Handheld Retinal Imaging Stabilization via AI Fixation Adjustment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Handheld retinal imaging devices face challenges in capturing stable images due to unwanted movement, leading to undesired positions and blurred images, as existing stabilization methods are either not applicable, too complex, or fail to provide the required image quality.
Innovation Solution
A system comprising a handheld retinal imaging device with an external computing device that uses a 3-D accelerometer and artificial intelligence to stabilize the image by adjusting the fixation stimulus and illuminating the eye with infrared and visible light sources, ensuring accurate capture of retinal images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If handheld retinal imaging device is used for mobility and ease of operation, then ease of operation is improved, but image stability deteriorates due to unwanted movement
Solution Approach 1:
The system uses real-time feedback from the camera to detect eye position and movement, then dynamically adjusts the fixation stimulus position on the display screen to compensate for device movement. This closed-loop feedback mechanism maintains image stability while preserving the handheld device's mobility and ease of operation.
Solution Approach 2:
The patent replaces mechanical stabilization systems (such as gimbals or motorized lens stabilization) with a software-based solution that uses camera imaging, AI analysis, and display feedback to achieve image stabilization. This eliminates complex mechanical components while maintaining stabilization effectiveness.
2Stability of the object's composition
If Optical Image Stabilization with motors is used to stabilize image, then image stability is improved, but device complexity increases
Solution Approach 1:
The system replaces motorized optical stabilization mechanisms with a software-based stabilization approach using camera imaging, AI-driven eye tracking, and dynamic fixation stimulus adjustment. This eliminates the need for motors, lenses, and complex calibration systems while achieving comparable or superior stabilization.
Solution Approach 2:
The patent introduces an intermediary software layer that processes camera images and controls the fixation stimulus position, acting as a mediator between the handheld device movement and the final retinal image. This software intermediary replaces the need for direct mechanical intervention in the optical path.
3Stability of the object's composition
If Digital Image Stabilization by cropping peripheral regions is used, then image stability is improved, but image quality deteriorates due to loss of peripheral area
Solution Approach 1:
Instead of cropping the image periphery to achieve stabilization, the system inverts the approach by dynamically repositioning the fixation stimulus within the full image frame. This allows the entire image area to be utilized while maintaining stability, avoiding the quality loss associated with peripheral cropping.
4Measurement precision
If AI-based real-time video analysis is used to determine stable image, then measurement precision is improved, but use of energy increases
Solution Approach 1:
The system performs AI analysis on selected frames or reduced-resolution video data rather than processing every frame at full resolution. This partial processing approach maintains sufficient measurement precision for eye tracking while significantly reducing computational energy consumption compared to exhaustive frame-by-frame 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
The system effectively stabilizes retinal images by compensating for device movement, ensuring accurate positioning and quality, overcoming the limitations of existing stabilization methods.
Implementation Method 1
The external computing device is configured for receiving position data from a 3-D accelerometer
Implementation Method 2
conveying a trigger to a light source of the handheld retinal imaging device for illuminating the eye of the subject
Data Source
AI summary
Disclosed is a system for retinal imaging. The type of image to be captured is selected by a user. Corresponding fixation stimulus is displayed on a fixation screen of the disclosed device for the subject to fix their gaze on. The device comprises illumination sources and, illumination optics and imaging optics, a 3-D accelerometer, and a camera. Signal from the 3-D accelerometer, and the video of the eye of the subject captured by camera are used by a computing device that uses AI techniques to vary the position of the fixation stimulus to obtain a stabilized image of the eye of the subject. The position of the fixation stimulus is varied to compensate for the movement of the handheld device to obtain a stabilized retinal image. Once stabilized, the image is captured.