MGNN: A Multimodal Graph Neural Network for Predicting the Survival of Cancer Patients

• DOI: 10.1145/3397271.3401214
• Publication Date: 2020-07-25T07:50:08Z
• Authors: Jianliang Gao; Tengfei Lyu; Fan Xiong; Jianxin Wang; Weimao Ke; Zhao Li
• Keywords: 0202 electrical engineering, electronic engineering, information engineering; 02 engineering and technology; 3. Good health

Predicting the survival of cancer patients holds significant meaning for public health, and has attracted increasing attention in medical information communities. In this study, we propose a novel framework prediction named Multimodal Graph Neural Network (MGNN), which explores features real-world multimodual data such as gene expression, copy number alteration clinical unified framework. order to explore inherent relation, first construct bipartite graphs between multimodal data. Subsequently, graph neural network is adopted obtain embedding each patient on different graphs. Finally, fusion layer designed fuse from modal The output our method classification short term or long patient. Experimental results one breast dataset demonstrate that MGNN outperforms all baselines. Furthermore, test trained model lung dataset, experimental verify strong robust by comparing with state-of-the-art methods.