Disease Risk Prediction from Clinical Texts

Abstract

Huge amounts of healthcare data collected can be leveraged to predict the prognosis of a particular patient. Risk prediction models can be used to estimate the probability of either having (diagnostic model) or developing a particular disease or outcome (prognostic model). In clinical practice, these models are used to inform patients and guide therapeutic management. As of 2015, cardiovascular diseases contribute to 32% of global deaths. A number of models like Framingham, SCORE, and QRISK are used for risk prediction of these diseases. In this proposed method we are focusing on predicting the risk of having coronary artery disease. Using unstructured data such as clinical texts allows the inclusion of a variety of information like patient history and lifestyle. These unstructured data if used along with structured clinical reports can contribute to a better prediction.