Continuous-Time Probabilistic Models for Longitudinal Electronic Health Records

TL;DR

This paper introduces an unsupervised probabilistic model for analyzing irregularly sampled longitudinal EHR data, capturing nonlinear relationships over continuous time to improve risk prediction in precision medicine.

Contribution

It presents a novel continuous-time probabilistic framework that handles heterogeneity and irregular sampling in EHR data, enabling better modeling of variable relationships and risk assessment.

Findings

Model captures nonlinear relationships in EHR data.

Likelihood ratio maps predict depression risk.

Applicable to diverse irregularly sampled data.

Abstract

Analysis of longitudinal Electronic Health Record (EHR) data is an important goal for precision medicine. Difficulty in applying Machine Learning (ML) methods, either predictive or unsupervised, stems in part from the heterogeneity and irregular sampling of EHR data. We present an unsupervised probabilistic model that captures nonlinear relationships between variables over continuous-time. This method works with arbitrary sampling patterns and captures the joint probability distribution between variable measurements and the time intervals between them. Inference algorithms are derived that can be used to evaluate the likelihood of future using under a trained model. As an example, we consider data from the United States Veterans Health Administration (VHA) in the areas of diabetes and depression. Likelihood ratio maps are produced showing the likelihood of risk for moderate-severe vs minimal depression as measured by the Patient Health Questionnaire-9 (PHQ-9).