Hierarchical Target-Attentive Diagnosis Prediction in Heterogeneous Information Networks

TL;DR

HTAD is a new model that uses target-aware hierarchical attention on heterogeneous EHR data to improve diagnosis prediction accuracy, incorporate non-categorical data, and enhance interpretability.

Contribution

The paper introduces HTAD, a novel target-aware hierarchical attention mechanism for diagnosis prediction in heterogeneous EHR networks, with improved performance and interpretability.

Findings

Significantly outperforms existing methods on benchmark datasets.

Incorporating non-categorical data improves prediction accuracy.

Predictions are easily interpretable.

Abstract

We introduce HTAD, a novel model for diagnosis prediction using Electronic Health Records (EHR) represented as Heterogeneous Information Networks. Recent studies on modeling EHR have shown success in automatically learning representations of the clinical records in order to avoid the need for manual feature selection. However, these representations are often learned and aggregated without specificity for the different possible targets being predicted. Our model introduces a target-aware hierarchical attention mechanism that allows it to learn to attend to the most important clinical records when aggregating their representations for prediction of a diagnosis. We evaluate our model using a publicly available benchmark dataset and demonstrate that the use of target-aware attention significantly improves performance compared to the current state of the art. Additionally, we propose a method for incorporating non-categorical data into our predictions and demonstrate that this technique leads to further performance improvements. Lastly, we demonstrate that the predictions made by our proposed model are easily interpretable.