Research on Colorectal Cancer Prediction and Survival Analysis with Data Fusion Based on Deep Learning
Abstract— Colorectal cancer is a highly aggressive type of cancer. Accurate prognosis prediction of
colorectal cancer can help patients and doctors choose the best treatment and avoid unnecessary costs.
Nevertheless, most of previous work relied mostly on selected gene expression data to create a predictive
model by traditional machine learning methods to reduce dimensions and predict cancer. Such methods
cannot fully consider the correlation between samples to effectively extract feature information, and the cost
of calculation is huge. In this paper, we propose a multi-modal neural network which integrates GCN and
DNN to train multi-modal data for the prediction of colorectal cancer and its prognosis. The novelty of the
method lies in the design of our method’s architecture and the fusion of multi-dimensional data, which can
give full play to the performance of the neural network. The comprehensive performance evaluation results
show that the proposed method achieves a better performance than the widely used prediction methods with
single dimensional data and other existing approaches.
Index Terms— Colorectal cancer, Cancer prognosis prediction, Graph convolution network, Multimodal deep neural network, Multi-modal data.
Shiqi Li, Jun Zheng
School of Computer Science and Software Engineering East China Normal University, CHINA
Department of General Surgery of Changzheng Hospital the Second Military Medical University, CHINA
Cite: Shiqi Li, Jun Zheng, Shuxun Wei, "Research on Colorectal Cancer Prediction and Survival Analysis with Data Fusion Based on Deep Learning," Proceedings of 2019 the 9th International Workshop on Computer Science and Engineering, pp. 643-651, Hong Kong, 15-17 June, 2019.