Bone Magnetic Resonance Imaging: From Conventional Methods to AI ‐Driven Solutions
Journal of Magnetic Resonance Imaging · 2026-07 · 64 (1) · 30-54
DOI: 10.1002/jmri.70330 ↗Kurz gefasst
Übersichtsarbeit in der Zeitschrift Journal of Magnetic Resonance Imaging, online erschienen am 12.04.2026. Verfasst von Yi He und 3 weiteren Autor:innen. Beteiligte Einrichtung: Department of Biomedical Engineering, College of Chemistry and Life Science, Beijing University of Technology, Beijing, China. Hauptthemen laut MEDLINE: Magnetic Resonance Imaging, Bone and Bones, Artificial Intelligence, Bone Diseases. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 176 Literaturangaben. Gefördert von: Beijing Municipal Natural Science Foundation, National Natural Science Foundation of China, Natural Science Foundation of Beijing Municipality.
Abstract (Originalsprache)
With global population aging and increased life expectancy, bone diseases affect a substantial proportion of individuals worldwide, imposing a significant socioeconomic burden. Although X‐ray‐based imaging techniques remain the clinical standard for bone assessment, they are constrained by ionizing radiation exposure and poor soft‐tissue contrast. Magnetic resonance imaging (MRI) has emerged as a promising radiation‐free alternative, enabling detailed evaluation of tissue properties, bone microstructure, functional status, and pathological changes. However, the intrinsically low proton density and rapid transverse relaxation of bone tissue present fundamental technical challenges. Over the past decades, numerous MRI‐based techniques have been developed to address these limitations. Yet, the literature remains fragmented. This review synthesizes recent advancements in bone MRI, covering conventional sequences (e.g., T1‐weighted and T2‐weighted imaging), advanced sequences (e.g., ultrashort echo time and zero echo time), metabolic imaging sequences (e.g., magnetic resonance spectroscopy and dynamic contrast‐enhanced MRI), and hardware innovations (e.g., ultra‐high‐field MRI at 7/14 T). The strengths and limitations of these techniques are discussed, and their roles are highlighted in early diagnosis, therapeutic monitoring, and management of diverse bone disorders, including osteoporosis, osteoarthritis, osteosarcoma, and osteonecrosis. Furthermore, we explore how artificial intelligence, particularly deep learning models, enhances MRI capabilities by reducing scanning time, adding synthetic contrasts like synthetic CT, improving image quality, and increasing diagnostic accuracy. Finally, we outline challenges and future directions to advance bone assessment. This comprehensive review may guide preclinical research and accelerate clinical translation of bone MRI techniques, ultimately improving musculoskeletal disease management. Level of Evidence 4. Technical Efficacy Stage 2.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Substanzen
Autor:innen und Institutionen
- Department of Biomedical Engineering, College of Chemistry and Life Science, Beijing University of Technology, Beijing, China.
- Department of Radiology, Ghent University Hospital, Ghent, Belgium.
ORCID-Profile: Yi He, Lieve Morbée, Haisheng Yang
Förderung
- Beijing Municipal Natural Science Foundation
- National Natural Science Foundation of China
- Natural Science Foundation of Beijing Municipality
Zitieren
APA
He, Y., Din, R. U., Morbée, L., & Yang, H. (2026). Bone Magnetic Resonance Imaging: From Conventional Methods to AI ‐Driven Solutions. Journal of Magnetic Resonance Imaging, 64(1), 30-54. https://doi.org/10.1002/jmri.70330
Vancouver
He Y, Din RU, Morbée L, Yang H. Bone Magnetic Resonance Imaging: From Conventional Methods to AI ‐Driven Solutions. Journal of Magnetic Resonance Imaging. 2026;64(1):30-54. doi:10.1002/jmri.70330
BibTeX
@article{He2026,
author = {He, Yi and Din, Rahman Ud and Morbée, Lieve and Yang, Haisheng},
title = {Bone Magnetic Resonance Imaging: From Conventional Methods to AI ‐Driven Solutions},
journal = {Journal of Magnetic Resonance Imaging},
year = {2026},
volume = {64},
number = {1},
pages = {30--54},
doi = {10.1002/jmri.70330}
}
EUVIMED