Multi-Spectral Retinal Imaging for Automated Biomarker Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional retinal imaging techniques, particularly color fundus cameras, are limited in their ability to penetrate deep tissues and provide accurate biomarker detection due to limited tissue penetration of visible light frequencies and broad spectrum imaging, necessitating high demand and heavy workload for experienced ophthalmologists.
Innovation Solution
Utilizing multi-spectral retinal imaging (MSI) with a wide wavelength range and deep learning-based artificial intelligence (AI) to automate biomarker identification, extraction, and quantification from multi-spectral retinal images, including specific illumination wavelengths and image registration to enhance visibility and analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional color fundus cameras use white light source and collect all-spectral combined color image, then the imaging system is simple and cost-effective, but the tissue penetration is limited and biomarker detection accuracy is reduced
Solution Approach 1:
The patent segments the broad spectrum white light into multiple discrete wavelength bands (e.g., 450nm, 530nm, 630nm, 780nm) using optical filters or dichroic mirrors. Each wavelength band captures specific tissue characteristics, enabling precise biomarker detection while maintaining system manageability through modular optical components
Solution Approach 2:
The patent transitions from 2D color imaging to multi-dimensional spectral imaging by adding the wavelength dimension. This enables detection of biomarkers across different spectral signatures, providing deeper tissue penetration and enhanced contrast for pathological features that are invisible in conventional color images
2Measurement precision
If multi-spectral retinal imaging uses multiple individual illumination wavelengths across wide wavelength range, then the biomarker detection capability is enhanced, but the device complexity and data processing burden increase
Solution Approach 1:
The patent employs a universal imaging platform that can capture multiple wavelength bands through a single optical path using dichroic mirrors and beam splitters. This multi-functional system handles visible and near-infrared wavelengths through the same camera sensor, reducing the need for multiple separate imaging devices while maintaining enhanced biomarker detection capabilities
Solution Approach 2:
The patent creates spectral copies of the retinal image at different wavelength bands, allowing simultaneous visualization of multiple tissue layers and biomarkers. The processed multi-spectral images serve as digital copies that can be analyzed independently or combined, providing flexibility in diagnosis without requiring physical multiple imaging systems
3Productivity
If automated biomarker segmentation is implemented using deep learning approaches, then the productivity and diagnostic efficiency are improved, but the need for large datasets and computational resources increases
Solution Approach 1:
The patent applies preliminary image processing steps including noise reduction, contrast enhancement, and registration before feeding images to the deep learning model. This preprocessing prepares the data in an optimal format, reducing the amount of training data needed and improving model convergence speed while maintaining high diagnostic accuracy
Solution Approach 2:
The patent focuses the deep learning model on specific local features and biomarkers relevant to each disease type rather than attempting to analyze the entire image globally. This localized approach reduces computational complexity and allows effective training with smaller, more targeted datasets while maintaining high productivity in diagnostic workflows
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 rapid and efficient automated retinal disease biomarker detection and quantification, reducing the burden on ophthalmologists and improving diagnostic accuracy and efficiency.
Implementation Method 1
Conventional color fundus cameras use a white light source and collect an all-spectral combined color image. A multi-spectral retinal ophthalmoscope, in contrast, uses multiple individual illumination wavelengths across a wide wavelength range from visible to near infra-red (NIR).
Implementation Method 2
The multi-spectral retinal imaging (MSI) technique has proven to be an effective tool for enhanced visual identification of many of the above-mentioned biomarkers compared with conventional color fundus imaging.
Data Source
AI summary
A method and system for the use of multi-spectral retinal images to achieve an effective, efficient, and AI enabled automated retinal disease biomarker detection using biomarker identification, segmentation, and quantification. A plurality of digital multi-spectral images of a retina of a patient at a plurality of illumination wavelengths can be done using a multi-spectral ophthalmoscope. The images can then be registered, processed, and assessed to automatically quantifying one or more biomarker based on the location and size of the biomarker in the plurality of processed multi-spectral images.