Spectral Sensor Mosaic Filtering for Non-Invasive Glucose Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing non-invasive blood glucose testing methods using near-infrared spectroscopy face challenges due to strong absorption by human tissues, leading to interference that reduces the accuracy of blood glucose calculation models, as water, muscle, bone, and protein in human tissues have strong absorption characteristics for near-infrared light, overlapping with blood glucose spectra and obscuring relevant information.
Innovation Solution
A data processing method and system that utilizes a periodic pixel-level light filtering structure on a spectral sensor to modulate testing light, forming a mosaic image, and calculates spectral data from testing and reference points to remove non-analyte interference, enabling accurate glucose concentration measurement without invasive methods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If near-infrared spectroscopy is used for non-invasive blood glucose testing, then non-invasive measurement is achieved, but tissue absorption interference reduces measurement accuracy
Solution Approach 1:
The patent segments the spectral measurement process into multiple wavelength points and applies principal component analysis to separate tissue interference components from blood glucose signal components. This segmentation of the spectral data allows independent analysis and removal of interference, resolving the contradiction between non-invasive measurement and accuracy.
Solution Approach 2:
The patent introduces an intermediary processing layer (spectral analysis algorithm with principal component analysis) between the raw spectral data and the final blood glucose concentration calculation. This intermediary step identifies and removes tissue interference components, enabling accurate measurement despite the non-invasive approach.
2Device complexity
If traditional spectral filtering methods are used, then device complexity is reduced, but interference components cannot be effectively removed
Solution Approach 1:
The patent replaces traditional mechanical/optical filtering methods with a computational approach using principal component analysis and spectral unmixing algorithms. This substitution maintains simple device structure while achieving effective removal of interference components through software-based signal processing.
Solution Approach 2:
The patent transforms the physical filtering problem into a mathematical parameter optimization problem by analyzing spectral data at multiple wavelength points and applying dimensionality reduction techniques. This parameter-based approach achieves superior interference removal without increasing hardware complexity.
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
The method improves accuracy by removing non-analyte interference, allows non-invasive testing, and facilitates real-time, low-cost glucose monitoring using fluorescence spectroscopy, with high signal-to-noise ratio and minimal system size.
Implementation Method 1
generating an image from the obtained reflection signal or excitation signal by a periodic pixel-level light filtering structure provided on a sensor surface
Implementation Method 2
facilitates real-time, low-cost glucose monitoring using fluorescence spectroscopy
Implementation Method 3
The near-infrared spectroscopy is considered as one of the most promising non-invasive blood glucose testing technologies
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
The present invention provides a data processing method and system for a spectral sensor, a medium, and a device. The method includes: data obtaining: obtaining a reflection signal or an excitation signal generated when an imaging area is irradiated by light; light filtering: generating a mosaic image from the obtained reflection signal or excitation signal by a periodic pixel-level light filtering structure provided on a sensor surface; and processing: based on a testing point candidate area and a reference point candidate area pre-divided in the imaging area, respectively selecting a testing point and a reference point from corresponding positions in the mosaic image, and respectively calculating spectral data of the testing point and spectral data of the reference point.