Frontiers in public health · 2026
Automatic grading of HER2 immunohistochemical whole-slide images based on a multi-scale hierarchical graph neural network.
СтатьяНазвание и аннотация: Перевод пока не подготовлен
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Jiang Yingjie, Sun Wenhao, Ye Chao, Xu Chun, Xue Jiewen, Wang Yiliang, Yan Shiju, Li Yiming, Bai Chenguang
Аннотация
OBJECTIVE: The HER2 immunohistochemical (IHC) score of breast cancer is a critical precursor for determining eligibility for anti-HER2 targeted therapy, as the final HER2 status (positive/negative/low) is determined by IHC results combined with FISH confirmation for equivocal 2 + cases according to ASCO/CAP guidelines. However, most existing artificial intelligence methods separate the detection of invasive cancer regions from HER2 grading and fail to adequately model the spatial heterogeneity of whole-slide tumors. This study proposes an end-to-end intelligent HER2 grading framework that integrates automated detection of invasive cancer regions with hierarchical graph reasoning, enabling automated and precise evaluation of HER2 IHC in breast cancer.
METHOD: A coarse-to-fine automated HER2 grading framework was developed. First, invasive tumor regions were automatically identified through supervised multi-scale tissue segmentation. Within the detected regions, patch-level HER2 grading integrates Delaunay cell image features with self-supervised visual representations, while WSI-level spatial representations model tumor heterogeneity to achieve fully automated HER2 grading across whole-slide images. The framework was trained using data from two scanning platforms and externally validated on 82 independent WSI samples from a third platform.
RESULTS: In independent external validation, the Dice coefficient for invasive region detection reached 0.89 (on 50 WSI images); the accuracy rate of HER2 grading at the patch level was 96.49% (95% CI, 96.08-96.87%) (7,924 patches), while that at the WSI level reached 98.0% (95% CI, 91.5-99.9%) (82 WSI images). The quadratic-weighted Cohen's κ was 0.97 (95% CI, 0.94-1.00), and all misclassifications at the WSI level were confined to the distinction between HER2 1 + and 2 +.
CONCLUSION: This framework enables fully automated HER2 IHC scoring and interpretation under multicenter and cross-scanner conditions, providing a computational tool for standardized IHC assessment, while final HER2 status determination for 2 + cases still requires FISH validation.
METHOD: A coarse-to-fine automated HER2 grading framework was developed. First, invasive tumor regions were automatically identified through supervised multi-scale tissue segmentation. Within the detected regions, patch-level HER2 grading integrates Delaunay cell image features with self-supervised visual representations, while WSI-level spatial representations model tumor heterogeneity to achieve fully automated HER2 grading across whole-slide images. The framework was trained using data from two scanning platforms and externally validated on 82 independent WSI samples from a third platform.
RESULTS: In independent external validation, the Dice coefficient for invasive region detection reached 0.89 (on 50 WSI images); the accuracy rate of HER2 grading at the patch level was 96.49% (95% CI, 96.08-96.87%) (7,924 patches), while that at the WSI level reached 98.0% (95% CI, 91.5-99.9%) (82 WSI images). The quadratic-weighted Cohen's κ was 0.97 (95% CI, 0.94-1.00), and all misclassifications at the WSI level were confined to the distinction between HER2 1 + and 2 +.
CONCLUSION: This framework enables fully automated HER2 IHC scoring and interpretation under multicenter and cross-scanner conditions, providing a computational tool for standardized IHC assessment, while final HER2 status determination for 2 + cases still requires FISH validation.