EUVIMEDEUVIMED entdecken
Crossref

How Generalizable Are Foundation Models When Applied to Different Demographic Groups and Settings?

Zhuxin Xiong iD, Xiaofei Wang iD, Yukun Zhou iD, Pearse A Keane iD, Yih Chung Tham iD, Ya Xing Wang iD, Tien Yin Wong iD

NEJM AI · 01.01.2025 · 2 (1)

DOI: 10.1056/aics2400497 ↗

Kurz gefasst

Fachartikel in der Zeitschrift NEJM AI, erschienen am 01.01.2025. Verfasst von Zhuxin Xiong und 6 weiteren Autor:innen. Beteiligte Einrichtung: Key Laboratory for Biomechanics and Mechanobiology, Ministry of Education, Beijing Advanced Innovation Center for Biomedical Engineering, School of B…. Die Arbeit wurde bisher 17-mal zitiert (Crossref) und nennt 5 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

AbstractRETFound is a retinal image-based foundational artificial intelligence (AI) model that can be fine-tuned to downstream tasks. However, its generalizability to Asian populations remains unclear. In this study, we fine-tuned RETFound on an Asian-specific dataset. We then evaluated the performance of RETFound versus a conventional Vision Transformer model (pretrained on ImageNet) in diagnosing glaucoma and coronary heart disease and predicting the 3-year risk of stroke in an Asian population. When fine-tuned on a "full" dataset, RETFound showed no significant improvement compared with a conventional Vision Transformer model (area under the curves [AUCs] of 0.863, 0.628, and 0.557 vs. 0.853, 0.621, and 0.543, respectively; all P≥0.2). Furthermore, in scenarios with limited training data (fine-tuned on ≤25% of the full dataset), RETFound showed a slight advantage (up to a maximum AUC increase of 0.03). However, these improvements were not statistically significant (all P≥0.2). These findings indicate the challenges foundational AI models face in adapting to diverse demographics, emphasizing the need for more diverse data in current foundation models and the importance of global collaboration on foundation model research.

Autor:innen und Institutionen

  • Key Laboratory for Biomechanics and Mechanobiology, Ministry of Education, Beijing Advanced Innovation Center for Biomedical Engineering, School of Biological Science and Medical Engineering, Beihang University, Beijing.
  • Centre for Medical Image Computing, University College London.
  • NIHR Biomedical Research Centre, Moorfields Eye Hospital NHS Foundation Trust, London.
  • Department of Medical Physics and Biomedical Engineering, University College London.
  • Institute of Ophthalmology, University College London.
  • Centre for Innovation and Precision Eye Health, Department of Ophthalmology, Yong Loo Lin School of Medicine, National University of Singapore.

ORCID-Profile: Zhuxin Xiong, Xiaofei Wang, Yukun Zhou, Pearse A Keane, Yih Chung Tham, Ya Xing Wang, Tien Yin Wong

Zitieren

APA

Xiong, Z., Wang, X., Zhou, Y., Keane, P. A., Tham, Y. C., Wang, Y. X., & Wong, T. Y. (2025). How Generalizable Are Foundation Models When Applied to Different Demographic Groups and Settings?. NEJM AI, 2(1). https://doi.org/10.1056/aics2400497

Vancouver

Xiong Z, Wang X, Zhou Y, Keane PA, Tham YC, Wang YX, et al. How Generalizable Are Foundation Models When Applied to Different Demographic Groups and Settings?. NEJM AI. 2025;2(1). doi:10.1056/aics2400497

BibTeX
@article{Xiong2025,
  author = {Xiong, Zhuxin and Wang, Xiaofei and Zhou, Yukun and Keane, Pearse A and Tham, Yih Chung and Wang, Ya Xing and Wong, Tien Yin},
  title = {How Generalizable Are Foundation Models When Applied to Different Demographic Groups and Settings?},
  journal = {NEJM AI},
  year = {2025},
  volume = {2},
  number = {1},
  doi = {10.1056/aics2400497}
}
Kostenlos starten →