Discovering Latent Representations for Patient State in Intensive Care Units
Abstract: The Intensive Care Unit (ICU) is playing an expanding role in acute hospital care, but the value of many treatments and interventions in the ICU is unproven, and high-quality data supporting or discouraging specific practices are sparse. Much prior work in clinical modeling has focused on building discriminating models to detect specific coded outcomes (e.g., hospital mortality) under specific settings, or understanding the predictive value of various types of clinical information without taking interventions into account. In this talk, we discuss our recent work on creating latent features for outcome prediction, and intervention prediction. In both settings, we use the publicly available MIMIC database to investigate whether we can predict mortality, and interventions in an empirically sound way. We demonstrate that latent representations of patient state are predictive of important clinical targets, and practically useful for creating data-driven clinical guidelines.