Insulin Delivery Visualization Tool for Glycemic Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current artificial pancreas systems generate complex data that overwhelms users and healthcare providers, lacking a suitable tool for interpretation and analysis, which can lead to suboptimal glycemic control and reduced trust in the system.
Innovation Solution
A visualization and analysis tool is integrated into the insulin delivery system to detect and display artificial pancreas activity events, providing metrics such as Hypo-APAEs and Hyper-APAEs, which quantify significant insulin-modulating actions and help users and caregivers understand the system's performance, thereby improving glycemic control and trust.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If autonomous insulin modulation algorithms (MPC/PID) are used to improve glycemic control, then glycemic control is improved, but data complexity increases and overwhelms users
Solution Approach 1:
The patent extracts and isolates specific meaningful patterns from the complex algorithm-generated data by defining activity events (APAEs) that represent significant insulin-modulating actions. This separates the clinically relevant information from the overwhelming raw data, allowing users to focus on actionable insights while maintaining the benefits of autonomous control.
Solution Approach 2:
The visualization and analysis tool serves as an intermediary between the complex autonomous insulin modulation algorithms and the end users. It translates algorithm outputs into comprehensible activity events and metrics, bridging the gap between sophisticated control systems and user understanding without compromising either glycemic control or usability.
2Reliability
If complex autonomous algorithms are implemented to improve glycemic control, then glycemic control is improved, but user trust decreases due to lack of interpretability
Solution Approach 1:
The patent implements feedback by providing users with visualizations of algorithm activity events and their impacts on glucose levels. Users can see how autonomous actions (APAEs) correlate with glycemic changes, creating a feedback loop that builds trust and understanding of the system's decision-making process while maintaining autonomous control performance.
Solution Approach 2:
The analysis tool acts as an intermediary that makes the black-box algorithm transparent to users by visualizing its actions and their effects, thereby preserving system interpretability without sacrificing the advanced glycemic control capabilities provided by autonomous algorithms.
3Reliability
If detailed algorithm activity data is provided to improve glycemic control, then glycemic control is improved, but ease of operation decreases due to data overload
Solution Approach 1:
The patent segments the continuous complex algorithm output into discrete, meaningful activity events (APAEs) that represent specific insulin-modulating actions. This segmentation transforms overwhelming continuous data into manageable, interpretable discrete events that are easier for users to analyze and understand while maintaining comprehensive glycemic control information.
Solution Approach 2:
The system extracts only the most clinically relevant algorithm activities as defined APAEs, filtering out unnecessary data complexity. This extraction provides detailed information about meaningful insulin modulation events while eliminating data overload, improving both glycemic control monitoring and user ease of operation.
Data Source
AI summary
A visualization and analysis tool is provided for an insulin delivery system, such as an artificial pancreas, in which insulin is delivered based upon a system algorithm using a pump that is patient controllable in order to adjust insulin delivery relative to a baseline delivery rate; e.g., pre-set basal rate and a sensor for measuring glucose levels. The tool is configured with a controller of the system to detect and log events that are based on differences between actual insulin delivered by the system and the baseline delivery rate. These detected events are metrics that provide information relating to the therapeutic value of the system which, without such metrics, may be overlooked or unnoticed, thereby fostering trust and confidence in the delivery system. In addition, information is provided which may enable further improved glucose control.