The purpose of this paper is to study insulin pump therapy and accurate monitoring of glucose levels in diabetic patients, which are current research trends in diabetology. Both problems have a wide margin for improvement and promising applications in the control of parameters and levels involved.
The authors have registered data for the levels of glucose in diabetic patients throughout a day with a temporal resolution of 5 minutes, the amount and time of insulin administered and time of ingestion. The estimated quantity of carbohydrates is also monitored. A mathematical model for Type 1 patients was fitted piecewise to these data and the evolution of the parameters was analyzed.
They have found that the parameters for the model change abruptly throughout a day for the same patient, but this set of parameters account with precision for the evolution of the glucose levels in the test patients. This fitting technique could be used to personalize treatments for specific patients and predict the glucose-level variations in terms of hours or even shorter periods of time. This way more effective insulin pump therapies could be developed.
The proposed model could allow for the development of improved schedules on insulin pump therapies.
Acedo, L., Botella, M., Cortés, J.C., Hidalgo, J.I., Maqueda, E. and Villanueva, R.J. (2018), "Swarm hybrid optimization for a piecewise model fitting applied to a glucose model", Journal of Systems and Information Technology, Vol. 20 No. 4, pp. 404-416. https://doi.org/10.1108/JSIT-10-2017-0103Download as .RIS
Emerald Publishing Limited
Copyright © 2018, Emerald Publishing Limited