Journal of multidisciplinary healthcare · 2026
Multicenter Validation of an Integrated DCE-MRI Radiomics, Deep Learning, and S-II Model for Predicting Axillary Lymph Node Metastasis in Breast Cancer.
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.
Lu Xinxin, Wang Mengshen, Liu Xiaohua, Lyu Di
Аннотация
INTRODUCTION/BACKGROUND: This study developed and externally validated an integrated model combining clinical variables, systemic immune-inflammation index (SII), DCE-MRI radiomics score (RadScore), and deep learning score (DLScore) for preoperative prediction of axillary lymph node metastasis (ALNM) in breast cancer.
MATERIALS AND METHODS: A retrospective dual-center cohort included 212 patients in the training cohort and 121 patients in the external validation cohort. Clinical data, inflammatory indices, and DCE-MRI images were analyzed. Radiomic and deep learning features were reduced to RadScore and DLScore, ten logistic regression models were compared using ROC analysis, calibration, and decision curve analysis.
RESULTS: ALNM rates were comparable in training and validation cohorts (47.17% and 47.11%). The integrated model achieved the best discrimination, with AUC of 0.972 in the training cohort and 0.942 in the external validation cohort, and showed good calibration and superior net benefit across clinically relevant threshold. Combining local DCE-MRI phenotypes with systemic inflammatory status improved predictive performance and external generalizability compared with single-modality models. These findings support multimodal integration as a non-invasive adjunct for preoperative ALNM risk stratification.
CONCLUSION: The integrated model may assist individualized preoperative axillary assessment in breast cancer patients. Prospective multicenter validation, workflow evaluation, and further interpretability analysis are still required before routine clinical implementation.
MATERIALS AND METHODS: A retrospective dual-center cohort included 212 patients in the training cohort and 121 patients in the external validation cohort. Clinical data, inflammatory indices, and DCE-MRI images were analyzed. Radiomic and deep learning features were reduced to RadScore and DLScore, ten logistic regression models were compared using ROC analysis, calibration, and decision curve analysis.
RESULTS: ALNM rates were comparable in training and validation cohorts (47.17% and 47.11%). The integrated model achieved the best discrimination, with AUC of 0.972 in the training cohort and 0.942 in the external validation cohort, and showed good calibration and superior net benefit across clinically relevant threshold. Combining local DCE-MRI phenotypes with systemic inflammatory status improved predictive performance and external generalizability compared with single-modality models. These findings support multimodal integration as a non-invasive adjunct for preoperative ALNM risk stratification.
CONCLUSION: The integrated model may assist individualized preoperative axillary assessment in breast cancer patients. Prospective multicenter validation, workflow evaluation, and further interpretability analysis are still required before routine clinical implementation.