Fundoscopic Image ROI Pre-Processing for Lesion Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing computer-assisted evaluation of medical images, particularly fundoscopic images, faces challenges in efficiency, complexity, and accuracy for detecting lesions such as retinopathy and macular disease, with a need for improved image pre-processing to enhance features and reduce noise.
Innovation Solution
A method and apparatus that focuses on a region of interest in fundoscopic images, applies smoothing processes, and adjusts pixel value differences to enhance features, using techniques like Gaussian blur and value enhancement, resulting in a normalized and enhanced image for improved recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual judgment of lesions by doctors is used, then diagnostic accuracy can be maintained, but evaluation efficiency is low and complexity is high
Solution Approach 1:
The patent segments the image processing into distinct modules: region of interest extraction, smoothing processing, and pixel value enhancement. This segmentation allows each module to perform a specific function independently, improving overall efficiency while keeping the complexity of individual modules manageable.
Solution Approach 2:
The patent extracts the region of interest (eyeball) from the full fundoscopic image and processes only this extracted region. This extraction significantly reduces the amount of data that needs to be processed, thereby improving evaluation efficiency while reducing the complexity of the processing system.
2Measurement precision
If image pre-processing is applied to enhance features, then recognition accuracy improves, but processing time increases
Solution Approach 1:
The patent applies smoothing and pixel value enhancement as preliminary actions before the main recognition task. By performing these actions in advance on the extracted region of interest, the system prepares the image data to improve recognition accuracy while minimizing the time required during the actual recognition phase.
Solution Approach 2:
The patent applies different processing qualities to different regions - specifically, it processes only the region of interest (eyeball) with enhanced smoothing and pixel value adjustment, while leaving the rest of the image unprocessed. This local quality approach improves recognition accuracy for the critical region without unnecessarily processing the entire image, thus reducing overall processing time.
3Loss of information
If the entire fundoscopic image is processed, then complete information is available, but processing complexity and time increase
Solution Approach 1:
The patent extracts only the region of interest (eyeball) from the complete fundoscopic image and processes this extracted region. This extraction maintains the necessary information for lesion detection while significantly reducing the processing workload, thereby improving processing speed without losing critical diagnostic information.
Solution Approach 2:
The patent segments the image processing into two parts: extraction of the region of interest and processing of this segmented region. This segmentation allows the system to maintain information completeness for the critical area while processing only the necessary portion, thus improving overall processing speed.
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
An image pre-processing method and an image processing apparatus for a fundoscopic image are provided. A region of interest (ROI) is obtained from a fundoscopic image to generate a first image. The ROI is focused on an eyeball in the fundoscopic image. A smoothing process is performed on the first image to generate a second image. A value difference between neighboring pixels in the second image is increased to generate a third image.