GRU-D Characterizes Age-Specific Temporal Missingness in MIMIC-IV

TL;DR

This paper introduces GRU-D, a model that captures age-specific temporal missingness in clinical time series data, demonstrating its ability to reveal meaningful missingness patterns and improve predictive performance.

Contribution

The paper presents GRU-D, a novel gated recurrent unit model with decay mechanisms, for analyzing and interpreting temporal missingness in clinical data, specifically in MIMIC-IV.

Findings

GRU-D achieved 0.780 AUROC and 0.810 AUPRC in classification tasks.

Model parameters revealed important missingness patterns in vital signs.

GRU-D can uncover meaningful temporal missingness structures in clinical data.

Abstract

Temporal missingness, defined as unobserved patterns in time series, and its predictive potentials represent an emerging area in clinical machine learning. We trained a gated recurrent unit with decay mechanisms, called GRU-D, for a binary classification between elderly - and young patients. We extracted time series for 5 vital signs from MIMIC-IV as model inputs. GRU-D was evaluated with means of 0.780 AUROC and 0.810 AUPRC on bootstrapped data. Interpreting trained model parameters, we found differences in blood pressure missingness and respiratory rate missingness as important predictors learned by parameterized hidden gated units. We successfully showed how GRU-D can be used to reveal patterns in temporal missingness building the basis of novel research directions.