Continuous Temporal Learning of Probability Distributions via Neural ODEs with Applications in Continuous Glucose Monitoring Data

TL;DR

This paper presents a neural ODE-based probabilistic model for tracking the evolution of distributions over time, demonstrated on continuous glucose monitoring data to improve longitudinal analysis in clinical trials.

Contribution

It introduces a novel Gaussian mixture model combined with Neural ODEs for interpretable, efficient, and subtle detection of distribution shifts in time-dependent data.

Findings

Effective in detecting subtle distribution changes.

Enables rigorous longitudinal comparisons in clinical trials.

Provides detailed characterizations beyond traditional summaries.

Abstract

Modeling the dynamics of probability distributions from time-dependent data samples is a fundamental problem in many fields, including digital health. The goal is to analyze how the distribution of a biomarker, such as glucose, changes over time and how these changes may reflect the progression of chronic diseases such as diabetes. We introduce a probabilistic model based on a Gaussian mixture that captures the evolution of a continuous-time stochastic process. Our approach combines a nonparametric estimate of the distribution, obtained with Maximum Mean Discrepancy (MMD), and a Neural Ordinary Differential Equation (Neural ODE) that governs the temporal evolution of the mixture weights. The model is highly interpretable, detects subtle distribution shifts, and remains computationally efficient. We illustrate the broad utility of our approach in a 26-week clinical trial that treats all continuous glucose monitoring (CGM) time series as the primary outcome. This method enables rigorous longitudinal comparisons between the treatment and control arms and yields characterizations that conventional summary-based clinical trials analytical methods typically do not capture.