Skip to content

Identification of Areas of Artificial Pancreas Algorithm Enhancements Through Big-Data Analysis (Part 1)

Deadlines are 5:00 PM (Eastern). No extensions will be granted.


Milestone Date Status
Letter of Intent N/A
Application Nov 28, 2018 Passed
Award Notification Apr 30, 2019 Passed
Earliest Start Jun 01, 2019 Passed

Background & Purpose

Please click on the “RFA ANNOUNCEMENT” link in the upper right corner for complete information.

PURPOSE

JDRF is committed to advancing the effectiveness and usability of artificial pancreas (AP) systems to treat type 1
diabetes (T1D). One focus area addressing both of these objectives is to improve the AP algorithm, which doses
insulin autonomously based—mostly—on readings from a continuous glucose monitor and records of previously
delivered insulin. JDRF challenges applicants to help understand how much better next-generation algorithms
might be if researchers/developers leverage a large real-world data set from which to glean insights and identify
areas for AP algorithm enhancements.

BACKGROUND

T1D is characterized by the loss of the body’s ability to produce insulin, a hormone which regulates blood glucose
levels tightly in individuals without diabetes. With no endogenous insulin production, people with T1D rely on
exogenous insulin. The delivery of this insulin, though, must be carefully regulated; too much insulin results in
potentially acutely dangerous low glucose levels (hypoglycemia), while too little insulin can result in problematic
high glucose levels (hyperglycemia) and/or a potentially dangerous condition known as diabetic ketoacidosis
(DKA). In short, it is essential that a person with T1D is dosed the right amount of insulin at the right time. Moreover, even if a person with T1D were able to provide this degree of control, it would necessitate an unacceptably high burden of self-management.

AP devices [interchangeably, automated insulin delivery (AID) devices] are currently the most advanced devicebased
treatment for T1D. These devices integrate three components to provide (at least partially) automated
insulin delivery:

1) The sensor: a continuous glucose monitor (CGM) which continuously measures glucose levels in the
body,
2) The actuator: an insulin pump, which has continuous access to the body to deliver insulin, and
3) The controller: an algorithm, which predominantly uses the information stream from the CGM and
information about previously delivered insulin to calculate the optimal insulin infusion dose for the current
conditions, and commands the insulin pump to deliver this calculated dose.