Glucose Data Segmentation for Activity Recommendation Ranking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
It is challenging for healthcare providers to determine whether specific activities have a positive or negative impact on an individual's glucose or pressure levels, as these metrics vary widely over time and are influenced by various factors, making it difficult to identify beneficial activities for improving health outcomes.
Innovation Solution
A system and method that collects glucose level data over a desired time span, separates it into subsamples, analyzes these subsamples using factors like average area under the curve, standard deviation, and biotransform algorithm results, and ranks them to identify the highest-ranked time period associated with desirable glucose levels, thereby recommending activities that likely positively influenced these readings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If glucose level data is collected continuously over time to identify activities that positively influence glucose levels, then the accuracy of activity recommendations is improved, but the complexity of data processing and analysis increases
Solution Approach 1:
The patent segments the continuous glucose data into discrete time periods (e.g., 24-hour periods) and creates subsamples based on these segments. This segmentation allows the system to manage large volumes of continuous data by breaking them into manageable units that can be independently analyzed for specific activities, thereby reducing processing complexity while maintaining accuracy.
Solution Approach 2:
The patent performs preliminary actions by pre-processing glucose data into standardized time periods and subsamples before conducting activity analysis. This preliminary organization of data includes calculating baseline glucose levels and preparing data structures in advance, which simplifies subsequent analysis and reduces the computational complexity during the actual recommendation generation phase.
2Reliability
If multiple factors are used to rank glucose subsamples (area under the curve, standard deviation, biotransform algorithm), then the reliability of identifying beneficial activities is improved, but the computational resources required increase
Solution Approach 1:
The patent applies partial action by selectively using multiple ranking factors (area under the curve, standard deviation, biotransform algorithm results) only for the subsamples that need detailed analysis, rather than applying all factors to all data uniformly. This approach maintains reliability by using comprehensive factors where needed while reducing overall computational resource consumption by limiting the scope of complex calculations.
3Loss of information
If glucose data is separated into multiple subsamples for detailed analysis, then the ability to identify specific beneficial activities is improved, but the time required for data processing increases
Solution Approach 1:
The patent segments glucose data into multiple subsamples organized by time periods, which enables detailed analysis of specific activities during different times. This segmentation preserves information about when activities occurred and their specific impact on glucose levels, while the structured organization allows for efficient processing by analyzing only relevant subsamples rather than the entire dataset at once.
Data Source
AI summary
A method and apparatus for determining and providing activity recommendations includes receiving glucose level data and activity data. The glucose level data is formed into two or more data sets, with each set representing a different time period. Each data set is evaluated and ranked against each other set according to one or more of several different individual factors and the individual ranking for each set are combined, resulting in an overall ranking for given data sets. A highest ranked data set is then determined, which is associated thereby with a highest ranked time period. Activities of the activity data which took place within the highest ranked time period are provided as recommendations to the user to encourage greater numbers of times those activities are undertaken.