Predicting axillary lymph node metastasis in breast cancer patients using CNN-GCN on DCE-MRI: a multicenter study.
Authors
Affiliations (14)
Affiliations (14)
- Department of Medical Imaging, Peking University Shenzhen Hospital, No. 1120, Lianhua Road, Shenzhen, Guangdong, 518036, China.
- Department of Medical Imaging, Guilin Municipal Hospital of Traditional Chinese Medicine, Guilin, 541100, China.
- Department of Pathology, The Affiliated Hospital of Qingdao University, Qingdao, 266000, China.
- Department of Radiology, Yantai Yuhuangding Hospital, Qingdao University, No. 20 Yuhuangding East Street, Yantai, 266000, China.
- Big Data and Artificial Intelligence Laboratory, Yantai Yuhuangding Hospital, No. 20 Yuhuangding East Street, Yantai, 266000, China.
- Department of Radiology, Weifang Hospital of Traditional Chinese Medicine, Weifang, 261041, China.
- School of Computer Science and Technology, Shandong Technology and Business University, Yantai, 264005, China.
- Department of Medical Imaging, Affiliated Hospital of Jining Medical University, No. 89, Guhuai Road, Jining, Shandong, 272029, China. [email protected].
- Department of Medical Imaging, Peking University Shenzhen Hospital, No. 1120, Lianhua Road, Shenzhen, Guangdong, 518036, China. [email protected].
- Department of Radiology, Yantai Yuhuangding Hospital, Qingdao University, No. 20 Yuhuangding East Street, Yantai, 266000, China. [email protected].
- Big Data and Artificial Intelligence Laboratory, Yantai Yuhuangding Hospital, No. 20 Yuhuangding East Street, Yantai, 266000, China. [email protected].
- Shandong Provincial Key Medical and Health Laboratory of Intelligent Diagnosis and Treatment for Women's Diseases, Yantai Yuhuangding Hospital, Qingdao University, Yantai, 264000, China. [email protected].
- Big Data and Artificial Intelligence Laboratory, Yantai Yuhuangding Hospital, No. 20 Yuhuangding East Street, Yantai, 266000, China. [email protected].
- Shandong Provincial Key Medical and Health Laboratory of Intelligent Diagnosis and Treatment for Women's Diseases, Yantai Yuhuangding Hospital, Qingdao University, Yantai, 264000, China. [email protected].
Abstract
To develop a deep learning model combining a convolution neural network (CNN) and a graph convolution network (GCN) based on dynamic contrast-enhanced (DCE) MRI for predicting axillary lymph node (ALN) metastasis in breast cancer patients while also aiming to explore the underlying biological mechanism by using RNA sequencing (RNA-seq) data. We retrospectively collected DCE-MRI and clinical data from 1002 patients across four centers and one public dataset. Various CNN-GCN models were trained on tumor and ALN images and compared to radiomics models, the MSKCC model, and radiologists. RNA-seq data from 11 patients were used to explore associated biological pathways. Model performance was evaluated by accuracy, sensitivity, specificity, AUC, and DeLong test. Participants were divided into a training set (<i>n</i> = 742, mean age: 53 years ±10 [SD]), an internal test set (<i>n</i> = 83, 53 years ±10), external test set 1 (<i>n</i> = 110, 50 years ±9) and external test set 2 (<i>n</i> = 67, 54 years ±11). The optimal CNN-GCN model, HRNet-GCN_<sub>tumor+ALN</sub>, achieved an AUC of 0.873 in the internal test set outperforming the LR_<sub>tumor+ALN</sub> (AUC: 0.790) and MSKCC models (AUC: 0.726) (DeLong test, <i>p</i> < 0.05). Radiologists’ performance improved with HRNet-GCN_<sub>tumor+ALN</sub> (in both the external test set 1 and 2, <i>p</i> < 0.05). High-risk group associated with pathways such as ribosome, synapse organization, and muscle contraction. The proposed CNN-GCN fusion deep learning model showed good performance for preoperatively predicting ALN status in breast cancer patients. The online version contains supplementary material available at 10.1186/s12880-025-01988-4.