Matthias Harders

This PhD project focuses on patient-specific liver deformation modelling for radio-frequency liver tumour ablation , where accurate prediction of organ movement is essential for treatment planning and intraoperative guidance. The project develops an AI-driven framework that combines strong modelling approaches to capture liver deformation at the patient level. A major part of the research is tackling the interpretability problem of deep learning systems. The PhD candidate will investigate methods for quantitative assessment of AI solutions , including failure mode analysis, model architecture analysis and reduction, and continual learning aspects. The work sits at the intersection of computer science and interventional radiology , with a clear translational medical application. The project is hosted at the University of Innsbruck and involves supervision from Matthias Harders in the Department of Computer Science and Reto Bale in the Department of Radiology at the Medical University Innsbruck. The position is part of an interdisciplinary doctoral environment with structured training across both universities and secondments at international partner institutions. The research environment combines AI, biomechanical simulation, and clinical imaging in a collaborative setting aimed at developing methods that are both technically strong and medically relevant.