Abstract：Neoadjuvant chemotherapy (NACT) is widely used in the treatment of breast cancer patients. Studies have shown that patients with a good chemotherapy response, particularly those achieving pathological complete response (pCR), have significantly improved survival rates. However, due to individual differences, some patients may experience poor responses or disease progression, therefore, early prediction of the response to NACT is crucial. Traditional convolutional neural networks (CNNs) have limitations in handling the spatial correlations and heterogeneity of tumor regions, which affects prediction accuracy. To address this, we proposed a graph network method combined with a deep graph attention network (DeepGAT) model and memory layers. We used the SLIC superpixel algorithm to segment the breast region and extract radiomic features from these superpixels. The superpixels were set as feature nodes, and edges were defined based on the weighted Euclidean distance and Pearson correlation coefficient of node features. By combining nodes and edges, we built a complete graph network. Our model included a DeepGAT module that dynamically weights node features using an attention mechanism to capture both local and global information. Additionally, we employed a memory pooling module based on clustering methods, which enhanced classification performance by capturing critical information and long-term dependencies. We used 214 samples, with 149 for training and 65 for testing. We also investigated the impact of graph sparsity coefficients on response to NACT performance, which showed that the best performance was achieved at a sparsity coefficient of 0.09. At this sparsity level, our model achieved AUCs of 0.822±0.027 and 0.811±0.041, significantly outperforming traditional GCN (AUC=0.726±0.045) and GAT (AUC=0.803±0.037) models. These results demonstrated the superior accuracy of our model in predicting response to NACT, providing a solid foundation for future research.
徐伟朝, 范明, 费思翔, 厉力华. 基于记忆深度图注意力网络的乳腺癌新辅助化疗疗效预测[J]. 中国生物医学工程学报, 2025, 44(4): 424-434.
Xu Weichao, Fan Ming, Fei Sixiang, Li Lihua. Deep Graph Attention Networks with Memory for Prediction of Response to Neoadjuvant Chemotherapy in Breast Cancer. Chinese Journal of Biomedical Engineering, 2025, 44(4): 424-434.
[1] Sung H, Ferlay J, Siegel R L, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries[J]. CA: a Cancer Journal for Clinicians, 2021, 71(3): 209-249.
[2] Burstein HJ, Curigliano G, Thürlimann B, et al. Customizing local and systemic therapies for women with early breast cancer: The St. Gallen international consensus guidelines for treatment of early breast cancer 2021 [J]. Annals of Oncology, 2021, 32(10): 1216-1235.
[3] Barrio AV, Montagna G, Mamtani A, et al. Nodal recurrence in patients with node-positive breast cancer treated with sentinel node biopsy alone after neoadjuvant chemotherapy—a rare event [J]. JAMA Oncology, 2021, 7(12): 1851-1855.
[4] Murchison S, Truong P. Locoregional therapy in breast cancer patients treated with neoadjuvant chemotherapy [J]. Expert Review of Anticancer Therapy, 2021,21(8):865-875.
[5] Conforti F, Pala L, Sala I, et al. Evaluation of pathological complete response as surrogate endpoint in neoadjuvant randomised clinical trials of early stage breast cancer: systematic review and meta-analysis[J]. The British Medical Journal (Clinical research ed.), 2021, 375: e066381.
[6] Cheng Rong, Li Jing, Ji Li, et al. Comparison of the diagnostic efficacy between ultrasound elastography and magnetic resonance imaging for breast masses[J]. Experimental and Therapeutic Medicine, 2018, 15(3): 2519-2524.
[7] Li Panli, Wang Xiuying, Xu Chongrui, et al. F-FDG PET/CT Radiomic predictors of pathologic complete response (pcr) to neoadjuvant chemotherapy in breast cancer patients [J]. European Journal of Nuclear Medicine and Molecular Imaging, 2020, 47(5): 1116-1126.
[8] Hosseini MP, Hosseini A, Ahi K. A review on machine learning for eeg signal processing in bioengineering [J]. IEEE Reviews in Biomedical Engineering, 2021, 14: 204-218.
[9] Zhang Shizhao, Li Xuelong, Zong Ming, et al. Efficient kNN classification with different numbers of nearest neighbors [J]. IEEE Transactions on Neural Networks and Learning Systems, 2018, 29(5): 1774-1785.
[10] Conti A, Duggento A, Indovina I, et al. Radiomics in breast cancer classification and prediction [J]. Seminars in Cancer Biology, 2021, 72: 238-250.
[11] Wagner J, Rapsomaniki M A, Chevrier S, et al. A single-cell atlas of the tumor and immune ecosystem of human breast cancer [J]. Cell, 2019, 177(5): 1330-1345.
[12] Satake H, Ishigaki S, Ito R, Naganawa S. Radiomics in breast mri: Current progress toward clinical application in the era of artificial intelligence [J]. Radiologia Medica, 2022, 127(1): 39-56.
[13] Abrol A, Fu Z, Salman M, et al. Deep learning encodes robust discriminative neuroimaging representations to outperform standard machine learning [J]. Nature Communications, 2021, 12(1): 353.
[14] Xiao Xiaolin, Zhou Yicong, Gong Yuejiao. Content-adaptive superpixel segmentation [J]. IEEE Transactions on Image Processing, 2018, 27(6): 2883-2896.
