Deep Cardiovascular Disease Prediction with Risk Factors Powered Bi-attention

摘要

Cardiovascular disease (CVD) is one of the serious diseases endangering human life and health. Therefore, using the electronic medical record information to automatically predict CVD has important application value in intelligent auxiliary diagnosis and treatment, and is a hot issue in intelligent medical research. In recent years, attention mechanism utilized in natural language processing has focused heeds on a small part of the context and congregated it using fixed-size vectors, coupling attention in time, and/or often forming a unidirectional attention. In this paper, we propose a CVD risk factor powered bi-directional attention network named as RFPBiA, which is a multi-stage hierarchical architecture, fusing the information at different granularity levels, and employs the bi-directional attention to obtain the text representation of risk factors without early aggregation. The experimental results show that the proposed method can obviously improve the performance of CVD prediction, and the F-score reaches 0.9424, which is better than the existing related methods.