Scholarlink

<p>Medical foundation models, pre-trained on large-scale unlabelled data, show strong performance and data efficiency when adapted to various clinically relevant applications. However, how pre-training data shape the generalisability and fairness of these models remains unexplored. Here we address this using two cohorts from Moorfields Eye Hospital (UK) and the Shanghai Diabetes Prevention Program (China), each containing 904,170 fundus photographs for model pre-training. Using identical pipelines, we train parallel foundation models using individual cohort and evaluate them on downstream tasks with publicly available datasets and held-out data from each site. The parallel models show competitive performance to data that differ substantially from their pre-training data. Nevertheless, we observe fairness gaps over age subgroups, whereas sex and ethnicity show minimal impact. These results demonstrate the good generalisability of retinal foundation models and indicate that pre-training demographic attributes shape fairness differently, highlighting the importance of domain-specific, fine-grained data curation for efficient foundation model development.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Understanding pre-training data effects in retinal foundation models using two large fundus cohorts

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年2月28日

查看全文
  • Yukun Zhou,
  • Zheyuan Wang,
  • Yilan Wu,
  • Ariel Yuhan Ong,
  • Siegfried K. Wagner,
  • Eden Ruffell,
  • Mark A. Chia,
  • Zhouyu Guan,
  • Lie Ju,
  • Justin Engelmann,
  • David A. Merle,
  • Tingyao Li,
  • Jia Shu,
  • Paul Nderitu,
  • Ke Zou,
  • Jocelyn Hui Lin Goh,
  • Qingshan Hou,
  • Xiaoxuan Liu,
  • Yaxing Wang,
  • Yih Chung Tham,
  • Andre Altmann,
  • Carol Y. Cheung,
  • Daniel C. Alexander,
  • Eric J. Topol,
  • Alastair K. Denniston,
  • Tien Yin Wong,
  • Bin Sheng,
  • Pearse A. Keane

摘要

Medical foundation models, pre-trained on large-scale unlabelled data, show strong performance and data efficiency when adapted to various clinically relevant applications. However, how pre-training data shape the generalisability and fairness of these models remains unexplored. Here we address this using two cohorts from Moorfields Eye Hospital (UK) and the Shanghai Diabetes Prevention Program (China), each containing 904,170 fundus photographs for model pre-training. Using identical pipelines, we train parallel foundation models using individual cohort and evaluate them on downstream tasks with publicly available datasets and held-out data from each site. The parallel models show competitive performance to data that differ substantially from their pre-training data. Nevertheless, we observe fairness gaps over age subgroups, whereas sex and ethnicity show minimal impact. These results demonstrate the good generalisability of retinal foundation models and indicate that pre-training demographic attributes shape fairness differently, highlighting the importance of domain-specific, fine-grained data curation for efficient foundation model development.