Handheld Multispectral Fundus Imager for Drusen Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current fundus imagers often fail to detect drusen and other eye diseases due to limitations in color fundus images, and there is low compliance and poor access to eye exams, leading to undetected diseases.
Innovation Solution
A handheld fundus imager that captures multispectral images using LEDs and spectral filters, allowing for the detection of details not visible in color images, and provides a diagnosis or recommendation for further eye care.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If color fundus images are used for eye disease screening, then the device is simple and easy to operate, but drusen and other eye diseases cannot be detected
Solution Approach 1:
The imaging system is segmented into multiple spectral channels, capturing images at different wavelengths (e.g., 450nm, 550nm, 650nm) separately. This allows extraction of specific spectral information from each channel that highlights different pathological features, thereby improving detection accuracy without requiring a completely complex new system design.
Solution Approach 2:
The system transitions from capturing only spatial information (2D color images) to capturing spectral information as an additional dimension. By adding the spectral wavelength dimension, the system can differentiate tissues based on their spectral signatures, enabling detection of drusen and other diseases that are invisible in conventional color images.
2Measurement precision
If multispectral imaging is implemented to detect drusen and eye diseases, then detection accuracy improves, but device complexity increases
Solution Approach 1:
The fundus imager is designed as a multi-functional device that can perform both conventional color fundus imaging and multispectral imaging modes. This universality allows the system to maintain simplicity for routine examinations while enabling enhanced disease detection when needed, without requiring completely separate simple and complex systems.
Solution Approach 2:
The system changes the illumination wavelength parameter to achieve different imaging modes. By adjusting the spectral characteristics of the light source (using LED arrays with different peak wavelengths), the same hardware can operate in simplified color mode or enhanced multispectral mode, managing complexity through parameter control rather than hardware changes.
3Ease of operation
If conventional color fundus imaging is used, then the device is portable and accessible, but diseases like macular degeneration go undetected
Solution Approach 1:
The system incorporates automated image analysis algorithms that process the captured multispectral images and automatically identify pathological features such as drusen. This self-service capability reduces the need for expert interpretation, maintaining ease of operation and accessibility while improving detection precision through computational analysis.
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 efficient detection of eye diseases like macular degeneration by intensifying contrast between healthy tissue and lesions, facilitating accurate analysis and diagnosis, even by non-experts, and improving access to eye care through portable and user-friendly technology.
Implementation Method 1
a lighting unit configured to illuminate an eye fundus, the lighting unit including one or more light-emitting diodes
Implementation Method 2
filtering light reflected from the eye fundus to capture a multispectral fundus image
Data Source
AI summary
A fundus imager includes a handheld housing that supports a lighting unit configured to illuminate an eye fundus. The lighting unit includes one or more light-emitting diodes. The housing further supports a camera configured to capture one or more images of the eye fundus, and a display configured to display the one or more images of the eye fundus. The fundus imager captures at least one multispectral fundus image using the camera, and displays the at least one multispectral fundus image on the display.