Medical Image Enhancement via Edge Detection and Local Histogram Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Medical imaging technologies like CT and X-ray generate ionizing radiation, leading to challenges in visualizing metal objects and other dense objects in low-dose, low-quality images, where they often appear faint or obscured in composite images.
Innovation Solution
A method involving edge detection, offset aggregation, and histogram correction to enhance the appearance of objects in medical images, including metal objects, by generating strength and index images, calculating offset and directional correlation images, and merging these with baseline images to improve image quality and visibility of objects of interest.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If low-dose imaging is used to reduce radiation exposure, then patient safety is improved, but image quality deteriorates and metal objects become faint or obscured
Solution Approach 1:
The patent merges a low-dose overlay image containing surgical tools with a baseline image containing anatomical details through composite image generation. The blending algorithm combines corresponding pixels from both images, allowing the overlay image's metal objects to become visible while the baseline image's anatomical structures are preserved, thereby solving the problem of faint metal objects in low-dose imaging without increasing radiation exposure.
Solution Approach 2:
The patent applies local histogram correction to specific regions of the composite image where metal objects are located. By detecting edges and identifying regions containing surgical tools, the system selectively enhances contrast and brightness in those local areas while preserving the overall low-dose imaging benefits, making metal objects visible without requiring high radiation exposure across the entire image.
2Device complexity
If standard flat blend of baseline and overlay images is used, then image processing is simplified, but metal objects appear faint and are difficult to detect
Solution Approach 1:
The patent performs preliminary edge detection and region identification on the overlay image before blending with the baseline image. By pre-identifying regions containing metal objects through edge detection algorithms, the system can apply targeted enhancement strategies during the blending process, ensuring metal objects are properly visualized while maintaining relatively simple overall processing complexity.
Solution Approach 2:
The patent applies local histogram correction specifically to regions identified as containing metal objects, rather than uniformly processing the entire image. This localized approach enhances the visibility of surgical tools in critical areas while keeping the processing complexity manageable by avoiding exhaustive analysis of every pixel in the image.
3Productivity
If conventional composite image methods are used, then processing speed is maintained, but visibility of objects of interest is insufficient
Solution Approach 1:
The patent performs preliminary edge detection and region classification on the overlay image before the blending operation. By pre-identifying which regions contain metal objects and which contain anatomical structures, the system can efficiently apply appropriate enhancement strategies during blending, maintaining processing speed while significantly improving object visibility through targeted rather than exhaustive processing.
Solution Approach 2:
The patent dynamically adjusts blending parameters and histogram correction parameters based on the detected content of different image regions. By changing parameters such as blend ratios and enhancement intensity according to whether a region contains metal objects or anatomical structures, the system maintains fast processing through efficient parameter selection while achieving superior object visibility compared to fixed-parameter methods.
Data Source
AI summary
Disclosed herein are systems and methods for enhancement of objects of interest in medical images.