Mechanistic machine learning: how data assimilation leverages physiologic knowledge using Bayesian inference to forecast the future, infer the present, and phenotype. (12th October 2018)

• Record Type: Journal Article
• Authors: Albers, David J; Levine, Matthew E; Stuart, Andrew; Mamykina, Lena; Gluckman, Bruce; Hripcsak, George
• Is Part Of: Journal of the American Medical Informatics Association. Volume 25:Number 10(2018)
• Page Start: 1392
• Page End: 1401
• Publication Date: 2018-10-12
• Subjects: data assimilation -- Bayesian inverse methods -- state space models -- self-monitoring data -- machine learning -- data mining -- type 2 diabetes -- Gaussian process model -- glucose forecasting -- precision medicine; Medical informatics -- Periodicals; Information Services -- Periodicals; Medical Informatics -- Periodicals; Médecine -- Informatique -- Périodiques; Informatica; Geneeskunde; Informatique médicale; Computer network resources; Electronic journals; 610.285
• DOI: 10.1093/jamia/ocy106
• Languages: English
• ISSNs: 1067-5027

Abstract

Abstract: We introduce data assimilation as a computational method that uses machine learning to combine data with human knowledge in the form of mechanistic models in order to forecast future states, to impute missing data from the past by smoothing, and to infer measurable and unmeasurable quantities that represent clinically and scientifically important phenotypes. We demonstrate the advantages it affords in the context of type 2 diabetes by showing how data assimilation can be used to forecast future glucose values, to impute previously missing glucose values, and to infer type 2 diabetes phenotypes. At the heart of data assimilation is the mechanistic model, here an endocrine model. Such models can vary in complexity, contain testable hypotheses about important mechanics that govern the system (eg, nutrition's effect on glucose), and, as such, constrain the model space, allowing for accurate estimation using very little data.