Uncertainty-Aware Deep Ensembles for Reliable and Explainable Predictions of Clinical Time Series

TL;DR

This paper introduces an uncertainty-aware deep ensemble method for clinical time series analysis, enhancing the reliability and explainability of deep learning predictions in healthcare applications.

Contribution

It proposes a novel ensemble approach that incorporates uncertainty in relevance scores, improving trustworthiness and consistency in clinical time series explanations.

Findings

More accurate in identifying relevant time steps

More consistent across different initializations

Enhances trustworthiness of model explanations

Abstract

Deep learning-based support systems have demonstrated encouraging results in numerous clinical applications involving the processing of time series data. While such systems often are very accurate, they have no inherent mechanism for explaining what influenced the predictions, which is critical for clinical tasks. However, existing explainability techniques lack an important component for trustworthy and reliable decision support, namely a notion of uncertainty. In this paper, we address this lack of uncertainty by proposing a deep ensemble approach where a collection of DNNs are trained independently. A measure of uncertainty in the relevance scores is computed by taking the standard deviation across the relevance scores produced by each model in the ensemble, which in turn is used to make the explanations more reliable. The class activation mapping method is used to assign a relevance score for each time step in the time series. Results demonstrate that the proposed ensemble is more accurate in locating relevant time steps and is more consistent across random initializations, thus making the model more trustworthy. The proposed methodology paves the way for constructing trustworthy and dependable support systems for processing clinical time series for healthcare related tasks.