EUVIMEDEUVIMED entdecken
Open Access

Accessible cartilage tumor malignancy prediction via vision-language foundation model adaptation

Xingxin He, Zachary E Stewart, Marcos R Gonzalez, Yin P Hung, Tara Shirin Ossiani, Yung Hsin Chen, Joseph Oliver Werenski, Ronald W Mercer, Zhaoye Zhou, Kendall Brown, Santiago A Lozano-Calderon, Fang Liu

Skeletal Radiology · 2026-05 · 55 (5) · 1073-1085

DOI: 10.1007/s00256-026-05131-4 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Skeletal Radiology, online erschienen am 17.01.2026. Verfasst von Xingxin He und 11 weiteren Autor:innen. Beteiligte Einrichtung: Athinoula A. Martinos Center for Biomedical Imaging, Harvard Medical School, Boston, MA, USA. Hauptthemen laut MEDLINE: Chondrosarcoma, Chondroma, Bone Neoplasms, Radiographic Image Interpretation, Computer-Assisted. Die Arbeit nennt 28 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: NIBIB NIH HHS, NIAMS NIH HHS, National Institute of Biomedical Imaging and Bioengineering.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

OBJECTIVE: To predict cartilage tumor malignancy from radiographic images combined with readily available non-imaging information based on a vision-language foundation model. MATERIALS AND METHODS: This single-institution study assembled a dataset of 3336 radiographs from 455 patients with enchondroma or chondrosarcoma that was assembled from two sources: (1) patients with histopathology-confirmed diagnoses of enchondroma or chondrosarcoma, and (2) patients with imaging-stable enchondroma without biopsy, confirmed through long-term imaging follow-up. An adapted vision-language foundation model based on the pre-trained CLIP (Contrastive Language-Image Pretraining) architecture was fine-tuned with our proposed Medical Knowledge Adapters and evaluated using 10-fold patient-level cross-validation to predict cartilage tumor malignancy from plain radiographs and demographic information. RESULTS AND CONCLUSION: Using radiographs alone, the model achieved an Areas Under the receiver operating characteristic Curve (AUC) of 0.91 ± 0.04. Incorporating demographics improved the AUC to 0.94 ± 0.02. Subgroup analysis demonstrated robust generalizability across tumor grades with an AUC of 0.91 ± 0.07 in distinguishing atypical cartilaginous tumors (ACT) previously known as low grade chondrosarcomas, and 0.95 ± 0.02 in differentiating high-grade chondrosarcomas from enchondromas. Within the clinically challenging extremity subgroup (enchondroma vs ACT/LGCS), the model achieved an AUC of 0.79 ± 0.14, reflecting diagnostic difficulty observed in clinical practice. This foundation model-based approach demonstrates strong performance using accessible data sources, offering a non-invasive, cost-efficient, and scalable solution for cartilage tumor assessment in musculoskeletal oncology.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

ChondrosarcomaChondromaBone NeoplasmsRadiographic Image Interpretation, Computer-AssistedHumansFemaleMaleAdultMiddle AgedAdolescentAged

Stichwörter der Autor:innen

Cartilage tumorMulti-modal deep learningMusculoskeletal oncologyVision-language foundation model

Autor:innen und Institutionen

  • Athinoula A. Martinos Center for Biomedical Imaging, Harvard Medical School, Boston, MA, USA.
  • Department of Radiology, Massachusetts General Hospital, Boston, MA, USA.
  • Department of Orthopaedic Surgery, Massachusetts General Hospital, Boston, MA, USA.
  • Department of Pathology, Massachusetts General Hospital, Boston, MA, USA.
  • Athinoula A. Martinos Center for Biomedical Imaging, Harvard Medical School, Boston, MA, USA. fliu12@mgh.harvard.edu.
  • Department of Radiology, Massachusetts General Hospital, Boston, MA, USA. fliu12@mgh.harvard.edu.

Förderung

  • NIBIB NIH HHS
  • NIAMS NIH HHS
  • National Institute of Biomedical Imaging and Bioengineering
  • National Institute of Arthritis and Musculoskeletal and Skin Diseases

Zitieren

APA

He, X., Stewart, Z. E., Gonzalez, M. R., Hung, Y. P., Ossiani, T. S., Chen, Y. H., Werenski, J. O., Mercer, R. W., Zhou, Z., Brown, K., Lozano-Calderon, S. A., & Liu, F. (2026). Accessible cartilage tumor malignancy prediction via vision-language foundation model adaptation. Skeletal Radiology, 55(5), 1073-1085. https://doi.org/10.1007/s00256-026-05131-4

Vancouver

He X, Stewart ZE, Gonzalez MR, Hung YP, Ossiani TS, Chen YH, et al. Accessible cartilage tumor malignancy prediction via vision-language foundation model adaptation. Skeletal Radiology. 2026;55(5):1073-1085. doi:10.1007/s00256-026-05131-4

BibTeX
@article{He2026,
  author = {He, Xingxin and Stewart, Zachary E and Gonzalez, Marcos R and Hung, Yin P and Ossiani, Tara Shirin and Chen, Yung Hsin and Werenski, Joseph Oliver and Mercer, Ronald W and Zhou, Zhaoye and Brown, Kendall and Lozano-Calderon, Santiago A and Liu, Fang},
  title = {Accessible cartilage tumor malignancy prediction via vision-language foundation model adaptation},
  journal = {Skeletal Radiology},
  year = {2026},
  volume = {55},
  number = {5},
  pages = {1073--1085},
  doi = {10.1007/s00256-026-05131-4}
}
Kostenlos starten →