Artificial Pancreas Closed-Loop Control for Variable Glucose Dynamics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current artificial pancreas systems lack effective control methods to manage abrupt and slow variations in glucose dynamics, relying on limited control algorithms that struggle with constant or slowly changing glucose levels, and do not adequately address rapid changes in glucose concentrations.
Innovation Solution
A closed-loop system utilizing a multi-model predictive controller (MMPC) algorithm that integrates glucose data from a continuous glucose monitor to determine optimal insulin doses, incorporating multiple state vectors and models to adapt to changes in metabolism and body dynamics, and includes a user interface for inputting user data to adjust medication delivery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a single fixed model or slowly adapting model is used in control algorithms, then the system is simpler to implement, but it cannot adequately handle abrupt changes in glucose dynamics
Solution Approach 1:
The patent implements a multiple model predictive controller that dynamically switches between multiple glucose dynamics models based on current physiological conditions. This allows the system to adapt to both abrupt and slow variations in glucose dynamics by selecting the most appropriate model for the current state, resolving the contradiction between adaptability and complexity through dynamic model selection rather than using a single complex adaptive model
Solution Approach 2:
The control algorithm is segmented into multiple distinct glucose dynamics models, each representing different physiological states (e.g., fasting, postprandial, exercise). By dividing the control problem into multiple specialized models rather than one general model, the system achieves high adaptability to different glucose dynamics while keeping each individual model relatively simple and computationally tractable
2Reliability
If conventional control algorithms are used, then the system structure is simpler, but the glycemic control effectiveness is reduced
Solution Approach 1:
The patent implements a closed-loop control system that continuously monitors glucose levels via CGM and adjusts insulin delivery based on real-time feedback. The multiple model predictive controller uses this feedback to predict future glucose levels and optimize insulin dosing, significantly improving glycemic control effectiveness compared to conventional open-loop or simple closed-loop algorithms while maintaining manageable system complexity through efficient predictive modeling
3Speed
If adaptive control algorithms are used to handle rapid changes, then glucose control improves, but the computational requirements and system complexity increase
Solution Approach 1:
The system uses dynamic model selection that rapidly switches between pre-defined glucose dynamics models based on detected physiological states. This approach enables fast response to abrupt glucose changes by selecting an appropriate model rather than computing adaptations in real-time, achieving high speed response while keeping computational complexity manageable through the use of pre-characterized models
Solution Approach 2:
Multiple glucose dynamics models are pre-computed and characterized for different physiological conditions before actual use. When rapid glucose changes occur, the system simply selects the pre-prepared model that best matches the current state rather than computing a new model on-the-fly, achieving fast adaptive response while minimizing real-time computational complexity
Data Source
AI summary
The present disclosure relates to systems and methods for controlling physiological glucose concentrations in a patient using a closed loop artificial pancreas. The systems and methods may utilize a controller with control logic operative to execute a multi-model predictive controller algorithm to determine a medication dose to the patient.