Virtual CGM Trace Generation Using Mathematical Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for creating Continuous Glucose Monitoring (CGM) traces fail to accurately capture the dynamic effects of meal and exercise behaviors on blood glucose levels in real-life settings, leading to incomplete or inaccurate representations of glycemic outcomes in type 1 and 2 diabetes management.
Innovation Solution
A method and system for virtualizing a continuous glucose monitoring trace using a mathematical model that estimates blood glucose levels based on episodic data, incorporating additional metabolic signals and weighing recent and historical data to improve accuracy, including meal, insulin, and physical activity inputs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing mathematical models are used to generate CGM traces, then the models can represent dynamic relationships of glucose and insulin transport, but they fail to accurately capture the wide excursions of blood glucose observed in real-life settings due to incomplete behavioral models
Solution Approach 1:
The patent creates a virtual CGM trace that copies and reconstructs the characteristics of a patient's actual CGM trace using episodic BG data and mathematical models. This virtual trace serves as a simplified representation that captures the essential glycemic patterns without requiring complex behavioral modeling, thereby achieving accurate blood glucose representation while avoiding the complexity of detailed behavioral models.
2Ease of operation
If physicians use known CGM trace generation methods, then traces can be created for retrospective analysis, but the traces are incomplete when patient information is unavailable or CGM data is incomplete
Solution Approach 1:
The system performs self-service by automatically generating virtual CGM traces from available episodic BG data without requiring complete patient information or existing CGM data. The mathematical models self-adjust to create accurate traces using only the data that is available, enabling retrospective analysis even when information is incomplete, thus maintaining both ease of operation and reliability.
3Productivity
If traditional interpolation methods are used to create CGM traces from episodic data, then traces can be generated quickly, but the traces fail to accurately represent the dynamic effects of meal and exercise behaviors
Solution Approach 1:
The patent transforms the approach by changing the parameters used in trace generation from simple interpolation to mathematical models that incorporate meal and exercise behaviors. By adjusting model parameters such as insulin sensitivity, carbohydrate ratios, and behavioral patterns, the system generates traces that accurately represent glycemic outcomes while maintaining efficient computation suitable for real-time or retrospective analysis.
Data Source
AI summary
A method and system use mathematical models and available patient information to virtualize a continuous glucose monitoring trace for a period of time. Such a method and system can generate the virtualized trace when episodic patient data is incomplete. Such a method and system can also rely on self-monitored blood glucose measurement information to improve the virtualized continuous glucose monitoring trace.