Adaptivity Modification Coefficient for Insulin Sensitivity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automated insulin delivery (AID) systems struggle to effectively manage rapid changes in insulin sensitivity, leading to prolonged periods of hyperglycemia or hypoglycemia due to noise spikes in user sensitivity signals, particularly in individuals with Type 2 diabetes, where existing algorithms are too slow to adapt and may not account for significant deviations in insulin needs.
Innovation Solution
A method that calculates an adaptivity modification coefficient based on the ratio of previous to current insulin delivery data, adjusting the rate of adaptivity and total daily insulin setting to rapidly respond to changes in insulin sensitivity, allowing for a more aggressive adaptation of insulin delivery schedules.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If AID algorithms use a slow rate of adaptivity to reduce vulnerability to temporary changes in sensitivity, then reliability is improved, but the system cannot respond quickly to significant changes in insulin needs
Solution Approach 1:
The patent implements a dynamic adaptivity rate that adjusts based on the detected magnitude of insulin sensitivity changes. The system transitions from a fixed slow adaptivity rate to a variable rate that accelerates when significant changes are detected, allowing the system to maintain reliability during normal operation while responding quickly to urgent changes in insulin needs
Solution Approach 2:
The patent changes the adaptivity rate parameter from a fixed value to a variable value that scales with the magnitude of detected sensitivity changes. By calculating an adaptivity modification coefficient based on the ratio of previous to current insulin delivery data, the system dynamically adjusts the rate of adaptivity to match the urgency of the situation
2Device complexity
If AID algorithms use a fixed adaptivity rate, then device complexity is reduced, but the system cannot accelerate adaptation when significant insulin sensitivity changes occur
Solution Approach 1:
The patent introduces an adaptivity modification coefficient that scales with the magnitude of insulin sensitivity changes. This parameter change allows the system to accelerate adaptation when needed while maintaining a relatively simple algorithmic structure based on standard statistical calculations of the ratio between previous and current insulin delivery data
3Speed
If the adaptivity rate is increased to respond quickly to sensitivity changes, then speed is improved, but the system becomes vulnerable to noise spikes in user sensitivity signals
Solution Approach 1:
The patent dynamically adjusts the adaptivity rate parameter based on the magnitude of detected changes, increasing speed only when significant sustained changes are detected rather than responding to minor fluctuations. This conditional parameter change allows fast response when needed while filtering out noise through the threshold-based detection mechanism
Solution Approach 2:
The system uses feedback from the ratio of previous to current insulin delivery data to modulate the adaptivity rate. By continuously monitoring this ratio and adjusting the adaptivity coefficient accordingly, the system can accelerate adaptation when feedback indicates significant changes while maintaining stability when feedback shows normal variation
Data Source
AI summary
Disclosed are techniques, devices and systems that modify an insulin delivery schedule based on how sensitive a diabetic user may to fluctuations in their total daily insulin or fluctuations in their blood glucose measurement values. As a control algorithm calculates how to adapt the calculation of the user's total daily insulin, a rate of adaptivity function may be used in the calculation. The rate of adaptivity may depend on a number of factors and the disclosed techniques, devices and systems enable calculation of the rate of adaptivity to provide effective implementation or modification of a diabetic treatment plan.