Assessment of Robustness of MRI Radiomic Features in Four Abdominal Organs: Impact of Deep Learning Reconstruction and Segmentation
Journal of Magnetic Resonance Imaging · 2026-08 · 64 (2) · 468-479
DOI: 10.1002/jmri.70342 ↗Kurz gefasst
Evaluationsstudie in der Zeitschrift Journal of Magnetic Resonance Imaging, online erschienen am 07.05.2026. Verfasst von Jingyu Zhong und 14 weiteren Autor:innen. Beteiligte Einrichtung: Department of Imaging, Tongren Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China. Hauptthemen laut MEDLINE: Magnetic Resonance Imaging, Radiomics, Deep Learning, Image Interpretation, Computer-Assisted. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Laboratory Open Fund of Key Technology and Materials in Minimally Invasive Spine Surgery, Research Found of Science and Technology Commission of Changing District, Shanghai Municipality, National Natural Science Foundation of China.
Abstract (Originalsprache)
Background
The impact of deep learning (DL) reconstruction and segmentation on MRI radiomics stability has not been fully assessed.
Purpose
To investigate the effects of acquisition, reconstruction, and segmentation on the reproducibility and variability of radiomic features in abdominal MRI. Study Type Prospective. Population 37 volunteers (22 men; mean age ± standard deviation, 37.4 ± 11.0 years). Field Strength/Sequence 3.0‐T; axial turbo spin echo T2 ‐weighted image, and fat‐suppressed T2 ‐weighted image using a half‐Fourier acquisition single‐shot turbo spin echo technique, each acquired four times with conventional or accelerated techniques, reconstructed with standard or DL algorithms. Assessment Regions of interest were automatically generated by a DL neural network for liver, spleen, and right and left kidneys, followed by manual correction. We extracted 107 features using PyRadiomics after z ‐score normalization. Statistical Tests The reproducibility between acquisitions, reconstructions, and segmentations was evaluated using intraclass correlation coefficient (ICC) and concordance correlation coefficient (CCC). The variability among the four scans was assessed by coefficient of variation (CV) and quartile coefficient of dispersion (QCD). p < 0.05 was considered significant.
Results
The mean ICC (0.518–0.608; 0.606–0.681) and CCC (0.515–0.603; 0.601–0.680) values were low for both manual and automatic segmentation regardless of image acquisition and reconstruction, using conventional acquisition with standard reconstruction as reference. The mean ICC (0.535–0.713) and CCC (0.531–0.714) values were low between manual and automatic segmentation, regardless of image acquisition and reconstruction. The median CV (10.0%–17.5%; 8.9%–15.5%) and QCD (5.3%–8.5%; 5.1%–8.3%) values were moderate but still adequate for both manual and automatic segmentation among different scans.
Conclusion
Given the substantial impact of accelerated acquisition and DL reconstruction on the robustness of radiomics features in abdominal MRI, caution should be exercised when utilizing images with different acquisition and reconstruction techniques in radiomics analysis. The automatic segmentation cannot replace manual segmentation due to insufficient robustness of radiomics features. Evidence Level 2. Technical Efficacy Stage 1.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Imaging, Tongren Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China.
- Shanghai Key Laboratory of Flexible Medical Robotics, Tongren Hospital, Institute of Medical Robotics, Shanghai Jiao Tong University, Shanghai, China.
- Department of Epidemiology and Population Health, Stanford University School of Medicine, Stanford, California, USA.
- Department of Biomedical Engineering, Boston University, Boston, Massachusetts, USA.
- Jacobi Medical Center, Albert Einstein College of Medicine, New York, New York, USA.
- MR Scientific Marketing, Siemens Healthineers Ltd., Shanghai, China.
ORCID-Profile: Jingyu Zhong, Junjie Lu, Yang Song, Dominik Nickel, Huan Zhang
Förderung
- Laboratory Open Fund of Key Technology and Materials in Minimally Invasive Spine Surgery
- Research Found of Science and Technology Commission of Changing District, Shanghai Municipality
- National Natural Science Foundation of China
- Research Found of Health Commission of Changing District, Shanghai Municipality
- Research Fund of Tongren Hospital, Shanghai Jiao Tong University School of Medicine
- Research Found of Center for Community Health Care, China Hospital Development Institute, Shanghai Jiao Tong University
- Research Found of Health Commission of Shanghai Municipality
Zitieren
APA
Zhong, J., Xing, Y., Hu, Y., Liu, X., Dai, S., Ding, D., Lu, J., Yang, J., Li, Y., Song, Y., Lu, M., Nickel, D., Lu, W., Zhang, H., & Yao, W. (2026). Assessment of Robustness of MRI Radiomic Features in Four Abdominal Organs: Impact of Deep Learning Reconstruction and Segmentation. Journal of Magnetic Resonance Imaging, 64(2), 468-479. https://doi.org/10.1002/jmri.70342
Vancouver
Zhong J, Xing Y, Hu Y, Liu X, Dai S, Ding D, et al. Assessment of Robustness of MRI Radiomic Features in Four Abdominal Organs: Impact of Deep Learning Reconstruction and Segmentation. Journal of Magnetic Resonance Imaging. 2026;64(2):468-479. doi:10.1002/jmri.70342
BibTeX
@article{Zhong2026,
author = {Zhong, Jingyu and Xing, Yue and Hu, Yangfan and Liu, Xianwei and Dai, Shun and Ding, Defang and Lu, Junjie and Yang, Jiarui and Li, Yue and Song, Yang and Lu, Minda and Nickel, Dominik and Lu, Wenjie and Zhang, Huan and Yao, Weiwu},
title = {Assessment of Robustness of MRI Radiomic Features in Four Abdominal Organs: Impact of Deep Learning Reconstruction and Segmentation},
journal = {Journal of Magnetic Resonance Imaging},
year = {2026},
volume = {64},
number = {2},
pages = {468--479},
doi = {10.1002/jmri.70342}
}
EUVIMED