Primitive Representation Learning for Unsupervised Dynamic Contrast Enhanced MRI Reconstruction

MCML Authors

Daniel Rückert

Prof. Dr.

Core PI

Artificial Intelligence in Healthcare and Medicine

Julia Schnabel

Prof. Dr.

Core PI

Computational Imaging and AI in Medicine

Abstract

Reliable quantitative analysis of dynamic contrast-enhanced MRI requires high-quality spatiotemporal reconstructions at high undersampling rates. Scan-specific reconstructions using Gaussian and Gabor primitives have shown promising results without the need for large training datasets, but have not addressed the additional dimension of dynamic contrast. We propose a multi-dimensional, primitive based framework for dynamic contrast-enhanced MRI reconstruction that disentangles the underlying anatomy, the dynamic contrast enhancement, and residual motion into separate temporal basis functions, thereby enabling a geometrical interpretation of the representation. We show that this architecture achieves performance competitive with conventional reconstruction methods, both in reconstruction quality and in the accuracy of extracted aorta and kidney enhancement curves. The modular tier design extends naturally to additional dynamic factors and higher acceleration rates.

inproceedings SHA+26

---

Off-Grid @MICCAI 2026

1st Workshop on Continuous Representations and Grid-Free Methods in Medical Imaging at the 29th International Conference on Medical Image Computing and Computer Assisted Intervention. Strasbourg, France, Sep 27-Oct 01, 2026. To be published. Preprint available.

---

Authors

V. Spieker • W. Huang • C. Ariyurek • L. Timms • D. Rückert • O. Afacan • J. A. Schnabel • S. Kurugol

---

Links

URL GitHub

---

Research Area

 C1 | Medicine