Analyte Sensor Fault Detection via Data Pattern Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing analyte monitoring systems, such as continuous glucose monitors, often provide false readings due to fault modes like end of sensor life and early signal attenuation, which are typically detected using in vitro reference glucose readings requiring user interaction and costly test strips.
Innovation Solution
A method that analyzes sensor data to identify fault modes by computing baseline and evaluation data points on a control grid, projecting vectors along risk contours, and comparing components to predefined thresholds, allowing for fault detection without in vitro reference readings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If in vitro reference glucose readings are used to detect fault modes, then measurement precision is improved, but device complexity and loss of time increase due to user interaction requirements
Solution Approach 1:
The system performs self-diagnosis by automatically comparing sensor data patterns against stored baseline patterns to detect fault modes without requiring external reference measurements or user intervention. The processor autonomously identifies sensor failures by analyzing changes in glucose reading patterns over time.
Solution Approach 2:
The system continuously monitors sensor performance by comparing current readings against established baseline patterns and provides feedback about sensor health status. This closed-loop approach enables automatic detection of drift and failure modes without requiring external calibration.
2Measurement precision
If in vitro reference glucose readings are used to detect fault modes, then measurement precision is improved, but loss of time increases due to required user interaction
Solution Approach 1:
The system performs self-diagnosis by automatically comparing sensor data patterns against stored baseline patterns to detect fault modes without requiring external reference measurements or user intervention. The processor autonomously identifies sensor failures by analyzing changes in glucose reading patterns over time.
Solution Approach 2:
The system continuously monitors sensor performance in real-time by comparing current readings against baseline patterns, enabling immediate detection of fault modes without requiring periodic user-initiated reference measurements. This continuous monitoring eliminates gaps in detection capability.
3Measurement precision
If in vitro reference glucose readings are used to detect fault modes, then measurement precision is improved, but loss of substance increases due to consumption of test strips
Solution Approach 1:
The system performs self-diagnosis by automatically comparing sensor data patterns against stored baseline patterns to detect fault modes without requiring external reference measurements or user intervention. The processor autonomously identifies sensor failures by analyzing changes in glucose reading patterns over time.
Solution Approach 2:
The system replaces the need for expensive, consumable test strips with a digital pattern recognition approach that uses stored baseline data. This eliminates material consumption while maintaining detection capability through computational analysis of sensor readings.
Data Source
AI summary
Analyte sensor faults are detected. Datasets of glucose values sensor electronics are coupled to a glucose sensor in fluid contact with interstitial fluid under a skin surface. Baseline median glucose value and glucose variability values are computed, based on the first dataset. A baseline data point is stored. Evaluation median glucose value and variability are computed, based on the second dataset of glucose values. An evaluation data point is stored. A magnitude of a vector that extends between the baseline data point and the evaluation data point is computed. A component of the magnitude of the vector that is parallel to a hypoglycemia risk contour line is computed and compared to a predefined threshold value. An indication that a sensor fault has been detected if the component is greater than a threshold is displayed.