Adaptive Spectral Glucose Testing With Cloud Model Updating

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for non-invasive blood glucose testing face challenges due to individual differences in spectral data characteristics, leading to inaccurate results, and existing systems are bulky, expensive, or require complex algorithms with multiple biological signals, making real-time testing difficult.

Innovation Solution

A method and system that utilizes spectral data analysis by inputting data into a local testing model, allowing for user-specific model updates, and employs fluorescence spectroscopy to obtain accurate spectral data through infrared and ultraviolet light, distinguishing between testing and reference points on the skin to exclude non-analyte influences, and uses a cloud platform for model training and updating.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If a unified testing model is used for all people, then the system is simple to operate, but the measurement precision deteriorates due to individual differences in spectral data characteristics

Engineering Contradiction:
Improvesystem operation simplicityVSAvoidblood glucose measurement accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The system transitions from a static unified model to a dynamic adaptive model that automatically adjusts to individual user characteristics. The model continuously learns from spectral data of different users and updates its parameters, enabling it to adapt to individual differences in skin properties, anatomy, and spectral characteristics while maintaining ease of use through automated adaptation

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system performs self-calibration and self-updating without requiring manual intervention. It automatically collects spectral data, identifies individual characteristics, and trains updated models using the collected data, allowing the system to serve itself in improving measurement accuracy for each user over time

Inventive Principle:
Principle #25Self-service

2Measurement precision

If laboratory-grade Raman spectroscopy system is used, then the measurement precision is improved, but the device complexity and cost increase significantly

Engineering Contradiction:
Improveblood glucose concentration measurement accuracyVSAvoidsystem size and cost
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces complex mechanical optical systems with computational methods. Instead of using sophisticated Raman spectroscopy hardware, it uses spectral data collection combined with machine learning algorithms and neural networks to achieve accurate measurements, substituting physical complexity with computational intelligence

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system changes the approach from measuring physical properties directly to measuring spectral characteristics and inferring concentrations through computational models. By transforming the measurement paradigm from direct physical detection to spectral analysis with AI processing, it achieves accuracy without requiring complex laboratory equipment

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If multiple biological signals are collected from different body positions, then the measurement precision is improved, but the device complexity and cost increase excessively

Engineering Contradiction:
Improveblood glucose concentration accuracyVSAvoidnumber of sensors and modules
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system extracts and focuses on the most critical spectral information from the available data. Instead of collecting and processing multiple biological signals from different body positions, it concentrates on extracting meaningful spectral characteristics from targeted measurement locations, filtering out redundant information while maintaining measurement accuracy

Inventive Principle:
Principle #2Taking out (Extraction)

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

Enables accurate, non-invasive, and cost-effective real-time blood glucose testing by minimizing interference from non-analyte factors, achieving high accuracy and miniaturization without the need for invasive procedures.

Implementation Method 1

the first image includes distribution data that indicate distribution in an imaging area of a reflection signal or an excitation signal generated by the analyte when irradiated by light

Methodology Applied
Scientific EffectReflection: Reflection

Implementation Method 2

the light in the second wavelength range can cause the analyte to excite a fluorescence radiation signal, and a main peak of a fluorescence spectrum of the fluorescence radiation signal is within the effective response range of the imaging spectrum detection apparatus

Methodology Applied
Scientific EffectFluorescence: Fluorescence

Data Source

PatentEP4681640A1Method and system for using spectral data of analyte, method and system for testing analyte, medium, and device
Publication Date: 2026.01.21 SENSURA PTE LTD
  • EP4681640A1 patent drawingFigure 1~2
  • EP4681640A1 patent drawingFigure 3~4
  • EP4681640A1 patent drawingFigure 5~6

AI summary

The present invention provides a method and a system for using spectral data of an analyte, a method and a system for testing an analyte, a medium, and a device, which relate to the field of optical analysis. The method includes: analyzing step: inputting obtained spectral data into a local testing model to obtain information about the analyte, and providing a correction option; information uploading step: after the correction option is triggered, obtaining correction information input by a user, and uploading the correction information and the obtained spectral data to a cloud platform; model training step: using, by the cloud platform, the obtained spectral data as an input and the obtained correction information as an output to train a cloud testing model; and model updating step: updating the local testing model based on a trained cloud testing model.