Retinal Biomarker Quantification Using Confounder-Corrected Spectral Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current medical imaging techniques for detecting biomarkers, such as amyloid in the retina, are expensive, time-consuming, and can be uncomfortable for patients, while hyperspectral imaging systems for biological tissues are costly and not easily accessible, and confounding factors like melanin content and cataracts affect the accuracy of these methods.
Innovation Solution
A method and system for quantifying biomarkers using hyperspectral and multispectral imaging that involves acquiring light reflectance data at multiple wavelengths, identifying and correcting for spectral confounders like melanin content and cataracts, and using machine learning algorithms to derive a spectral model that minimizes confounder effects, thereby improving accuracy and accessibility.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If PET scanning is used to detect amyloid in the retina, then diagnostic accuracy is improved, but cost and time consumption increase
Solution Approach 1:
The patent replaces expensive PET scanning with a low-cost hyperspectral imaging system that uses standard camera equipment and software algorithms. The system captures retinal images and processes them through spectral analysis to detect amyloid deposits, achieving comparable diagnostic accuracy without the high costs and time requirements of PET scanning.
Solution Approach 2:
The patent substitutes the complex mechanical and radioactive systems of PET scanning with optical imaging and computational algorithms. Hyperspectral imaging captures reflected light across multiple wavelengths, and machine learning algorithms analyze the spectral signatures to identify amyloid, replacing the invasive and time-consuming PET process.
2Measurement precision
If commercial hyperspectral imaging systems are used, then biomarker quantification capability is improved, but cost and accessibility worsen
Solution Approach 1:
The patent divides the complex hyperspectral imaging system into separate functional components: standard camera hardware for image capture, software modules for spectral processing, and algorithms for biomarker quantification. This segmentation allows each component to be developed and optimized independently, reducing overall system cost and improving accessibility.
Solution Approach 2:
The patent creates a universal hyperspectral imaging platform that can detect multiple biomarkers (amyloid, melanin, cataracts) using the same hardware system. The system processes reflected light across a broad spectral range and uses algorithmic analysis to identify different tissue properties, making the equipment versatile and cost-effective for various diagnostic applications.
3Productivity
If spectral analysis is performed without correcting for confounders, then processing speed is improved, but measurement precision deteriorates
Solution Approach 1:
The patent performs preliminary identification and correction of spectral confounders (melanin content, cataracts) before final biomarker quantification. The system first detects and characterizes these interfering factors using spectral analysis, then applies correction algorithms to remove their effects, ensuring accurate amyloid measurement without sacrificing processing efficiency.
Solution Approach 2:
The patent converts the harmful effect of spectral confounders into a beneficial diagnostic feature. By identifying and quantifying melanin and cataract spectral signatures, the system not only corrects for their interference but also uses this information to improve the overall accuracy of retinal tissue characterization and biomarker detection.
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 allows for accurate and cost-effective detection of biomarkers like amyloid beta in the retina, providing a reliable screening tool for conditions like Alzheimer's disease, reducing the need for invasive tests and improving patient comfort.
Implementation Method 1
acquiring light reflectance data from the tissue at a plurality of wavelengths within a main wavelength interval acquired by an imaging sensor
Data Source
Figure 1
Figure 2A~2C
Figure 3
AI summary
The present disclosure relates to a method and a system for quantifying a biomarker of a biological tissue. Two spectral sub-intervals are determined from a plurality of images of the tissue acquired at discrete wavelengths within a main wavelength interval using an imaging sensor in a manner such that the combination of the image data in the at least two spectral sub-intervals are correlated with a clinical variable of interest. The tissue is illuminated using one or more light sources with wavelengths within the at least two spectral sub-intervals. A measurement is of the light reflected by the tissue is acquired using an imaging sensor; and a measure of the biomarker of the tissue is calculated using the acquired measurement. The main wavelength interval is broader than each of the at least two spectral sub-intervals and at least one spectral confounder causing a spectral variability of the acquired image is present in the tissue.