Global Deep Forecasting with Patient-Specific Pharmacokinetics

TL;DR

This paper introduces a hybrid global-local deep learning architecture with a pharmacokinetic encoder for personalized healthcare time series forecasting, significantly improving accuracy in blood glucose prediction over existing models.

Contribution

It presents a novel hybrid architecture and PK encoder that incorporate patient-specific treatment effects, enhancing forecasting accuracy in healthcare applications.

Findings

PK encoder outperforms baselines by up to 16.4% on simulated data and 4.9% on real data.

Hybrid global-local architecture surpasses patient-specific PK models by 15.8%.

Approach improves early detection of adverse health events.

Abstract

Forecasting healthcare time series data is vital for early detection of adverse outcomes and patient monitoring. However, it can be challenging in practice due to variable medication administration and unique pharmacokinetic (PK) properties of each patient. To address these challenges, we propose a novel hybrid global-local architecture and a PK encoder that informs deep learning models of patient-specific treatment effects. We showcase the efficacy of our approach in achieving significant accuracy gains in a blood glucose forecasting task using both realistically simulated and real-world data. Our PK encoder surpasses baselines by up to 16.4% on simulated data and 4.9% on real-world data for individual patients during critical events of severely high and low glucose levels. Furthermore, our proposed hybrid global-local architecture outperforms patient-specific PK models by 15.8%, on average.