EUVIMEDEUVIMED entdecken
Open Access

Noise‐Level Adaptive Diffusion Priors for Zero‐Shot Low‐Field MRI Quality Enhancement

Jiacai Cai iD, Shoujin Huang iD, Zihao Wang iD, Shaojun Liu, Huaishui Yang iD, Min Wang iD, Lingfang Xiao, Zhuangzhuang Miao, Jiangbin Zhang, Yilong Liu iD, Mengye Lyu iD, Ed X Wu iD

NMR in Biomedicine · 2026-09 · 39 (9)

DOI: 10.1002/nbm.70373 ↗

Kurz gefasst

Fachartikel in der Zeitschrift NMR in Biomedicine, online erschienen am 14.08.2026. Verfasst von Jiacai Cai und 11 weiteren Autor:innen. Beteiligte Einrichtung: College of Health Science and Environmental Engineering, Shenzhen Technology University, Shenzhen, China. Hauptthemen laut MEDLINE: Signal-To-Noise Ratio, Magnetic Resonance Imaging, Image Enhancement, Diffusion Magnetic Resonance Imaging. Die Arbeit nennt 49 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: National Natural Science Foundation of China, Zhejiang Provincial Natural Science Foundation of China, National Key Research and Development Program of China.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Low‐field MRI (LF‐MRI) improves imaging accessibility, but its practical utility is often limited by reduced signal‐to‐noise ratio, weaker tissue contrast, and variable image quality. These challenges are further compounded by the limited availability of low‐field training data across scanners and the fact that raw k‐space data are not always accessible in research and clinical workflows. In this study, we adapted a high‐field‐trained diffusion model for LF‐MRI quality enhancement without low‐field retraining or fine‐tuning. During inference, the pretrained diffusion prior was combined with noise‐level adaptive measurement guidance to suppress low‐field noise while maintaining consistency with measurement‐supported anatomical structures. With phase augmentation, this framework was further extended to magnitude‐only DICOM‐exported inputs when measured phase information was unavailable. The method was evaluated on 10 locally recruited healthy volunteers scanned at 0.05, 0.3, and 3 T, a public 0.064‐T healthy‐subject dataset with 10 paired cases, and patient cases acquired at 0.05 and 0.35 T. Compared with raw LF input and representative baselines including BM3D, MiDiffusion, and NAFNet, the proposed method (Nila) improved visual image quality, noise suppression, and tissue contrast, while preserving lesion‐region appearance in patient cases. On the 0.3 and 0.064 T datasets with coregistered 3‐T references, Nila achieved the best LPIPS and the highest or tied‐highest NMI across contrasts, indicating improved similarity to the high‐field reference. Multisample posterior inference further produced uncertainty maps that provide an explainable‐AI view of spatial output variability, with elevated variance mainly localized to tissue boundaries and ambiguous regions. These results suggest that high‐field diffusion priors can serve as practical and reusable tools for LF‐MRI enhancement across heterogeneous systems.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Signal-To-Noise RatioMagnetic Resonance ImagingImage EnhancementDiffusion Magnetic Resonance ImagingHumansAlgorithmsDiffusion

Stichwörter der Autor:innen

diffusion modelelectromagnetic interferenceimage restorationlow‐field MRImagnitude‐only DICOMuncertainty estimationzero‐shot enhancement

Autor:innen und Institutionen

  • College of Health Science and Environmental Engineering, Shenzhen Technology University, Shenzhen, China.
  • Key Laboratory for Biomedical Engineering of Ministry of Education, College of Biomedical Engineering and Instrument Science, Zhejiang University, Hangzhou, China.
  • Shanghai Soundwise Technology Co. Ltd., Shanghai, China.
  • Department of Neurosurgery, Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China.
  • Department of Radiology, Huayin People's Hospital, Huayin City, China.
  • Guangdong-Hongkong-Macau CNS Regeneration Institute, Key Laboratory of CNS Regeneration (Jinan University)-Ministry of Education, Jinan University, Guangzhou, China.

ORCID-Profile: Jiacai Cai, Shoujin Huang, Zihao Wang, Huaishui Yang, Min Wang, Yilong Liu, Mengye Lyu, Ed X Wu

Förderung

  • National Natural Science Foundation of China
  • Zhejiang Provincial Natural Science Foundation of China
  • National Key Research and Development Program of China

Zitieren

APA

Cai, J., Huang, S., Wang, Z., Liu, S., Yang, H., Wang, M., Xiao, L., Miao, Z., Zhang, J., Liu, Y., Lyu, M., & Wu, E. X. (2026). Noise‐Level Adaptive Diffusion Priors for Zero‐Shot Low‐Field MRI Quality Enhancement. NMR in Biomedicine, 39(9). https://doi.org/10.1002/nbm.70373

Vancouver

Cai J, Huang S, Wang Z, Liu S, Yang H, Wang M, et al. Noise‐Level Adaptive Diffusion Priors for Zero‐Shot Low‐Field MRI Quality Enhancement. NMR in Biomedicine. 2026;39(9). doi:10.1002/nbm.70373

BibTeX
@article{Cai2026,
  author = {Cai, Jiacai and Huang, Shoujin and Wang, Zihao and Liu, Shaojun and Yang, Huaishui and Wang, Min and Xiao, Lingfang and Miao, Zhuangzhuang and Zhang, Jiangbin and Liu, Yilong and Lyu, Mengye and Wu, Ed X},
  title = {Noise‐Level Adaptive Diffusion Priors for Zero‐Shot Low‐Field MRI Quality Enhancement},
  journal = {NMR in Biomedicine},
  year = {2026},
  volume = {39},
  number = {9},
  doi = {10.1002/nbm.70373}
}
Kostenlos starten →