Closed-Loop Insulin Control for Unannounced Meal Response
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current closed-loop control systems for Type 1 diabetes mellitus (T1DM) struggle to automate insulin delivery in response to unannounced meals, leading to prolonged hyperglycemia due to delays in CGM sensing and insulin action, and are prone to hypoglycemia from aggressive insulin administration.
Innovation Solution
A closed-loop control system using Model Predictive Control (MPC) and a Bolus Priming System (BPS) to predict glycemic fluctuations, automatically adjust basal insulin dosing, and deliver insulin boluses based on probability of unannounced meals, minimizing hyperglycemia and hypoglycemia risks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If hybrid closed-loop control systems are used to automatically modulate insulin infusions, then glycemic control during overnight periods is improved, but the systems cannot prevent prolonged hyperglycemia following unannounced carbohydrate consumption
Solution Approach 1:
The system performs preliminary detection of glycemic disturbance patterns using CGM data analysis before significant hyperglycemia occurs. The disturbance detection mechanism identifies early signs of unannounced meal consumption by monitoring glucose level changes and rates of change, enabling the system to prepare for corrective action in advance
Solution Approach 2:
The system implements continuous feedback through CGM measurements to detect glycemic disturbances. By monitoring glucose levels and their rates of change in real-time, the system receives feedback about unannounced carbohydrate consumption and automatically adjusts insulin delivery in response to maintain glycemic control
2Ease of operation
If commercial hybrid APs require manual carbohydrate bolus requests, then user control is maintained, but boluses must be proportional to a priori meal-size estimation which is often inaccurate
Solution Approach 1:
The system performs self-service by automatically detecting unannounced glycemic disturbances through CGM data analysis and autonomously administering appropriate insulin boluses. The disturbance detection mechanism eliminates the need for user estimation of meal size by independently monitoring glucose level changes and determining when carbohydrate consumption has occurred
Solution Approach 2:
The system replaces the manual mechanical process of user bolus request with an automated electronic detection and control system. The CGM-based disturbance detection mechanism substitutes human judgment and manual operation with automated sensor-based detection and algorithmic decision-making
3Extent of automation
If fully automated APs are implemented to reject glycemic disturbances, then mealtime insulin boluses are automatically administered, but inherent delays in CGM sensing and insulin action remain
Solution Approach 1:
The system performs preliminary detection of glycemic disturbance patterns using CGM data analysis before significant hyperglycemia occurs. By monitoring glucose level changes and rates of change in real-time, the system identifies unannounced meal consumption early, enabling faster response despite inherent sensing and insulin action delays
Solution Approach 2:
The system dynamically adjusts its response based on the severity and progression of detected glycemic disturbances. By continuously monitoring CGM data and adapting insulin delivery in real-time, the system optimizes its corrective action to compensate for time delays in the control loop
Data Source
AI summary
A method, system, and computer-readable medium are provided for a dual mode Closed-Loop Control (CLC) system integrating each of (i) an adaptive, personalized Model Predictive Control (MPC) control law that modulates the control strength of insulin infusion depending on recent past control actions, glucose measurements, and their derivative(s), (ii) an automatic Bolus Priming System (BPS) that commands additional insulin injections upon the detection of enabling metabolic conditions (e.g., an unannounced meal), and (iii) a hyperglycemia mitigation system (HMS) to avoid prevailing hyperglycemia.