Risk factor identification for incident heart failure using neural network distillation and variable selection

TL;DR

This study introduces methods to interpret deep learning models for heart failure risk prediction, identifying relevant risk factors at both population and individual levels to enhance clinical understanding and hypothesis generation.

Contribution

The paper presents novel techniques of model distillation and variable selection to interpret deep neural networks trained on electronic health records for heart failure risk factors.

Findings

Identified 598 diseases associated with heart failure at the population level.

Highlighted less known links worth further investigation.

Developed an individual-level interpretation approach for personalized risk assessment.

Abstract

Recent evidence shows that deep learning models trained on electronic health records from millions of patients can deliver substantially more accurate predictions of risk compared to their statistical counterparts. While this provides an important opportunity for improving clinical decision-making, the lack of interpretability is a major barrier to the incorporation of these black-box models in routine care, limiting their trustworthiness and preventing further hypothesis-testing investigations. In this study, we propose two methods, namely, model distillation and variable selection, to untangle hidden patterns learned by an established deep learning model (BEHRT) for risk association identification. Due to the clinical importance and diversity of heart failure as a phenotype, it was used to showcase the merits of the proposed methods. A cohort with 788,880 (8.3% incident heart failure) patients was considered for the study. Model distillation identified 598 and 379 diseases that were associated and dissociated with heart failure at the population level, respectively. While the associations were broadly consistent with prior knowledge, our method also highlighted several less appreciated links that are worth further investigation. In addition to these important population-level insights, we developed an approach to individual-level interpretation to take account of varying manifestation of heart failure in clinical practice. This was achieved through variable selection by detecting a minimal set of encounters that can maximally preserve the accuracy of prediction for individuals. Our proposed work provides a discovery-enabling tool to identify risk factors in both population and individual levels from a data-driven perspective. This helps to generate new hypotheses and guides further investigations on causal links.