EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Hybrid learning: a combination of self-supervised and supervised learning for joint MRI reconstruction and denoising in low-field MRI

Haoyang Pei iD, Nikola Janjušević, Renqing Luo, Ding Xia, Xiang Xu, William Moore iD, Yao Wang, Hersh Chandarana iD, Li Feng iD

Physics in Medicine & Biology · 28.06.2026 · 71 (12) · 125029

DOI: 10.1088/1361-6560/ae792b ↗

Kurz gefasst

Fachartikel in der Zeitschrift Physics in Medicine & Biology, online erschienen am 25.06.2026. Verfasst von Haoyang Pei und 8 weiteren Autor:innen. Beteiligte Einrichtung: Bernard and Irene Schwartz Center for Biomedical Imaging, Department of Radiology, New York University Grossman School of Medicine, New York, NY, Uni…. Hauptthemen laut MEDLINE: Magnetic Resonance Imaging, Supervised Machine Learning, Signal-To-Noise Ratio, Image Processing, Computer-Assisted. Die Arbeit nennt 65 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: NIBIB NIH HHS, National Institutes of Health.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Objective. Deep learning has demonstrated strong potential for magnetic resonance imaging (MRI) reconstruction. However, conventional supervised learning requires high-quality, high-signal-to-noise-ratio (SNR) reference data for network training, which are often difficult or impossible to obtain, particularly in low-field MRI. Self-supervised learning (SSL) eliminates the need for reference training data but may suffer from degraded performance under low-SNR conditions. To address these limitations, we propose hybrid learning, a new training framework that integrates self-supervised and supervised learning for joint MRI reconstruction and denoising when only low-SNR training data are available. Approach. Hybrid learning is implemented in two sequential stages. In the first stage, SSL is applied to fully sampled low-SNR data to generate higher-quality pseudo-references. In the second stage, these pseudo-references are then used as targets for supervised learning to reconstruct and denoise undersampled, noisy data. The proposed method was evaluated in four experiments using simulated and real noisy MRI data of the breast, lung, and brain across different field strengths (0.3 T to 3 T), sampling trajectories (Cartesian, spiral, and radial), noise levels, and undersampling ratios. Main Results. Hybrid learning consistently improved reconstruction quality relative to both supervised and self-supervised baselines under different acceleration rates, noise levels, and sampling patterns in all experiments. Compared with standard supervised learning using noisy references, it achieved up to 167.70% higher structural similarity index measure (SSIM), 95.41% lower normalized mean squared error (NMSE), and 90.70% lower high-frequency error norm (HFEN). Compared with standard SSL, it achieved up to 23.88% higher SSIM, 60.85% lower NMSE, and 49.13% lower HFEN. Significance. Hybrid learning enables improved MRI reconstruction under low-SNR imaging conditions by jointly addressing noise and undersampling. It provides a practical solution for robust deep learning-based reconstruction and is particularly well suited for applications such as low-field MRI, where image quality is limited by reduced SNR.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Magnetic Resonance ImagingSupervised Machine LearningSignal-To-Noise RatioImage Processing, Computer-AssistedHumans

Stichwörter der Autor:innen

deep learningdenoisinghybrid learninglow-fieldlung MRIreconstructionself-supervised learning

Autor:innen und Institutionen

  • Bernard and Irene Schwartz Center for Biomedical Imaging, Department of Radiology, New York University Grossman School of Medicine, New York, NY, United States of America.
  • Center for Advanced Imaging Innovation and Research (CAI2R), Department of Radiology, New York University Grossman School of Medicine, New York, NY, United States of America.
  • Department of Electrical and Computer Engineering, NYU Tandon School of Engineering, New York, NY, United States of America.
  • Biomedical Engineering and Imaging Institute and Department of Radiology, Icahn School of Medicine at Mount Sinai, New York, NY, United States of America.

ORCID-Profile: Haoyang Pei, William Moore, Hersh Chandarana, Li Feng

Förderung

  • NIBIB NIH HHS
  • National Institutes of Health

Zitieren

APA

Pei, H., Janjušević, N., Luo, R., Xia, D., Xu, X., Moore, W., Wang, Y., Chandarana, H., & Feng, L. (2026). Hybrid learning: a combination of self-supervised and supervised learning for joint MRI reconstruction and denoising in low-field MRI. Physics in Medicine & Biology, 71(12), 125029. https://doi.org/10.1088/1361-6560/ae792b

Vancouver

Pei H, Janjušević N, Luo R, Xia D, Xu X, Moore W, et al. Hybrid learning: a combination of self-supervised and supervised learning for joint MRI reconstruction and denoising in low-field MRI. Physics in Medicine & Biology. 2026;71(12):125029. doi:10.1088/1361-6560/ae792b

BibTeX
@article{Pei2026,
  author = {Pei, Haoyang and Janjušević, Nikola and Luo, Renqing and Xia, Ding and Xu, Xiang and Moore, William and Wang, Yao and Chandarana, Hersh and Feng, Li},
  title = {Hybrid learning: a combination of self-supervised and supervised learning for joint MRI reconstruction and denoising in low-field MRI},
  journal = {Physics in Medicine & Biology},
  year = {2026},
  volume = {71},
  number = {12},
  pages = {125029},
  doi = {10.1088/1361-6560/ae792b}
}
Kostenlos starten →