Personalized glucose forecasting for type 2 diabetes using data assimilation.

David J Albers,Henry Ginsberg,Lena Mamykina,Bruce Gluckman,George Hripcsak,Matthew Levine,Feilim Mac Gabhann

doi:10.1371/journal.pcbi.1005232

David J Albers, Henry Ginsberg + Show 5 more

Open Access

https://doi.org/10.1371/journal.pcbi.1005232

Copy DOI

Abstract

Type 2 diabetes leads to premature death and reduced quality of life for 8% of Americans. Nutrition management is critical to maintaining glycemic control, yet it is difficult to achieve due to the high individual differences in glycemic response to nutrition. Anticipating glycemic impact of different meals can be challenging not only for individuals with diabetes, but also for expert diabetes educators. Personalized computational models that can accurately forecast an impact of a given meal on an individual’s blood glucose levels can serve as the engine for a new generation of decision support tools for individuals with diabetes. However, to be useful in practice, these computational engines need to generate accurate forecasts based on limited datasets consistent with typical self-monitoring practices of individuals with type 2 diabetes. This paper uses three forecasting machines: (i) data assimilation, a technique borrowed from atmospheric physics and engineering that uses Bayesian modeling to infuse data with human knowledge represented in a mechanistic model, to generate real-time, personalized, adaptable glucose forecasts; (ii) model averaging of data assimilation output; and (iii) dynamical Gaussian process model regression. The proposed data assimilation machine, the primary focus of the paper, uses a modified dual unscented Kalman filter to estimate states and parameters, personalizing the mechanistic models. Model selection is used to make a personalized model selection for the individual and their measurement characteristics. The data assimilation forecasts are empirically evaluated against actual postprandial glucose measurements captured by individuals with type 2 diabetes, and against predictions generated by experienced diabetes educators after reviewing a set of historical nutritional records and glucose measurements for the same individual. The evaluation suggests that the data assimilation forecasts compare well with specific glucose measurements and match or exceed in accuracy expert forecasts. We conclude by examining ways to present predictions as forecast-derived range quantities and evaluate the comparative advantages of these ranges.

Highlights

One promise of data science is the application of elegant solutions to new and important problems
Our objectives here are to reconstruct the state of the endocrine system relevant to glucose-insulin dynamics regulation from the sporadic blood-glucose measurements clinically expected from a patient with diabetes
Because of the combination of the nonlinear, high-dimensional dynamics of the endocrine models and the sparsity of the data, we utilize a variant of an unscented Kalman filter (UKF) or Bayesian sigma point processor [7, 58,59,60,61] for Data assimilation (DA)

Summary

Introduction

One promise of data science is the application of elegant solutions to new and important problems. In this way, personal health care can be seen as a prediction challenge: Identifying the disease the patient has contracted—estimating the current state, forecasting the disease progression under different interventions—estimating the evolution of the state, and therapy—optimization of feedback to achieve target state. The ability to predict outcomes is important in selection of treatment. To be useful in the real world, such methods need to generate accurate and actionable predictions based on realistic datasets that are consistent with current standards for health monitoring

Objectives

Methods

Results

Discussion

Conclusion