Biological Information Analysis Device Anomaly Detection Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional techniques face difficulties in detecting anomalies in biological information measured by wearable devices due to daily variations, making it challenging to ascertain anomalies in user data.
Innovation Solution
A biological information analysis apparatus and method that includes a sensor data acquisition unit, data analysis unit, measurement anomaly detection unit, and presentation unit, which calculate representative values from time-series data across multiple time intervals and detect anomalies based on these values, outlier determination, and statistical summarization to identify deviations or inconsistencies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Stability of the object's composition
If conventional measurement techniques are used to continuously measure biological information, then measurement stability is improved, but the ability to detect anomalies in the presence of daily variations deteriorates
Solution Approach 1:
The patent segments the continuous biological information measurement data into multiple time intervals (e.g., morning, afternoon, evening). By dividing the measurement period into discrete segments and calculating representative values for each segment, the system can compare corresponding time points across different days to detect anomalies while accounting for daily variations. This segmentation allows the system to maintain measurement stability while improving anomaly detection precision.
2Measurement precision
If representative values are calculated from multiple pieces of biological information across multiple time intervals, then anomaly detection capability is improved, but data processing complexity increases
Solution Approach 1:
The patent extracts only the essential features from the raw biological information data by calculating representative values (such as averages or typical values) for each time interval. This extraction process reduces the volume of data that needs to be processed while retaining the key characteristics necessary for anomaly detection. By taking out only the relevant representative values rather than processing all raw data points, the system improves anomaly detection capability while managing data processing complexity.
3Measurement precision
If outlier determination is performed based on preset criteria to detect anomalies, then detection accuracy is improved, but false positive rate may increase
Solution Approach 1:
The patent employs feedback mechanisms where the system continuously compares representative values from corresponding time intervals across multiple days and uses this comparative information to adjust anomaly detection thresholds. By incorporating feedback from historical data and establishing dynamic reference ranges based on individual user patterns, the system can distinguish between normal variations and genuine anomalies more accurately. This feedback approach improves detection accuracy while reducing false positives by adapting to the user's unique biological rhythms.
Data Source
AI summary
A biological information analysis apparatus includes a sensor data acquirer that acquires biological information of a user measured by a sensor, a data analyzer that analyzes time-series data for the biological information over a plurality of time intervals and, from multiple pieces of biological information that were acquired at mutually corresponding times of measurement in respective ones of the plurality of time intervals, calculates representative values for the multiple pieces of biological information, a measurement anomaly detector that detects an anomaly contained in the measured biological information based on time-series data with the representative values calculated by the data analyzer or on the time-series data for the multiple pieces of biological information that was used in calculation of the time-series data with the representative values, and a presenter that outputs the calculated representative values for the multiple pieces of biological information and information indicating the detected anomaly.