Scholarlink

<p>Colorectal liver metastases (CRLM) remain a major cause of cancer-related mortality. Early, reliable detection on CT imaging is essential for curative treatment planning, yet the diagnostic performance of AI models often declines across scanners and institutions, limiting clinical generalizability. In this study, we developed and evaluated a multi-center foundation-model-based AI pipeline for patient-level classification and lesion-level detection of CRLM on contrast-enhanced CT. Using data from the EuCanImage consortium (<i>n</i> = 2437) and TCIA_CRLM (<i>n</i> = 197, all CRLM), we benchmarked several pretrained foundation models and identified UMedPT as the optimal encoder. The final model achieved an AUC of 0.89 and a sensitivity of 0.82 on the EuCanImage test set, with a sensitivity of 0.85 on the external TCIA cohort. Excluding the most uncertain 20% of cases improved AUC to 0.90 and balanced accuracy to 0.85. Decision-curve analysis indicated superior net benefit over “treat-all” and “treat-none” strategies for threshold probabilities between 0.35 and 0.75. The lesion-level detector identified 69.1% of lesions overall, increasing substantially with lesion size. Grad-CAM maps showed strong correspondence between attention regions and metastases in high-confidence predictions. These findings demonstrate that foundation-model-based pipelines enable robust, generalizable, and interpretable solutions for CRLM detection and classification in multi-center CT imaging.</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

Robust multicentre detection and classification of colorectal liver metastases on CT: application of foundation models

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月27日

查看全文
  • Shruti Atul Mali,
  • Zohaib Salahuddin,
  • Yumeng Zhang,
  • Andre Aichert,
  • Xian Zhong,
  • Henry C. Woodruff,
  • Maciej Bobowicz,
  • Katrine Riklund,
  • Juozas Kupčinskas,
  • Lorenzo Faggioni,
  • Roberto Francischello,
  • Razvan L. Miclea,
  • Philippe Lambin,
  • Karim Lekadir,
  • Lidia Garrucho,
  • Henry Woodruff,
  • Martijn Starmans,
  • Josep Lluis Gelpi,
  • Salvador Capella-Gutierrez,
  • Jordi Rambla,
  • Fred Prior,
  • Melanie Goisauf,
  • Pilar Nicolas,
  • Davide Zaccagnini,
  • Pär Kragsterman,
  • Samuele Ghezzo,
  • Jeroen Cant,
  • Tobias Heimann,
  • Peter Gordebeke,
  • Emanuele Neri,
  • Jane Smith,
  • Lorenzo Faggioni,
  • Jordi Rimola,
  • Xavier Bargalló,
  • Javier del Riego

摘要

Colorectal liver metastases (CRLM) remain a major cause of cancer-related mortality. Early, reliable detection on CT imaging is essential for curative treatment planning, yet the diagnostic performance of AI models often declines across scanners and institutions, limiting clinical generalizability. In this study, we developed and evaluated a multi-center foundation-model-based AI pipeline for patient-level classification and lesion-level detection of CRLM on contrast-enhanced CT. Using data from the EuCanImage consortium (n = 2437) and TCIA_CRLM (n = 197, all CRLM), we benchmarked several pretrained foundation models and identified UMedPT as the optimal encoder. The final model achieved an AUC of 0.89 and a sensitivity of 0.82 on the EuCanImage test set, with a sensitivity of 0.85 on the external TCIA cohort. Excluding the most uncertain 20% of cases improved AUC to 0.90 and balanced accuracy to 0.85. Decision-curve analysis indicated superior net benefit over “treat-all” and “treat-none” strategies for threshold probabilities between 0.35 and 0.75. The lesion-level detector identified 69.1% of lesions overall, increasing substantially with lesion size. Grad-CAM maps showed strong correspondence between attention regions and metastases in high-confidence predictions. These findings demonstrate that foundation-model-based pipelines enable robust, generalizable, and interpretable solutions for CRLM detection and classification in multi-center CT imaging.