Julian P. Merkofer

My research lies at the intersection of signal processing, machine learning, and biomedical data analysis. I am interested in developing learning methods that remain closely connected to the signal models and physical processes underlying the data, with a focus on inverse problems, uncertainty estimation, and interpretable inference. I currently apply these ideas to magnetic resonance spectroscopy (MRS), where I study how data-driven approaches can be combined with simulations and model-based methods. Alongside this work, I contribute to open-source tools that support reproducible and collaborative research in the MRS community.

I recently defended my PhD at TU/e on Model-Based Machine Learning for Magnetic Resonance Spectroscopy, and will be joining the University of Oxford as an NWO Rubicon postdoctoral fellow in late 2026.