PhD in Credible Hybrid AI–Mechanistic Models for Clinical Decision-Making: Sensitivity Analysis for ECMO Weaning Eindhoven University of Technology in Netherlands

This fully funded 4-year PhD position at the Department of Biomedical Engineering, Eindhoven University of Technology (TU/e), offers an exciting opportunity to advance clinical decision support through credible hybrid AI–mechanistic models. The project, HY-Credibility, focuses on developing trustworthy, explainable models for cardiovascular biomechanics, specifically targeting safer and more personalized ECMO (extracorporeal membrane oxygenation) weaning. You will be embedded in the Cardiovascular Biomechanics group led by Professor Huberts and co-supervised by Dr. Jemima Tabeart from the Computational Science group, Centre for Analysis, Scientific Computing and Applications (Department of Mathematics & Computer Science).

Your research will combine first-principles cardiovascular pathophysiology with data-driven AI, leveraging sensitivity analysis (SA) and uncertainty quantification (UQ) to enhance model transparency and trustworthiness. The project involves designing hybrid models that integrate fast physiological signals with slower patient-specific dynamics, building an SA/UQ toolbox for high-dimensional systems, and working with unique datasets from an advanced ECMO mock loop. You will conduct simulation and data assimilation studies, evaluate model credibility, robustness, and generalizability, and collaborate closely with clinicians and interdisciplinary experts.

The impact of your work will be direct and significant, contributing to safer, more personalized care for critically ill patients by reducing trial-and-error decisions and improving outcomes. More broadly, you will help establish trustworthy hybrid AI methodologies for safety-critical healthcare applications. The research environment at TU/e is highly collaborative and inclusive, with strong links to clinical partners and leading expertise in computational modelling, digital twins, and uncertainty quantification.

Applicants should have a strong background in biomedical engineering, applied mathematics, computer science, or related fields, with experience in simulation, data assimilation, sensitivity analysis, or uncertainty quantification highly desirable. The position offers opportunities to publish results, contribute to open datasets and tools, and supervise BSc/MSc students. Apply online via the AcademicTransfer portal and join a multidisciplinary team working to revolutionize personalized medicine.