A Master Class on Reproducibility: A Student Hackathon on Advanced MRI Reconstruction Methods

MCML Authors

Lina Felsner

Dr.

→ Group Julia Schnabel
 Computational Imaging and AI in Medicine

Daniel Rückert

Prof. Dr.

Director

Artificial Intelligence in Healthcare and Medicine

Abstract

We report the design, protocol, and outcomes of a student reproducibility hackathon focused on replicating the results of three influential MRI reconstruction papers: (a) MoDL, an unrolled model-based network with learned denoising [1]; (b) HUMUS-Net, a hybrid unrolled multi-scale CNN+Transformer architecture [2]; and (c) an untrained, physics-regularized dynamic MRI method that uses a quantitativeMRmodel for early stopping [3]. We describe the setup of the hackathon and present reproduction outcomes alongside additional experiments, and we detail fundamental practices for building reproducible codebases.

inproceedings FKE+26

---

BVM 2026

German Conference on Medical Image Computing -Bildverarbeitung für die Medizin. Lübeck, Germany, Mar 15-17, 2026.

---

Authors

L. Felsner • S. G. Kafali • H. Eichhorn • A. A. J. Leth • A. Batvinskas • A. Datchev • F. Klemm • J. Aulich • P. Leepagorn • R. Klinger • D. Rückert • J. A. 

---

Links

DOI

---

Research Area

 C1 | Medicine