Physical Color Calibration of Digital Pathology Scanners for Robust Artificial Intelligence–Assisted Cancer Diagnosis
Modern Pathology · 2025
DOI: 10.1016/j.modpat.2025.100715 ↗Kurz gefasst
Fachartikel in der Zeitschrift Modern Pathology, 2025. Verfasst von Xiaoyi Ji und 16 weiteren Autor:innen. Beteiligte Einrichtung: Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden. Hauptthemen laut MEDLINE: Prostatic Neoplasms, Artificial Intelligence, Image Interpretation, Computer-Assisted. Die Arbeit wurde bisher 29-mal zitiert (Crossref) und nennt 46 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Innovate UK, Swedish Cancer Society, Oskar Huttunen Foundation.
Abstract (Originalsprache)
The potential of artificial intelligence (AI) in digital pathology is limited by technical inconsistencies in the production of whole slide images (WSIs). This causes degraded AI performance and poses a challenge for widespread clinical application, as fine-tuning algorithms for each site is impractical. Changes in the imaging workflow can also compromise diagnostic accuracy and patient safety. Physical color calibration of scanners, relying on a biomaterial-based calibrant slide and a spectrophotometric reference measurement, has been proposed for standardizing WSI appearance, but its impact on AI performance has not been investigated. We evaluated whether physical color calibration can enable robust AI performance. We trained fully supervised and foundation model-based AI systems for detecting and Gleason grading prostate cancer using WSIs of prostate biopsies from the STHLM3 clinical trial (n = 3651) and evaluated their performance in 3 external cohorts (n = 1161) with and without calibration. With physical color calibration, the fully supervised system's concordance with pathologists' grading (Cohen linearly weighted κ) improved from 0.439 to 0.619 in the Stavanger University Hospital cohort (n = 860), from 0.354 to 0.738 in the Karolinska University Hospital cohort (n = 229), and from 0.423 to 0.452 in the Aarhus University Hospital cohort (n = 72). The foundation model's concordance improved as follows: from 0.739 to 0.760 (Karolinska), from 0.424 to 0.459 (Aarhus), and from 0.547 to 0.670 (Stavanger). This study demonstrated that physical color calibration provides a potential solution to the variation introduced by different scanners, making AI-based cancer diagnostics more reliable and applicable in diverse clinical settings.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden.
- PathQA Ltd, United Kingdom.
- Institute of Biomedicine, University of Turku, Turku, Finland.
- Department of Pathology, Stavanger University Hospital, Stavanger, Norway; Faculty of Health Sciences, University of Stavanger, Stavanger, Norway.
- Department of Radiology, Aarhus University Hospital, Aarhus, Denmark; Department of Clinical Medicine, Aarhus University, Aarhus, Denmark.
- Department of Clinical Medicine, Aarhus University, Aarhus, Denmark; Department of Molecular Medicine, Aarhus University Hospital, Aarhus, Denmark.
Förderung
- Innovate UK
- Swedish Cancer Society
- Oskar Huttunen Foundation
- David and Astrid Hagelen Foundation
- Vetenskapsrådet
- Aarhus Universitetshospital
- Karolinska Institutet
- Research Council of Finland
Zitieren
APA
Ji, X., Salmon, R., Mulliqi, N., Khan, U., Wang, Y., Blilie, A., Olsson, H., Pedersen, B. G., Sørensen, K. D., Ulhøi, B. P., Kjosavik, S. R., Janssen, E. A. M., Rantalainen, M., Egevad, L., Ruusuvuori, P., Eklund, M., & Kartasalo, K. (2025). Physical Color Calibration of Digital Pathology Scanners for Robust Artificial Intelligence–Assisted Cancer Diagnosis. Modern Pathology. https://doi.org/10.1016/j.modpat.2025.100715
Vancouver
Ji X, Salmon R, Mulliqi N, Khan U, Wang Y, Blilie A, et al. Physical Color Calibration of Digital Pathology Scanners for Robust Artificial Intelligence–Assisted Cancer Diagnosis. Modern Pathology. 2025. doi:10.1016/j.modpat.2025.100715
BibTeX
@article{Ji2025,
author = {Ji, Xiaoyi and Salmon, Richard and Mulliqi, Nita and Khan, Umair and Wang, Yinxi and Blilie, Anders and Olsson, Henrik and Pedersen, Bodil Ginnerup and Sørensen, Karina Dalsgaard and Ulhøi, Benedicte Parm and Kjosavik, Svein R and Janssen, Emilius A M and Rantalainen, Mattias and Egevad, Lars and Ruusuvuori, Pekka and Eklund, Martin and Kartasalo, Kimmo},
title = {Physical Color Calibration of Digital Pathology Scanners for Robust Artificial Intelligence–Assisted Cancer Diagnosis},
journal = {Modern Pathology},
year = {2025},
doi = {10.1016/j.modpat.2025.100715}
}
EUVIMED