Nicholas (Nick) Konz

Email: nicholas (dot) konz (at) duke (dot) edu
Bluesky 🦋: @nickkonz.bsky.social
I’m a Ph.D. candidate studying machine learning at Duke University, working under Maciej Mazurowski. My research focuses on deep learning/computer vision for medical image analysis on a spectrum which ranges from application-oriented to foundational work, with an emphasis on topics like generative models, domain adaptation, and image-to-image translation.
I am particularly interested in how foundational deep learning concepts–such as generalization, intrinsic geometric properties of real-world datasets, and image distribution distance metrics–behave differently in medical image analysis and other secondary computer vision domains. This includes exploring how these concepts need to be adapted for unique challenges in these fields.
Beyond medical imaging, I’m drawn to the intersection of machine learning and science: understanding deep learning through a scientific lens, and leveraging it for scientific modeling, discovery, and applications in science-adjacent domains.
Previously, I worked as a research intern in the Math, Stats, and Data Science Group at PNNL. I earned my undergraduate degree at UNC, double-majoring in physics and mathematics, where I conducted research on statistical techniques for astronomy.
To learn more about my research, check out my full list of research topics and papers.
news
Aug 12, 2025 | Our paper, “Are Vision Foundation Models Ready for Out-of-the-Box Medical Image Registration?” (link here), has been accepted and selected for an oral presentation at the Deep-Brea3th Workshop at MICCAI 2025! |
---|---|
Jun 18, 2025 | Our paper, “SegmentAnyMuscle: A universal muscle segmentation model across different locations in MRI”, has been released on the arXiv! |
Jun 15, 2025 | Our paper, “Are Vision Foundation Models Ready for Out-of-the-Box Medical Image Registration?”, has been released on the arXiv! |
Jun 13, 2025 | Our paper, “MRI-CORE: A Foundation Model for Magnetic Resonance Imaging”, has been released on the arXiv! |
Jun 3, 2025 | Our paper, “Fréchet Radiomic Distance (FRD): A Versatile Metric for Comparing Medical Imaging Datasets”, has been released on the arXiv! |