AUG 11, 2026 · JOURNAL ARTICLE
A deep learning model utilizing H&E-stained images for predicting HER2-low expression in breast cancer
Diagnostic Pathology
Single-centre retrospective study with moderate diagnostic accuracy (AUC ~0.61) on a surrogate imaging endpoint; model performance is modest and requires external validation before clinical use.
Reported
Validation set AUC0.613 ± 0.118
Test set AUC0.608 ± 0.104
Sample size776