Gland surgery · 2026
Ultrasound radiomics for preoperative prediction of high axillary nodal burden in node-positive breast cancer: comparison of intratumoral and peritumoral features with external validation.
СтатьяНазвание и аннотация: Перевод в очереди
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.
Ye Hao, Cai Shaopan, Ye Jiaxuan, Li Weipeng, Xu Zhirong, Cai Danhong
Аннотация
BACKGROUND: Accurate preoperative assessment of axillary nodal burden is important for individualized axillary management, but conventional ultrasound and needle biopsy have limited ability to characterize the overall nodal burden. This study aimed to develop and externally validate ultrasound radiomics models for preoperative prediction of high axillary nodal burden in node-positive breast cancer and to compare the value of intratumoral and peritumoral features.
METHODS: This retrospective dual-center study included 404 patients with pathologically confirmed breast cancer and axillary lymph node metastasis, including 294 patients in the training cohort and 110 in the external validation cohort. Axillary nodal burden was categorized as low burden (1-2 metastatic lymph nodes) or high burden (≥3 metastatic lymph nodes) according to postoperative pathology. Intratumoral and 3-mm peritumoral ultrasound radiomics features were extracted. To avoid information leakage, the training and validation cohorts were separated before preprocessing and model development. Six prespecified models were developed in the training cohort and independently evaluated in the external validation cohort.
RESULTS: Cortical thickening, hilum status, and ultrasound-reported axillary lymph node status were independent clinical predictors of high nodal burden. In external validation, all models showed limited discrimination, with areas under the receiver operating characteristic curve (AUCs) ranging from 0.585 to 0.621. The intratumoral radiomics model achieved the numerically highest validation AUC of 0.621, with a sensitivity of 0.762 and specificity of 0.382, followed by the combined models (AUC =0.613), clinical model (AUC =0.597), and peritumoral model (AUC =0.585). Bootstrap comparisons showed no significant differences between models. Calibration performance was suboptimal in the external validation cohort.
CONCLUSIONS: Under a leakage-free external validation framework, ultrasound radiomics showed limited but exploratory value for predicting high axillary nodal burden in node-positive breast cancer. The current models should not be used as standalone tools for axillary management.
METHODS: This retrospective dual-center study included 404 patients with pathologically confirmed breast cancer and axillary lymph node metastasis, including 294 patients in the training cohort and 110 in the external validation cohort. Axillary nodal burden was categorized as low burden (1-2 metastatic lymph nodes) or high burden (≥3 metastatic lymph nodes) according to postoperative pathology. Intratumoral and 3-mm peritumoral ultrasound radiomics features were extracted. To avoid information leakage, the training and validation cohorts were separated before preprocessing and model development. Six prespecified models were developed in the training cohort and independently evaluated in the external validation cohort.
RESULTS: Cortical thickening, hilum status, and ultrasound-reported axillary lymph node status were independent clinical predictors of high nodal burden. In external validation, all models showed limited discrimination, with areas under the receiver operating characteristic curve (AUCs) ranging from 0.585 to 0.621. The intratumoral radiomics model achieved the numerically highest validation AUC of 0.621, with a sensitivity of 0.762 and specificity of 0.382, followed by the combined models (AUC =0.613), clinical model (AUC =0.597), and peritumoral model (AUC =0.585). Bootstrap comparisons showed no significant differences between models. Calibration performance was suboptimal in the external validation cohort.
CONCLUSIONS: Under a leakage-free external validation framework, ultrasound radiomics showed limited but exploratory value for predicting high axillary nodal burden in node-positive breast cancer. The current models should not be used as standalone tools for axillary management.