20Aug 2024
Predicting HER2-low in Breast Cancer Using Weakly-Supervised Deep Learning

Predicting HER2-low in Breast Cancer Using Weakly-Supervised Deep Learning

This study introduces a weakly-supervised deep learning approach to predict HER2-low breast cancer directly from H&E-stained histopathological slides. By analyzing 1,437 images from 1,351 patients, the research demonstrates the potential of deep learning models to classify HER2 statuses, including HER2-negative, HER2-low, and HER2-high. The findings underscore the promise of AI-driven methods in enhancing HER2 assessment, offering a cost-effective, efficient alternative to traditional diagnostic tests.

  • #oncology

Like

Save

Share