Your tasks:
• •Develop data-driven surrogate models to replace computationally expensive physics-based simulations for aneurysm treatment, namely, computational fluid dynamics (CFD) and finite element modeling (FEM), enabling rapid, on-demand simulation of patient-specific treatment scenarios
• •Design and implement machine learning (ML) pipelines using full-fidelity CFD and FEM data, ensuring models generalize across patient geometries and device configurations
• •Integrate trained surrogate models into the broader in silico trial framework and evaluate their accuracy, robustness, and computational efficiency against full-fidelity simulations
• •Explore AI-based methods for treatment outcome prediction and device optimization, contributing to the framework's clinical decision-support capabilities
• •Collaborate closely with clinicians, scientists, and engineers to develop and validate cutting-edge simulation techniques
• •Lead and co-author scientific publications, and represent the project at international conferences

• Your profile:
• •PhD in Computer Science, Applied Mathematics, Physics, Engineering, or a related field with a strong computational focus
• •Strong practical experience in machine learning for scientific applications — including surrogate modeling, reduced-order models (ROM), or physics-informed machine learning (PIML/PINNs) — is the core technical foundation of the role
• •Proficiency in Python and relevant ML frameworks (e.g., PyTorch, TensorFlow, JAX) is essential
• •Familiarity with CFD, FEM, or other numerical simulation methods is a significant advantage
• •Experience with high-performance computing (HPC) environments is beneficial

Please submit a motivation letter (max. 1 page) and CV (max. 2 pages), including, where applicable, degree certificates and job references, as a single PDF until 27. October 2026, indicating the reference number 28652.

For questions about the role, please contact the project leader, Dr.-Ing. Mariya Pravdivtseva at [email protected].