Physics-informed neural networks for denoising high b-value diffusion-weighted images
Computerized Medical Imaging and Graphics · 2025
DOI: 10.1016/j.compmedimag.2025.102579 ↗Kurz gefasst
Fachartikel in der Zeitschrift Computerized Medical Imaging and Graphics, 2025. Verfasst von Qiaoling Lin und 10 weiteren Autor:innen. Beteiligte Einrichtung: Department of Electronic Science, Fujian Provincial Key Laboratory of Plasma and Magnetic Resonance, Xiamen University, Xiamen, 361102, Fujian, China. Hauptthemen laut MEDLINE: Diffusion Magnetic Resonance Imaging, Neural Networks, Computer, Prostatic Neoplasms, Image Processing, Computer-Assisted. Die Arbeit nennt 57 Literaturangaben. Gefördert von: National Key Research and Development Program of China, Xiamen University, Ministry of Education of the People's Republic of China.
Abstract (Originalsprache)
Diffusion-weighted imaging (DWI) is widely applied in tumor diagnosis by measuring the diffusion of water molecules. To increase the sensitivity to tumor identification, faithful high b-value DWI images are expected by setting a stronger strength of gradient field in magnetic resonance imaging (MRI). However, high b-value DWI images are heavily affected by reduced signal-to-noise ratio due to the exponential decay of signal intensity. Thus, removing noise becomes important for high b-value DWI images. Here, we propose a Physics-Informed neural Network for high b-value DWI images Denoising (PIND) by leveraging information from physics-informed loss and prior information from low b-value DWI images with high signal-to-noise ratio. Experiments are conducted on a prostate DWI dataset that has 125 subjects. Compared with the original noisy images, PIND improves the peak signal-to-noise ratio from 31.25 dB to 36.28 dB, and structural similarity index measure from 0.77 to 0.92. Our schemes can save 83% data acquisition time since fewer averages of high b-value DWI images need to be acquired, while maintaining 98% accuracy of the apparent diffusion coefficient value, suggesting its potential effectiveness in preserving essential diffusion characteristics. Reader study by 4 radiologists (3, 6, 13, and 18 years of experience) indicates PIND's promising performance on overall quality, signal-to-noise ratio, artifact suppression, and lesion conspicuity, showing potential for improving clinical DWI applications.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Electronic Science, Fujian Provincial Key Laboratory of Plasma and Magnetic Resonance, Xiamen University, Xiamen, 361102, Fujian, China.
- Department of Communication Engineering, Xiamen University, Xiamen, 361102, Fujian, China.
- Department of Radiology, Yuyao People's Hospital, Yuyao, 315400, Zhejiang, China.
- Intelligent Instrument and Equipment, Pen-Tung Sah Institute of Micro-Nano Science and Technology, Xiamen University, Xiamen, 361102, Fujian, China.
- China Mobile (Suzhou) Software Technology Company Limited, Suzhou, 215163, Jiangsu, China.
- Department of Radiology, Yuyao People's Hospital, Yuyao, 315400, Zhejiang, China. Electronic address: yyfsk123@qq.com.
Förderung
- National Key Research and Development Program of China
- Xiamen University
- Ministry of Education of the People's Republic of China
- Natural Science Foundation of Fujian Province
- National Natural Science Foundation of China
Zitieren
APA
Lin, Q., Yang, F., Yan, Y., Zhang, H., Xie, Q., Zheng, J., Yang, W., Qian, L., Liu, S., Yao, W., & Qu, X. (2025). Physics-informed neural networks for denoising high b-value diffusion-weighted images. Computerized Medical Imaging and Graphics. https://doi.org/10.1016/j.compmedimag.2025.102579
Vancouver
Lin Q, Yang F, Yan Y, Zhang H, Xie Q, Zheng J, et al. Physics-informed neural networks for denoising high b-value diffusion-weighted images. Computerized Medical Imaging and Graphics. 2025. doi:10.1016/j.compmedimag.2025.102579
BibTeX
@article{Lin2025,
author = {Lin, Qiaoling and Yang, Fan and Yan, Yang and Zhang, Haoyu and Xie, Qing and Zheng, Jiaju and Yang, Wenze and Qian, Ling and Liu, Shaoxing and Yao, Weigen and Qu, Xiaobo},
title = {Physics-informed neural networks for denoising high b-value diffusion-weighted images},
journal = {Computerized Medical Imaging and Graphics},
year = {2025},
doi = {10.1016/j.compmedimag.2025.102579}
}
EUVIMED