Glucose Sequence Similarity Modeling for Intervention Inference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional continuous glucose monitoring (CGM) systems fail to utilize the sequential properties of glucose value data, limiting the inference of information between sequences with the same aggregate metrics, such as averages, which prevents effective determination of interventions that can maintain healthy blood glucose levels.
Innovation Solution
A similarity model is employed to compute similarity scores between sequences of glucose values by comparing probabilities given context data, identifying a particular sequence with a maximum similarity score, and generating an indication of associated externalities for user interfaces.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If aggregate data metrics (e.g., average TIR) are used to analyze glucose data, then clinically useful information is provided, but the sequential property of the data is lost and cannot be utilized
Solution Approach 1:
The patent segments the glucose data into discrete time points and compares sequences of glucose values at different times. By breaking down the continuous glucose data into discrete temporal segments and comparing them, the system preserves sequential information while maintaining computational manageability through structured comparison frameworks.
Solution Approach 2:
The patent introduces a temporal dimension to the analysis by comparing glucose sequences across different time points rather than using only aggregate metrics. This dimensional approach to time allows the system to capture sequential patterns and relationships between glucose values at different times, transforming the data from static aggregates to dynamic temporal patterns.
2Loss of information
If each glucose value is treated as an independent observation, then simple aggregate metrics can be calculated, but information about the sequential property of CGM data is not captured
Solution Approach 1:
The patent performs preliminary organization of glucose data into sequences at discrete time points before analysis. By pre-structuring the data into temporal sequences and preparing comparison frameworks in advance, the system enables precise capture of temporal relationships while maintaining measurement accuracy through systematic comparison of pre-organized data structures.
Data Source
AI summary
In implementations of systems for determining a similarity of sequences of glucose values, a computing device implements a similarity system to receive input data describing a sequence of user glucose values measured by a continuous glucose monitoring (CGM) system. The similarity system computes similarity scores for a plurality of sequences of glucose values by comparing each glucose values included in the sequence of user glucose values with ever glucose value included in each sequence of the plurality of sequences. A particular sequence of glucose values that is associated with a highest similarity score is identified. The similarity system determines an externality associated with the particular sequence. The similarity system generates an indication of the externality for display in a user interface.