Cumulative Stay-time Representation for Electronic Health Records in Medical Event Time Prediction

TL;DR

This paper introduces a novel cumulative stay-time representation (CTR) for electronic health records, improving the prediction of disease development timing by directly modeling cumulative health conditions.

Contribution

The paper proposes a new CTR data representation and neural network-based construction that effectively captures cumulative health information for better disease time prediction.

Findings

CTR achieves high prediction accuracy on synthetic and real data.

Combining CTR with existing models improves their performance.

CTR is scalable to high-dimensional EHR data.

Abstract

We address the problem of predicting when a disease will develop, i.e., medical event time (MET), from a patient's electronic health record (EHR). The MET of non-communicable diseases like diabetes is highly correlated to cumulative health conditions, more specifically, how much time the patient spent with specific health conditions in the past. The common time-series representation is indirect in extracting such information from EHR because it focuses on detailed dependencies between values in successive observations, not cumulative information. We propose a novel data representation for EHR called cumulative stay-time representation (CTR), which directly models such cumulative health conditions. We derive a trainable construction of CTR based on neural networks that has the flexibility to fit the target data and scalability to handle high-dimensional EHR. Numerical experiments using synthetic and real-world datasets demonstrate that CTR alone achieves a high prediction performance, and it enhances the performance of existing models when combined with them.