Developing control systems for medical applications poses particular challenges, some of which are unique to this field. Physiological systems involve a multitude of interacting subsystems and networks, with multiple feedforward and feedback loops, and interactions at many levels. The dynamics vary from one individual to another, and within the same individual over time. Many times important states can not be measured, and at times, not even estimated. Quantifying clinical objectives is another challenge, as they do not easily translate into the mathematical performance measures common in control systems theory.
In our current research we are aiming to develop better algorithms to improve insulin dosing in type 1 diabetic individuals. We have two parallel tracks, one aiming at more immediate application with current technology, and a second to have algorithms in place once reliable continuous sensors become available.
Given current constraints of sparse measurements, we are applying a run-to-run strategy to determine the insulin dosing related to meals, as well as for the adjustment of basal infusion rates. This can be incorporated into a continuous control algorithm at a later stage, in which the run-to-run structure can provide feedforward action based on input from the user related to meal consumption.
The long term objective of this work is to develop new algorithmic approaches to optimize the delivery of insulin in an automated fashion to people with type 1 diabetes. Specifically, novel approaches to patient characterization will be developed to predict glucose profiles under various conditions of stress, exploiting developments in pattern recognition from the engineering literature. The net result will be the development of an algorithm that predicts the dosages of insulin delivered to the patient by the clinical team. This will involve the identification of recurring patterns of glucose response to meal and other stimuli. The algorithm will be tested in both simulation and clinical trials for varying degrees of patient stress and meal stimuli, as well as robustness to sensor noise and patient characterization uncertainty.
The AP Database is a new tool for locating, analyzing, and comparing published clinical studies of the Artificial Pancreas. The database allows users to find all clinical studies meeting specified criteria, as well as immediately see the most relevant details from those studies.
Click Here to access the database!
- A Pediatric Diabetes Gamechanger
NIH-funded research on a pediatric artificial pancreas could make sleepless nights and high anxiety a thing of the past for parents of children with Type 1 diabetes To view the complete story, view the press release. Also see the story from UCSB. Related Group Members
- UCSB diabetes researcher refines fine art of forecasting behavior
A recent news story highlighted in the Pacific Coast Business Times. http://www.pacbiztimes.com/2015/01/16/dubroff-ucsb-diabetes-researcher-refines-fine-art-of-forecasting-behavior/
- Artificial Pancreas Clinical Database
Our new Artificial Pancreas Clinical Database brings together information from published studies conducted around the world, allowing protocol-specific searches and cross-study comparisons.
- Research featured in recent article from the Goleta chamber of commerce
See a recent news article from Goleta Magazine highlighting research from the biomedical control project: Goleta _Magazine_2014 Related Group Members
F.J. Doyle III, L. M. Huyett, J. B. Lee, H. C. Zisser, E. Dassau , "Closed- Loop Artificial Pancreas Systems: Engineering the Algorithms," Diabetes Care, May 2014. [DOI]
A. Srinivasan, J.B. Lee, E. Dassau, F.J. Doyle III, "Novel insulin delivery profiles for mixed meals for sensor-augmented pump and closed-loop artificial pancreas therapy for type 1 diabetes mellitus," Journal of Diabetes Science and Technology, vol. 8, no. 5, pp. 957-68,Sep 2014. [DOI]
J. B. Lee, E. Dassau, D. Seborg, F.J. Doyle III, "Model-Based Personalization Scheme of an Artificial Pancreas for Type 1 Diabetes Applications," Proceedings of the American Controls Conference 2013.
E. Dassau, H. Zisser, R.A. Harvey, M.W. Percival, B. Grosman, W. Bevier, E. Atlas, S. Miller, R. Nimri, L. Jovanovic, F.J. Doyle III "Clinical evaluation of a personalized artificial pancreas," Diabetes Care, vol. 36, no. 4, pp. 801-9, Apr 2013. [DOI]