EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Impact of tissue staining and scanner variation on the performance of pathology foundation models: a study of sarcomas and their mimics

Binghao Chai iD, Jianan Chen, Paul Cool, Fatine Oumlil, Anna Tollitt, David F Steiner, Tapabrata Chakraborti iD, Adrienne M Flanagan iD

The Journal of Pathology: Clinical Research · 2026-03 · 12 (2)

DOI: 10.1002/2056-4538.70080 ↗

Kurz gefasst

Fachartikel in der Zeitschrift The Journal of Pathology: Clinical Research, online erschienen am 24.02.2026. Verfasst von Binghao Chai und 7 weiteren Autor:innen. Beteiligte Einrichtung: Research Department of Pathology, UCL Cancer Institute, University College London, London, UK. Hauptthemen laut MEDLINE: Sarcoma, Staining and Labeling, Soft Tissue Neoplasms, Image Interpretation, Computer-Assisted. Die Arbeit wurde bisher 7-mal zitiert (Crossref) und nennt 47 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Sarcoma UK, Royal National Orthopaedic Hospital Research, Engineering and Physical Sciences Research Council.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Histopathological analysis is considered the gold standard for the diagnosis and prognostication of cancer. Recent advances in AI, driven by large‐scale digitisation and pan‐cancer foundation models, are opening new opportunities for clinical integration. However, it remains unclear how robust these foundation models are to real‐world sources of variability, particularly in H&E staining and scanners produced by different manufacturers. In this study, we use soft tissue tumours, a rare and morphologically diverse tumour type, as a challenging test case to systematically investigate the colour‐related robustness and generalisability of seven AI models. Controlled staining and scanning experiments were utilised to assess model performance across diverse real‐world data sources. Foundation models, particularly UNI‐v2, Virchow and TITAN, demonstrated encouraging robustness to staining and scanning variation, particularly when a small number of stain‐varied slides were included in the training loop, highlighting their potential as adaptable and data‐efficient tools for real‐world digital pathology workflows.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

SarcomaStaining and LabelingSoft Tissue NeoplasmsImage Interpretation, Computer-AssistedHumansArtificial Intelligence

Stichwörter der Autor:innen

cancer diagnosisclinical artificial intelligencedigital pathologyhistology stainingsarcomasoft tissue tumourswhole slide images

Autor:innen und Institutionen

  • Research Department of Pathology, UCL Cancer Institute, University College London, London, UK.
  • The Robert Jones and Agnes Hunt Orthopaedic Hospital, Gobowen, Oswestry, UK.
  • Keele University, Keele, UK.
  • Cellular and Molecular Pathology, Royal National Orthopaedic Hospital, Stanmore, UK.
  • Google Health, Mountain View, CA, USA.
  • The Alan Turing Institute, London, UK.

ORCID-Profile: Binghao Chai, Tapabrata Chakraborti, Adrienne M Flanagan

Förderung

  • Sarcoma UK
  • Royal National Orthopaedic Hospital Research
  • Engineering and Physical Sciences Research Council
  • Bone Cancer Research Trust

Zitieren

APA

Chai, B., Chen, J., Cool, P., Oumlil, F., Tollitt, A., Steiner, D. F., Chakraborti, T., & Flanagan, A. M. (2026). Impact of tissue staining and scanner variation on the performance of pathology foundation models: a study of sarcomas and their mimics. The Journal of Pathology: Clinical Research, 12(2). https://doi.org/10.1002/2056-4538.70080

Vancouver

Chai B, Chen J, Cool P, Oumlil F, Tollitt A, Steiner DF, et al. Impact of tissue staining and scanner variation on the performance of pathology foundation models: a study of sarcomas and their mimics. The Journal of Pathology: Clinical Research. 2026;12(2). doi:10.1002/2056-4538.70080

BibTeX
@article{Chai2026,
  author = {Chai, Binghao and Chen, Jianan and Cool, Paul and Oumlil, Fatine and Tollitt, Anna and Steiner, David F and Chakraborti, Tapabrata and Flanagan, Adrienne M},
  title = {Impact of tissue staining and scanner variation on the performance of pathology foundation models: a study of sarcomas and their mimics},
  journal = {The Journal of Pathology: Clinical Research},
  year = {2026},
  volume = {12},
  number = {2},
  doi = {10.1002/2056-4538.70080}
}
Kostenlos starten →