Infrared Vessel Image Processing for Non-Invasive Analyte Testing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing analyte testing technologies face challenges such as invasiveness, high cost, bulkiness, and inaccurate measurements due to mixed spectral signals and interference from non-analyte components, particularly in non-invasive optical methods like Raman spectroscopy and hyperspectral data analysis.
Innovation Solution
An image processing method that involves obtaining infrared grayscale images, distinguishing dark and bright spots based on pixel gradients, selecting target candidate points using standard deviation, and performing outlier detection to accurately identify blood vessel areas for optical data collection, utilizing fluorescence spectroscopy for analyte testing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If Raman spectroscopy is used for non-invasive blood glucose measurement, then measurement accuracy is improved, but device complexity and cost increase significantly
Solution Approach 1:
The patent extracts and emphasizes the use of fluorescence spectroscopy instead of Raman spectroscopy, taking out the complex Raman system and replacing it with a simpler fluorescence-based approach that can be implemented with portable devices while maintaining non-invasive measurement capability
Solution Approach 2:
The patent adopts a cost-effective fluorescence spectroscopy system that can be implemented with affordable, portable equipment rather than requiring expensive laboratory-grade Raman spectroscopy systems, making the technology accessible for widespread use
2Ease of operation
If hyperspectral data analysis is used for non-invasive testing, then non-invasive measurement is achieved, but measurement accuracy deteriorates due to mixed spectral signals
Solution Approach 1:
The patent segments the imaging area into distinct regions of interest (dark spot area and bright spot area) based on pixel gradient analysis, allowing separate analysis of spectral signals from different tissue types to isolate the blood glucose signal from confounding tissue signals
Solution Approach 2:
The patent applies local quality analysis by examining pixel gradient characteristics to distinguish between different tissue regions, using the local optical properties to identify blood vessel locations and extract accurate spectral signals specific to the analyte of interest
3Quantity of substance
If spectral signals from different wavelengths are collected together, then comprehensive data is obtained, but signal separation difficulty increases
Solution Approach 1:
The patent segments the spectral data analysis by first segmenting the spatial domain into dark spot and bright spot areas based on pixel gradients, then analyzing spectral signals separately in each region to facilitate easier separation and interpretation of mixed wavelength signals
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 analyte measurement by distinguishing between blood vessel and non-vessel areas, reducing outlier influence, and correlating spectral data strongly with analyte concentration, facilitating real-time testing with a miniaturized system.
Implementation Method 1
obtaining an infrared grayscale image by imaging a first area
Implementation Method 2
calculating a grayscale gradient of a pixel combination including each pixel in the dark spot area and a surrounding pixel
Implementation Method 3
performing outlier point detection on the first target candidate points, obtaining a selected first target candidate point by using a standard deviation with a preset multiple
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: image obtaining step: obtaining an infrared grayscale image by imaging a first area; dividing: based on a pixel gradient, distinguishing a dark spot from a bright spot, and dividing the infrared grayscale image into a dark spot area and a bright spot area; point selecting step: calculating and selecting first target candidate points; and selecting step: performing outlier point detection and selection on the first target candidate points, and obtaining a position of a first target.