EUVIMEDEUVIMED entdecken
Crossref

Generating synthetic CT from MR images for radiation-free computer-assisted quantification of hip morphology

Kangqing Ye, Wenyuan Sun, Guoyan Zheng

European Journal of Radiology · 2026

DOI: 10.1016/j.ejrad.2026.112962 ↗

Kurz gefasst

Fachartikel in der Zeitschrift European Journal of Radiology, 2026. Verfasst von Kangqing Ye und 2 weiteren Autor:innen. Beteiligte Einrichtung: Institute of Medical Robotics, School of Biomedical Engineering, Shanghai Jiao Tong University, 200240 Shanghai, China. Hauptthemen laut MEDLINE: Tomography, X-Ray Computed, Magnetic Resonance Imaging, Hip Joint, Image Interpretation, Computer-Assisted. Die Arbeit nennt 33 Literaturangaben. Gefördert von: Key Technologies Research and Development Program, National Key Research and Development Program of China, National Natural Science Foundation of China.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

OBJECTIVES: To compare MR image-based synthetic CT (sCT) with conventional CT for computer-assisted quantification of hip morphology by evaluating osseous structure segmentation quality and hip morphological parameters. MATERIALS AND METHODS: This retrospective study included patients undergoing both CT and MR scans. The sCT images generated from both high-resolution (HR) and simulated low-resolution (LR) MR images were evaluated. Segmentation accuracy was assessed using Dice similarity coefficient (DSC) and average symmetric surface distance (ASSD). Two readers independently measured seven typical hip morphological parameters for computer-assisted diagnosis of developmental hip dysplasia: anteversion, inclination, lateral center-edge (LCE) angle, extrusion index, and three femoral head coverage measurements (total, anterior, and posterior). Inter-modality, inter-reader, and intra-reader agreements were assessed using intraclass correlation coefficient (ICC) and mean absolute difference. RESULTS: The study included 70 patients, with 35 for model development and remaining 35 for testing. Compared with real CT, segmentation of the pelvis demonstrated high accuracy for both HR MR image-based sCT (DSC = 93.8 ± 1.7%, ASSD = 0.80 ± 0.36 mm) and LR MR image-based sCT (DSC = 93.6 ± 1.8%, ASSD = 0.82 ± 0.37 mm). Inter-modality agreement for most hip morphological parameters was excellent (ICC ≥ 0.90), while LCE angle for LR MR image-based sCT showed good agreement (ICC ≥ 0.75). Inter-reader and intra-reader agreements were excellent (ICC ≥ 0.90). No clinically relevant differences were observed between real CT and sCT for any morphological parameter. CONCLUSION: The MR image-based sCT provides accurate osseous structure segmentation and clinically acceptable hip morphological parameters, supporting the feasibility of radiation-free quantification of hip morphology.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Tomography, X-Ray ComputedMagnetic Resonance ImagingHip JointImage Interpretation, Computer-AssistedHumansReproducibility of ResultsFemaleMaleSensitivity and SpecificityAdolescentGenerative Artificial IntelligenceChildAdultAlgorithmsRetrospective Studies

Stichwörter der Autor:innen

Deep learningHip morphologyMagnetic resonance imagingSynthetic computed tomography

Autor:innen und Institutionen

  • Institute of Medical Robotics, School of Biomedical Engineering, Shanghai Jiao Tong University, 200240 Shanghai, China.
  • Institute of Medical Robotics, School of Biomedical Engineering, Shanghai Jiao Tong University, 200240 Shanghai, China. Electronic address: guoyan.zheng@sjtu.edu.cn.

Förderung

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

Zitieren

APA

Ye, K., Sun, W., & Zheng, G. (2026). Generating synthetic CT from MR images for radiation-free computer-assisted quantification of hip morphology. European Journal of Radiology. https://doi.org/10.1016/j.ejrad.2026.112962

Vancouver

Ye K, Sun W, Zheng G. Generating synthetic CT from MR images for radiation-free computer-assisted quantification of hip morphology. European Journal of Radiology. 2026. doi:10.1016/j.ejrad.2026.112962

BibTeX
@article{Ye2026,
  author = {Ye, Kangqing and Sun, Wenyuan and Zheng, Guoyan},
  title = {Generating synthetic CT from MR images for radiation-free computer-assisted quantification of hip morphology},
  journal = {European Journal of Radiology},
  year = {2026},
  doi = {10.1016/j.ejrad.2026.112962}
}
Kostenlos starten →