Multivariate Time-Series Similarity Assessment via Unsupervised Representation Learning and Stratified Locality Sensitive Hashing: Application to Early Acute Hypotensive Episode Detection

Abstract

An accurate and timely prediction of clinically critical events in intensive care unit (ICU) is important for improving care and survival rate. Most of the existing approaches are based on the application of various classification methods on explicitly extracted statistical features from the vital signals of ICU patients, such as the mean, the standard deviation, and the skew. In this article, we propose to eliminate the high cost of engineering hand-crafted features from multivariate time-series of physiologic signals by learning its representation with a sequence-to-sequence autoencoder. We then propose to hash the learned multivariate time-series representations of labeled dataset to enable signal similarity assessment for the prediction of critical events. We evaluate this methodological framework to predict acute hypotensive episodes (AHE) on a large and diverse dataset of vital signal recordings extracted from eICU collaborative research database. Experiments demonstrate the ability of the presented framework in accurately predicting an upcoming AHE.