Categories | Inventors
Technology Overview
OHSU # 2526 — Adaptive exercise-enabled proportional-derivative artificial pancreas control algorithm
Summary
The current technology is an adaptive proportional-derivative (APD) control algorithm that provides automated delivery amounts of insulin and optionally glucagon to a person with type 1 diabetes using continuous glucose measurements (CGM) sensed from the person’s subcutaneous tissue. The APD algorithm includes a safety layer decision tree to prevent dosing of insulin and glucagon under certain physician-specified conditions.
Technology Overview
The APD algorithm includes a method for adapting postprandial insulin amounts in response to prior postprandial hypoglycemia episodes. This method is called “Adaptive Learning Postprandial Hypoglycemia Prevention Algorithm” (ALPHA). This APD control algorithm can be used to improve glycemic control in people with type 1 diabetes. The person with type 1 diabetes does not need to dose insulin manually through either an insulin pump or an insulin pen. People with type 1 diabetes have difficulty managing their glucose during exercise. We have shown recently through several publications that automatically detecting exercise and adjusting dosing during exercise can help avoid exercise-induced hypoglycemia. A random forest machine learning algorithm has been developed that can automatically predict hypoglycemia during exercise at the onset of exercise.
A traditional proportional-derivative or proportional-integral-derivative algorithm would never be appropriate for dosing a life-supporting drug like insulin. This is because there are certain conditions that can cause negative physiologic outcomes (e.g. too much glucagon can cause nausea). Such physiologic outcomes are challenging to be met by a PD or PID controller. Therefore, we have incorporated a safety-layer decision tree on top of the APD algorithm to ensure that these safety conditions are met.
Publications
Jacobs PG, et al., “Automated control of an adaptive bihormonal, dual-sensor artificial pancreas and evaluation during inpatient studies.” IEEE Trans Biomed Eng. 2014 Oct;61(10):2569-81. Link
Resalat N, et al., “Adaptive Control of an Artificial Pancreas Using Model Identification, Adaptive Postprandial Insulin Delivery, and Heart Rate and Accelerometry as Control Inputs.” J Diabetes Sci Technol. 2019 Nov;13(6):1044-1053. Link
Licensing Opportunity
This antibody is available for licensing.
Inventors:
Categories: