EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Exploring the risks of over-reliance on AI in diagnostic pathology. What lessons can be learned to support the training of young pathologists?

Yaëlle Bellahsen-Harrar iD, Mélanie Lubrano, Charles Lépine iD, Aurélie Beaufrère iD, Claire Bocciarelli, Anaïs Brunet, Elise Decroix iD, Franck Neil El-Sissy, Bettina Fabiani, Aurélien Morini, Cyprien Tilmant, Thomas Walter und 1 weitere

PLOS One · 28.08.2025 · 20 (8) · e0323270

DOI: 10.1371/journal.pone.0323270 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PLOS One, online erschienen am 28.08.2025. Verfasst von Yaëlle Bellahsen-Harrar und 12 weiteren Autor:innen. Beteiligte Einrichtung: Department of Pathology, Hôpital Européen Georges-Pompidou, APHP, Paris, France. Hauptthemen laut MEDLINE: Artificial Intelligence, Pathologists, Pathology, Clinical. Die Arbeit wurde bisher 13-mal zitiert (Crossref) und nennt 27 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: ANRT.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The integration of Artificial Intelligence (AI) algorithms into pathology practice presents both opportunities and challenges. Although it can improve accuracy and inter-rater reliability, it is not infallible and can produce erroneous diagnoses, hence the need for pathologists to always check predictions. This critical judgment is particularly important when algorithm errors could lead to high-impact negative clinical outcomes, such as missing an invasive carcinoma. However, the influence of AI tools on pathologists’ decision-making is not well explored. This study aims to evaluate the impact of a previously developed AI tool on the diagnostic accuracy and inter-rater reliability among pathologists, while assessing whether pathologists maintain independent judgment of AI predictions. Eight pathologists from different hospitals and with varying levels of experience, participated in the study. Each of them reviewed 115 slides of laryngeal biopsies, including benign epithelium, low-grade and high-grade dysplasia, and invasive squamous carcinomas. The study compared diagnostic outcomes with and without AI assistance. The reference labels were established by an expert’s double-blind review. Results show that assisted pathologists had a higher accuracy for high-grade dysplasia, invasive carcinoma and improved inter-rater reliability. However, cases of over-reliance on AI have been observed, resulting in the omission of correctly diagnosed invasive carcinomas during the unassisted examination. The false predictions on these carcinoma slides were labeled with a low confidence score, which was not considered by the less experienced pathologists, showing the risk that they would follow the AI prediction without enough critical judgment or expertise. Our study emphasizes the potential over-reliance of pathologists on AI models and the potential harmful consequences, even with the advancement of powerful algorithms. The integration of confidence scores and the education of pathologists to use this tool could help to optimize the safe integration of AI into pathology practice.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Artificial IntelligencePathologistsPathology, ClinicalHumansReproducibility of ResultsAlgorithmsObserver VariationDiagnostic Errors

Autor:innen und Institutionen

  • Department of Pathology, Hôpital Européen Georges-Pompidou, APHP, Paris, France.
  • Université Paris Cité, Paris, France.
  • Centre for Computational Biology (CBIO), Mines Paris, PSL University, Paris, France.
  • Tribun Health, Paris, France.
  • Department of Pathology, CHU Nantes, Nantes, France.
  • UMR 1302/EMR6001, INCIT, Inserm, Nantes, France.

ORCID-Profile: Yaëlle Bellahsen-Harrar, Charles Lépine, Aurélie Beaufrère, Elise Decroix

Förderung

  • ANRT

Zitieren

APA

Bellahsen-Harrar, Y., Lubrano, M., Lépine, C., Beaufrère, A., Bocciarelli, C., Brunet, A., Decroix, E., El-Sissy, F. N., Fabiani, B., Morini, A., Tilmant, C., Walter, T., & Badoual, C. (2025). Exploring the risks of over-reliance on AI in diagnostic pathology. What lessons can be learned to support the training of young pathologists?. PLOS One, 20(8), e0323270. https://doi.org/10.1371/journal.pone.0323270

Vancouver

Bellahsen-Harrar Y, Lubrano M, Lépine C, Beaufrère A, Bocciarelli C, Brunet A, et al. Exploring the risks of over-reliance on AI in diagnostic pathology. What lessons can be learned to support the training of young pathologists?. PLOS One. 2025;20(8):e0323270. doi:10.1371/journal.pone.0323270

BibTeX
@article{BellahsenHarrar2025,
  author = {Bellahsen-Harrar, Yaëlle and Lubrano, Mélanie and Lépine, Charles and Beaufrère, Aurélie and Bocciarelli, Claire and Brunet, Anaïs and Decroix, Elise and El-Sissy, Franck Neil and Fabiani, Bettina and Morini, Aurélien and Tilmant, Cyprien and Walter, Thomas and Badoual, Cécile},
  title = {Exploring the risks of over-reliance on AI in diagnostic pathology. What lessons can be learned to support the training of young pathologists?},
  journal = {PLOS One},
  year = {2025},
  volume = {20},
  number = {8},
  pages = {e0323270},
  doi = {10.1371/journal.pone.0323270}
}
Kostenlos starten →