DuETT: Dual Event Time Transformer for Electronic Health Records

TL;DR

DuETT is a novel Transformer architecture designed for electronic health records, effectively capturing time and event type relationships, improving representation quality, and outperforming existing models on multiple clinical prediction tasks.

Contribution

This paper introduces DuETT, a dual-attention Transformer that models both time and event type dimensions in EHR data, with improved efficiency and performance.

Findings

Outperforms state-of-the-art models on MIMIC-IV and PhysioNet-2012 datasets.

Reduces computational complexity enabling larger, deeper models.

Effectively captures structured relationships in sparse, irregular EHR data.

Abstract

Electronic health records (EHRs) recorded in hospital settings typically contain a wide range of numeric time series data that is characterized by high sparsity and irregular observations. Effective modelling for such data must exploit its time series nature, the semantic relationship between different types of observations, and information in the sparsity structure of the data. Self-supervised Transformers have shown outstanding performance in a variety of structured tasks in NLP and computer vision. But multivariate time series data contains structured relationships over two dimensions: time and recorded event type, and straightforward applications of Transformers to time series data do not leverage this distinct structure. The quadratic scaling of self-attention layers can also significantly limit the input sequence length without appropriate input engineering. We introduce the DuETT architecture, an extension of Transformers designed to attend over both time and event type dimensions, yielding robust representations from EHR data. DuETT uses an aggregated input where sparse time series are transformed into a regular sequence with fixed length; this lowers the computational complexity relative to previous EHR Transformer models and, more importantly, enables the use of larger and deeper neural networks. When trained with self-supervised prediction tasks, that provide rich and informative signals for model pre-training, our model outperforms state-of-the-art deep learning models on multiple downstream tasks from the MIMIC-IV and PhysioNet-2012 EHR datasets.