Learning Insulin-Glucose Dynamics in the Wild

Andrew C. Miller, Nicholas J. Foti, Emily Fox
; Proceedings of the 5th Machine Learning for Healthcare Conference, PMLR 126:172-197, 2020.

Abstract

We develop a new model of insulin-glucose dynamics for forecasting blood glucose in type 1 diabetics. We augment an existing biomedical model by introducing time-varying dynamics driven by a machine learning sequence model. Our model maintains a physiologically plausible inductive bias and clinically interpretable parameters — e.g., insulin sensitivity — while inheriting the flexibility of modern pattern recognition algorithms. Critical to modeling success are the flexible, but structured representations of subject variability with a sequence model. In contrast, less constrained models like the LSTM fail to provide reliable or physiologically plausible forecasts. We conduct an extensive empirical study. We show that allowing biomedical model dynamics to vary in time improves forecasting at long time horizons, up to six hours, and produces forecasts consistent with the physiological effects of insulin and carbohydrates.

Cite this Paper


BibTeX
@InProceedings{pmlr-v126-miller20a, title = {Learning Insulin-Glucose Dynamics in the Wild}, author = {Miller, Andrew C. and Foti, Nicholas J. and Fox, Emily}, booktitle = {Proceedings of the 5th Machine Learning for Healthcare Conference}, pages = {172--197}, year = {2020}, editor = {Finale Doshi-Velez and Jim Fackler and Ken Jung and David Kale and Rajesh Ranganath and Byron Wallace and Jenna Wiens}, volume = {126}, series = {Proceedings of Machine Learning Research}, address = {Virtual}, month = {07--08 Aug}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v126/miller20a/miller20a.pdf}, url = {http://proceedings.mlr.press/v126/miller20a.html}, abstract = {We develop a new model of insulin-glucose dynamics for forecasting blood glucose in type 1 diabetics. We augment an existing biomedical model by introducing time-varying dynamics driven by a machine learning sequence model. Our model maintains a physiologically plausible inductive bias and clinically interpretable parameters — e.g., insulin sensitivity — while inheriting the flexibility of modern pattern recognition algorithms. Critical to modeling success are the flexible, but structured representations of subject variability with a sequence model. In contrast, less constrained models like the LSTM fail to provide reliable or physiologically plausible forecasts. We conduct an extensive empirical study. We show that allowing biomedical model dynamics to vary in time improves forecasting at long time horizons, up to six hours, and produces forecasts consistent with the physiological effects of insulin and carbohydrates.} }
Endnote
%0 Conference Paper %T Learning Insulin-Glucose Dynamics in the Wild %A Andrew C. Miller %A Nicholas J. Foti %A Emily Fox %B Proceedings of the 5th Machine Learning for Healthcare Conference %C Proceedings of Machine Learning Research %D 2020 %E Finale Doshi-Velez %E Jim Fackler %E Ken Jung %E David Kale %E Rajesh Ranganath %E Byron Wallace %E Jenna Wiens %F pmlr-v126-miller20a %I PMLR %J Proceedings of Machine Learning Research %P 172--197 %U http://proceedings.mlr.press %V 126 %W PMLR %X We develop a new model of insulin-glucose dynamics for forecasting blood glucose in type 1 diabetics. We augment an existing biomedical model by introducing time-varying dynamics driven by a machine learning sequence model. Our model maintains a physiologically plausible inductive bias and clinically interpretable parameters — e.g., insulin sensitivity — while inheriting the flexibility of modern pattern recognition algorithms. Critical to modeling success are the flexible, but structured representations of subject variability with a sequence model. In contrast, less constrained models like the LSTM fail to provide reliable or physiologically plausible forecasts. We conduct an extensive empirical study. We show that allowing biomedical model dynamics to vary in time improves forecasting at long time horizons, up to six hours, and produces forecasts consistent with the physiological effects of insulin and carbohydrates.
APA
Miller, A.C., Foti, N.J. & Fox, E.. (2020). Learning Insulin-Glucose Dynamics in the Wild. Proceedings of the 5th Machine Learning for Healthcare Conference, in PMLR 126:172-197

Related Material