Automated Dynamic Bayesian Networks for Predicting Acute Kidney Injury Before Onset

Several algorithms for learning the structure of dynamic Bayesian networks (DBNs) require an a priori ordering of variables, which influences the determined graph topology. However, it is often unclear how to determine this order if feature importance is unknown, especially as an exhaustive search is usually impractical. In this paper, we introduce Ranking Approaches for Unknown Structures (RAUS), an automated framework to systematically inform variable ordering and learn networks end-to-end. RAUS leverages existing statistical methods (Cramer’s V (CV), chi-squared test (Chi 2 ), and information gain (IG)) to compare variable ordering, resultant generated network topologies, and DBN performance. We evaluate RAUS on the task of predicting impending acute kidney injury (AKI) from clinical laboratory data. Longitudinal observations from 67,460 inpatients were collected from our electronic health record (EHR) as well as 13,109 patients from the MIMIC-III dataset. Kidney Disease Improving Global Outcomes (KDIGO) criteria were then applied to define AKI events. RAUS learned multiple DBNs in parallel to predict a future AKI event at different time points (i.e., 24-, 48-, 72-hours in advance of AKI). We also compared the results of the learned AKI prediction models and variable orderings to baseline techniques (logistic regression, random forests, extreme gradient boosting, multi-layer perceptron). The DBNs generated by RAUS achieved up to 87% area under the receiver operating characteristic curve (AUCROC) within 24-hours before AKI; and up to 78% AUCROC within 48-hours before AKI of any stage in a 7-day observation window. CV-DBN and IG-DBN ranked serum albumin and estimated glomerular filtration rate (eGFR) as the top two features within 48-hours before AKI for our site, resulting in higher overall performance. Insights from this automated framework can help efficiently implement and interpret DBNs for clinical decision support.