RF Glucose Monitoring Training Data via Convolution Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current glucose monitoring devices require invasive blood sampling and are prone to contamination, with existing training data generation methods for supervised machine learning algorithms being inefficient, necessitating improved data quality for more accurate and less intrusive glucose level monitoring.
Innovation Solution
A system and method for generating training data using a waveform database of glucose waveforms and health parameters, where radio-frequency glucose detection signals are transmitted and received through a two-dimensional array of antennas, with convolution matching and labeling to extract and correlate pulse wave signals with corresponding blood glucose levels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If blood sampling methods are used for glucose monitoring, then measurement precision is improved, but ease of operation deteriorates due to invasiveness and contamination risk
Solution Approach 1:
The patent replaces the mechanical blood sampling process with a radio frequency electromagnetic field-based detection system. The system uses RF signals to interact with glucose molecules in tissue without requiring physical blood contact, thereby eliminating needle pricking and contamination risks while maintaining measurement capability through dielectric property detection.
Solution Approach 2:
The patent introduces radio frequency electromagnetic fields as an intermediary between the measurement system and the glucose molecules. Instead of direct mechanical contact with blood, the RF fields serve as a mediator to detect glucose concentration through changes in tissue dielectric properties, enabling non-invasive measurement.
2Productivity
If filtering methods are used to generate training data, then data processing speed is improved, but manufacturing precision deteriorates due to reduced data quality
Solution Approach 1:
The patent performs preliminary actions by collecting and storing large volumes of raw RF signal data and corresponding glucose measurements during normal operation. This pre-collected dataset is then used for training machine learning models, eliminating the need for time-consuming filtering processes while maintaining high data quality through the use of matched pairs of RF signals and ground truth glucose values.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enables the creation of high-quality training data for health monitoring systems, improving the accuracy and non-invasiveness of glucose level monitoring by using radio-frequency signals to generate and label data, reducing contamination risks and enhancing the reliability of glucose monitoring devices.
Implementation Method 1
radio frequency scanning data that corresponds to radio waves that have responded from below the skin surface of a person
Implementation Method 2
one or more transmit antennas configured to transmit radio-frequency (RF) glucose detection signals into a user and one or more receive antennas configured to detect RF glucose signals
Data Source
AI summary
A method for generating training data for use in monitoring a health parameter of a person in which a waveform database of glucose waveforms and health parameters is created. The device receives a pulse wave signal that is generated from radio frequency scanning data that corresponds to radio waves that have responded from below the skin surface of a person, wherein the radio frequency scanning data is collected through a two-dimensional array of receive antennas over a range of radio frequencies. Then at least one of the pulse wave signals is extracted in response to using convolution matching to the waveform database to generate an extracted signal. Then the data is labeled corresponding to the extracted signal with a corresponding blood glucose level to generate training data.