Scholarlink

Vision-language models have demonstrated remarkable success in general medical image analysis, yet their application in pediatric imaging remains significantly underexplored. These models show limited performance on pediatric datasets, primarily due to domain gaps stemming from anatomical differences, lower radiation doses, and pediatric-specific diseases. To this end, we present the first pediatric vision-language pre-training framework, dubbed PedCLIP, trained on a comprehensive pediatric imaging dataset comprising 404,670 X-rays of pediatric patients across diverse anatomical regions. To address anatomical diversity, we introduce a Mixture of Body part Experts design, with each expert specializing in learning features from distinct anatomical regions. Experimental evaluation across eleven downstream tasks demonstrates that our model significantly outperforms current state-of-the-art vision-language models, achieving superior diagnostic accuracy in challenging pediatric conditions, including rare diseases such as pediatric inflammatory arthritis. Code is available: https://github.com/tadeephuy/PedCLIP

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

PedCLIP: A Vision-Language Model for Pediatric X-Rays with Mixture of Body Part Experts

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Ta Duc Huy,
  • Abin Shoby,
  • Sen Tran,
  • Yutong Xie,
  • Qi Chen,
  • Phi Le Nguyen,
  • Akshay Gole,
  • Lingqiao Liu,
  • Antonios Perperidis,
  • Mark Friswell,
  • Rebecca Linke,
  • Andrea Glynn,
  • Minh-Son To,
  • Anton van den Hengel,
  • Johan Verjans,
  • Zhibin Liao,
  • Minh Hieu Phan

摘要

Vision-language models have demonstrated remarkable success in general medical image analysis, yet their application in pediatric imaging remains significantly underexplored. These models show limited performance on pediatric datasets, primarily due to domain gaps stemming from anatomical differences, lower radiation doses, and pediatric-specific diseases. To this end, we present the first pediatric vision-language pre-training framework, dubbed PedCLIP, trained on a comprehensive pediatric imaging dataset comprising 404,670 X-rays of pediatric patients across diverse anatomical regions. To address anatomical diversity, we introduce a Mixture of Body part Experts design, with each expert specializing in learning features from distinct anatomical regions. Experimental evaluation across eleven downstream tasks demonstrates that our model significantly outperforms current state-of-the-art vision-language models, achieving superior diagnostic accuracy in challenging pediatric conditions, including rare diseases such as pediatric inflammatory arthritis. Code is available: https://github.com/tadeephuy/PedCLIP