Early Prediction of Vital Signs Using Generative Boosting via LSTM Networks

Vital signs including heart rate, respiratory rate, body temperature and blood pressure, are critical in the clinical decision making process. Abnormal vital signs help to alert medical practitioners to potential health problems. Effective and long-range early prediction of vital signs may prevent adverse health outcomes and reduce cost. In this paper, we suggest a new approach called generative boosting, in order to effectively perform long-range early prediction of vital signs. Generative boosting consists of a generative model, to generate synthetic data for the next few time steps, and a predictive model, to directly make long-range predictions based on observed and generated data. We explore generative boosting via long short-term memory (LSTM) for both the predictive and generative models, leading to a scheme called generative LSTM (GLSTM). Our experiments indicate that GLSTM outperforms a diverse range of strong benchmark models, with and without generative boosting. As expected, the results also indicate that more accurately generated data leads to more accurate long-range predictions. In light of this, we use a mutual information based clustering algorithm to select a more representative dataset to train the generative model. This leads to significantly improved accuracy of the long-range prediction of high variation vital signs such as heart rate and systolic blood pressure. Overall, our best results indicate that the proposed method is able to predict heart rate and systolic blood pressure 20 minutes in advance, with a mean absolute percentage error of 7.41% and 6.17%, respectively.