Analyte Sensor Dropout Detection During Repeating Data Excursions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing analyte monitoring systems, such as continuous glucose monitors, often provide false low readings known as 'dropouts', which are difficult to detect and can trigger unnecessary alarms, complicating diabetes management.
Innovation Solution
A method to analyze analyte monitoring system sensor data by segmenting it into time series associated with repeating events, applying curve smoothing, time dilation, and dynamic range scaling to normalize and compare data, identifying dropouts by detecting anomalous low levels across multiple days.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If analyte monitoring systems provide continuous sensor data, then glucose level monitoring capability is improved, but false low readings (dropouts) occur that are difficult to detect
Solution Approach 1:
The patent segments continuous sensor data into discrete time series associated with repeating events (meals, snacks). Each time series represents a specific eating event and can be independently analyzed for dropouts. This segmentation allows focused comparison between corresponding time periods across different days, improving dropout detection without compromising continuous monitoring capability.
Solution Approach 2:
The patent performs preliminary normalization of sensor data through curve smoothing, time dilation, and dynamic range scaling before comparison. This preliminary processing aligns the temporal and amplitude characteristics of different time series, enabling accurate detection of anomalous low readings that would otherwise be masked by natural variability in glucose responses.
2Reliability
If dropout detection algorithms are developed, then false alarm reduction is improved, but system complexity increases
Solution Approach 1:
By segmenting data into event-specific time series, the patent reduces the complexity of analyzing entire continuous datasets. Each segmented time series can be independently normalized and compared, breaking down the complex dropout detection task into manageable, repetitive operations that are computationally efficient and easier to implement.
Solution Approach 2:
The patent creates normalized copies of sensor data through curve smoothing and time dilation transformations. These processed copies preserve the essential characteristics of original data while removing variability that complicates comparison, enabling simpler dropout detection algorithms to achieve high accuracy without processing the raw complex data directly.
3Measurement precision
If data normalization techniques are applied, then comparison accuracy across different days is improved, but processing time increases
Solution Approach 1:
The patent applies normalization techniques selectively to only the portions of data that need comparison - specifically the time series segments corresponding to repeating events. Rather than normalizing entire datasets, the method focuses processing on relevant portions, achieving sufficient accuracy for dropout detection while minimizing unnecessary processing time.
Solution Approach 2:
The patent transforms data parameters through curve smoothing, time dilation, and dynamic range scaling to enable direct comparison. These parameter changes standardize the temporal and amplitude characteristics of glucose responses across different days, allowing accurate dropout identification without requiring extensive processing of raw variable data.
Data Source
AI summary
Methods, devices, and systems are provided for identifying dropouts in analyte monitoring system sensor data including segmenting sensor data into a plurality of time series wherein each time series is associated with a different instance of a repeating event, selecting a first time series to analyze for dropouts from the plurality of time series; comparing the selected first time series to a second time series among the plurality of time series, determining whether the selected first time series includes a portion that is more than a predefined threshold lower than a corresponding portion of the second time series, and displaying, on a computer system display, an indication that the selected first time series includes a dropout if the selected first time series includes a portion that is more than the predefined threshold lower than the corresponding portion of the second time series.