EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Memory-efficient image reconstruction using diffusion models for accelerated 3D non-Cartesian UTE imaging

Jonas Petersen, Stefan Sommer, Thomas Küstner

Magnetic Resonance Materials in Physics, Biology and Medicine · 19.08.2026

DOI: 10.1007/s10334-026-01403-1 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Magnetic Resonance Materials in Physics, Biology and Medicine, online erschienen am 19.08.2026. Verfasst von Jonas Petersen und 2 weiteren Autor:innen. Beteiligte Einrichtung: Medical Image and Data Analysis (MIDAS.lab), Department of Diagnostic and Interventional Radiology, University of Tübingen, Tübingen, Germany. Jonas.…. Die Arbeit nennt 35 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Deutsche Forschungsgemeinschaft, Universitätsklinikum Tübingen.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Purpose

Accelerated 3D non-Cartesian MRI presents unique challenges in balancing high-resolution reconstruction with computational and memory constraints. In this work, a novel, memory-efficient image reconstruction framework using score-based diffusion models tailored for highly undersampled 3D radial UTE acquisitions is proposed.

Methods

The method leverages a volumetric extension of the NCSN++ architecture with efficient NUFFT operations and integrates a conjugate gradient-based data consistency mechanism within a predictor–corrector sampling scheme.

Results

This approach enables robust and high-fidelity reconstructions from as few as 4000 spokes—representing up to 50 × undersampling relative to Nyquist. Generalizable performance across knee, ankle, and out-of-distribution shoulder datasets is demonstrated and superior image quality compared to compressed sensing and conventional iterative methods, especially at high acceleration factors is shown.

Conclusion

The proposed framework marks a significant step toward practical deployment of diffusion models in non-Cartesian MRI reconstruction.

Schlagwörter

Stichwörter der Autor:innen

Deep learningDiffusion modelsImage reconstructionMSK

Autor:innen und Institutionen

  • Medical Image and Data Analysis (MIDAS.lab), Department of Diagnostic and Interventional Radiology, University of Tübingen, Tübingen, Germany. Jonas.petersen@siemens-healthineers.com.
  • Research and Clinical Translation, Magnetic Resonance, Siemens Healthineers AG, Erlangen, Germany. Jonas.petersen@siemens-healthineers.com.
  • Swiss Center for Musculoskeletal Imaging (SCMI), Balgrist Campus, Zurich, Switzerland.
  • Swiss Innovation Hub, Siemens Healthineers International AG, Zurich, Switzerland.
  • Medical Image and Data Analysis (MIDAS.lab), Department of Diagnostic and Interventional Radiology, University of Tübingen, Tübingen, Germany.

Förderung

  • Deutsche Forschungsgemeinschaft
  • Universitätsklinikum Tübingen

Zitieren

APA

Petersen, J., Sommer, S., & Küstner, T. (2026). Memory-efficient image reconstruction using diffusion models for accelerated 3D non-Cartesian UTE imaging. Magnetic Resonance Materials in Physics, Biology and Medicine. https://doi.org/10.1007/s10334-026-01403-1

Vancouver

Petersen J, Sommer S, Küstner T. Memory-efficient image reconstruction using diffusion models for accelerated 3D non-Cartesian UTE imaging. Magnetic Resonance Materials in Physics, Biology and Medicine. 2026. doi:10.1007/s10334-026-01403-1

BibTeX
@article{Petersen2026,
  author = {Petersen, Jonas and Sommer, Stefan and Küstner, Thomas},
  title = {Memory-efficient image reconstruction using diffusion models for accelerated 3D non-Cartesian UTE imaging},
  journal = {Magnetic Resonance Materials in Physics, Biology and Medicine},
  year = {2026},
  doi = {10.1007/s10334-026-01403-1}
}
Kostenlos starten →