Development of a Behavioral Observer for Type 1 Diabetes Mellitus (Phase1)
Development of a bio-behavioral stochastic model-predictive controller (SMPC) for use as an artificial pancreas in T1DM requires fundamental behavioral and physiology studies, as well as translational modeling and engineering development. In order to be successful, closed-loop control in Type 1 Diabetes Mellitus (T1DM) must adapt to individual physiologic characteristics and to the behavioral profile of each person. An essential part of this adaptation is biosystem (patient) observation. The investigators propose to lay the foundation for a closed-loop control system which will include algorithmic observers of patients' behavior and metabolic state.
|Study Design:||Intervention Model: Single Group Assignment
Masking: Open Label
|Official Title:||Development of a Behavioral Observer for Type 1 Diabetes Mellitus|
- Development of behavioral observer [ Time Frame: Over a period of six - 8 weeks. ] [ Designated as safety issue: No ]A behavioral observer - will be able to track over time key recurrent elements, such as wake-up time, meals, exercise, and daily patterns of risks for hypo- or hyperglycemia.
|Study Start Date:||April 2010|
|Study Completion Date:||June 2011|
|Primary Completion Date:||June 2011 (Final data collection date for primary outcome measure)|
This is a field study that will investigate behavioral events (e.g. meals, exercise) in T1DM and daily glucose patterns using an insulin pump, continuous glucose monitoring (CGM) data, and frequent SMBG tagged with behavioral markers (recent food and activity). From these data, a learning algorithm - behavioral observer - will be able to track over time key recurrent elements, such as wake-up time, meals, exercise, and daily patterns of risks for hypo- or hyperglycemia. In future controllers, behavioral observation such as this will be used to forecast upcoming routine events, enabling open-loop and closed-loop control algorithms to deal with the probabilistic patterns of patients' self-treatment behavior.
This intensive descriptive study will follow 60 adults with T1DM who are currently experienced with insulin pump use for a two-week training period plus a one month active study period during which the DexCom SEVEN® PLUS Continuous Glucose Monitor (CGM) will be used in tandem with the OmniPod® Insulin Management System. The OmniPod® has a built in FreeStyle glucometer that allows tagging of food and activity-related treatment behaviors with each self-monitoring blood glucose (SMBG) value. The OmniPod® personal digital assistant (PDA) also stores information about insulin delivery and meal size in relation to the carbohydrate content. Parallel recording of CGM and behavioral data, as well as psychometric instruments will produce a rich synchronized data set for each person that will ultimately lead to the development of a behavioral event generator for use in future open-loop and closed-loop control algorithms for intelligent insulin dosing.
|United States, Virginia|
|University of Virginia - Center for Diabetes Technology|
|Charlottesville, Virginia, United States, 22901|
|Principal Investigator:||Boris P Kovatchev, Ph.D.||University of Virginia|