Fluorescence Image Filtering for Accurate Non-Invasive Glucose Testing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fluorescence analysis methods for blood glucose testing are inaccurate due to pigment areas in human skin, which cause deviations in spectral data collection, and existing Raman spectroscopy systems are bulky and expensive, while hyperspectral data analysis struggles with separating spectral signals of different wavelengths and is affected by skin color and thickness, making accurate measurement challenging.
Innovation Solution
An image processing method using ultraviolet-excited fluorescence grayscale imaging to distinguish pigment areas, calculate average grayscale values, and filter out points based on variance thresholds, combined with a convolutional neural network model for analyzing spectral data to extract accurate glucose concentration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If spectral data collection is performed on pigment areas, then testing can be conducted, but large deviation occurs between spectral data of normal skin and pigment areas resulting in inaccurate test results
Solution Approach 1:
The patent extracts and removes pigment area data from the spectral data collection process. By identifying pigment areas through image processing and excluding them from analysis, the harmful interference of pigment variations on measurement accuracy is eliminated while preserving the utility of spectral testing on normal skin areas.
Solution Approach 2:
The patent applies different processing qualities to different regions of the skin. Normal skin areas undergo standard spectral analysis while pigment areas are identified and handled differently through image processing techniques, allowing each region to be processed according to its specific characteristics to maintain overall measurement accuracy.
2Measurement precision
If Raman spectroscopy system is used for blood glucose testing, then higher accuracy is achieved, but the system becomes bulky and expensive
Solution Approach 1:
The patent replaces the complex mechanical Raman spectroscopy system with a simpler optical system combining fluorescence excitation and image processing. This substitution maintains measurement accuracy by using fluorescence signals from glucose molecules while eliminating the need for bulky Raman spectroscopy equipment through clever use of standard UV light sources and cameras.
Solution Approach 2:
The patent changes the measurement parameter from Raman scattering signals to fluorescence emission signals. By exciting glucose molecules with UV light and detecting their fluorescence response, the system achieves comparable accuracy to Raman spectroscopy but with simpler, smaller, and more cost-effective equipment.
3Ease of operation
If hyperspectral data analysis is used for non-invasive testing, then non-invasive measurement is achieved, but spectral signals of different wavelengths are mixed together making fine separation difficult
Solution Approach 1:
The patent extracts only the relevant fluorescence signal information from the hyperspectral data while discarding redundant wavelength information. By focusing on the specific fluorescence emission spectrum of glucose rather than analyzing all wavelengths, the system achieves non-invasive testing while simplifying signal processing and avoiding the complexity of separating mixed spectral signals.
Solution Approach 2:
The patent applies specialized processing to the fluorescence signal region of interest within the hyperspectral data. Rather than attempting to separate all spectral components, the system focuses analysis on the specific wavelength range where glucose fluorescence occurs, making signal extraction straightforward while maintaining non-invasive capabilities.
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
Effectively reduces the interference of pigment areas, enables non-invasive testing, and achieves accurate glucose concentration measurement with a miniaturized and cost-effective system, providing high testing accuracy and comfort.
Implementation Method 1
irradiating a first area by ultraviolet light within a preset wavelength, to obtain a fluorescence grayscale image of an imaging area
Data Source
AI summary
The present invention provides an image processing method and system in analyte testing, a medium, and a device. The method includes: ultraviolet-excited fluorescence grayscale image obtaining: irradiating a first area by ultraviolet light within a preset wavelength, to obtain a fluorescence grayscale image of an imaging area; grayscale value obtaining: marking three or more pre-selected collection points into the fluorescence grayscale image, and obtaining corresponding grayscale values D; and filtering: calculating an average value Mu and a variance Sigma of grayscale values of all collection points, filtering out a collection point of |D−Mu|>Sigma×threshold, where the threshold is a preset threshold, and using remaining collection points as optional points. According to the present invention, a pigment area in the skin can be effectively distinguished, and the interference of the pigment area on spectral data can be reduced as much as possible.