[15] Zhao C, Zhu W, Feng S. Superpixel guided deformable convolution network for hyperspectral image classification [J]. IEEE Transactions on Image Processing, 2022, 31: 3838-3851.
[16] Zhao Xue, Bai Jingwen, Guo Qiu, et al. Clinical applications of deep learning in breast MRI [J]. Biochimica et Biophysica Acta-Reviews on Cancer, 2023, 1878(2): 188864.
[17] Soffer S, Ben-Cohen A, Shimon O, et al. Convolutional neural networks for radiologic images: a radiologist′s guide [J]. Radiology, 2019, 290(3): 590-606.
[18] Wu Zonghan, Pan Shirui, Chen Fengwen, et al. A comprehensive survey on graph neural networks [J]. IEEE Transactions on Neural Networks and Learning Systems, 2021, 32(1): 4-24.
[19] Jiang Bo, Wang Leiling, Cheng Jian, et al. GPENs: graph data learning with graph propagation-embedding networks [J]. IEEE Transactions on Neural Networks and Learning Systems, 2023, 34(8): 3925-3938.
[20] Xia Tianyi, Zhao Ben, Li Binrong, et al. MRI-based radiomics and deep learning in biological characteristics and prognosis of hepatocellular carcinoma [J]. Magnetic Resonance Imaging, 2024, 59(3): 767-783.
[21] Chen Jie, Chen Shouzhen, Bai Mingyuan, et al. Graph decoupling attention Markov networks for semisupervised graph node classification [J]. IEEE Transactions on Neural Networks and Learning Systems, 2023, 34(12): 9859-9873.
[22] Shi Yong, Quan Pei, Xiao Yang, et al. Graph influence network [J]. IEEE Transactions on Cybernetics, 2023, 53(10): 6146-6159.
[23] Gubern-Mérida A, Martí R, Melendez J, et al. Automated localization of breast cancer in DCE-MRI [J]. Medical Image Analysis, 2015, 20(1): 265-274.
[24] Riner AN, Girma S, Vudatha V, et al. Eligibility criteria perpetuate disparities in enrollment and participation of black patients in pancreatic cancer clinical trials [J]. Clinical Oncology, 2022, 40(20): 2193-2202.
[25] Tan WCC, Nerurkar SN, Cai HY, et al. Overview of multiplex immunohistochemistry/immunofluorescence techniques in the era of cancer immunotherapy [J]. Cancer Communications (London), 2020, 40(4): 135-153.
[26] Liang Xuzhi, Li Jijia. Optimize statistical analysis via propensity score matching and repeated-measures analysis of variance [J]. JACC: Cardiovascular Interventions, 2023, 16(3): 361-362.
[27] Mishro PK, Agrawal S, Panda R, et al. A novel type-2 fuzzy C-means clustering for brain MR image segmentation [J]. IEEE Transactions on Cybernetics, 2021, 51(8): 3901-3912.
[28] Feng Shuanglang, Zhao Heming, Shi Fei, et al. CPFNet: context pyramid fusion network for medical image segmentation [J]. IEEE Transactions on Medical Imaging, 2020, 39(10): 3008-3018.
[29] Chen Zhihong, Yao Lisha, Liu Yue, et al. Deep learning-aided 3D proxy-bridged region-growing framework for multi-organ segmentation [J]. Scientific Reports, 2024, 14(1): 9784.
[30] de Oliveira ES, Lião LM, Silva AK, et al. A t-test ranking-based discriminant analysis for classification of free-range and barn-raised broiler chickens by H NMR spectroscopy [J]. Food Chemistry, 2023, 399: 134004.
[31] Terlep TA, Bell MR, Talavage TM, et al. Euclidean distance approximation from substitute product images [J]. IEEE Transactions on Image Processing, 2022, 31: 125-137.
[32] O'Brien JJ, Gunawardena HP, Qaqish BF, et al. Row versus column correlations: avoiding the ecological fallacy in RNA/protein expression studies [J]. Briefings in Bioinformatics, 2018, 19(5): 946-953.
[33] Lingjærde C, Richardson S. Stab JGL: a stability approach to sparsity and similarity selection in multiple-network reconstruction [J]. Bioinformatics Advances, 2023, 3(1): vbad185.
[34] Wang Jinxian, Liu Xuejun, Shen Siyuan, et al. DeepDDS: Deep graph neural network with attention mechanism to predict synergistic drug combinations [J]. Briefings in Bioinformatics, 2022, 23(1): bbab390.
[35] Chen Zhaoliang, Wu Zhihao, Lin Zhenghong, et al. AGNN: alternating graph-regularized neural networks to alleviate over-smoothing [J]. IEEE Transactions on Neural Networks and Learning Systems, 2024, 35(10): 13764-13776.
[36] Jahmunah V, Ng EYK, Tan RS, et al. Explainable detection of myocardial infarction using deep learning models with Grad-CAM technique on ECG signals [J]. Computers in Biology and Medicine, 2022, 146: 105550.
[37] Li B, Nabavi S. A multimodal graph neural network framework for cancer molecular subtype classification [J]. BMC Bioinformatics, 2024, 25(1): 27.
[38] Qu Yuhong, Zhu Haitao, Kun Cao, et al. Prediction of pathological complete response to neoadjuvant chemotherapy in breast cancer using a deep learning (DL) method[J]. Thoracic Cancer, 2020, 11(3): 651-658.
