Intraluminal Image Abnormality Detection Using Gradient-Based Region Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image processing techniques for intraluminal images struggle to accurately detect abnormal parts, such as lesions, without mistakenly identifying groove positions or contour portions as abnormalities, due to the complexity of biological tissue structures.
Innovation Solution
An image processing apparatus and method that calculates gradient information, creates closed regions excluding pixels with high gradient strength and curved boundaries, and detects abnormalities within these regions to prevent false positives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional image processing techniques are used to detect abnormal parts in intraluminal images, then detection sensitivity is improved, but false positive rate increases due to misidentification of groove positions and contour portions as abnormalities
Solution Approach 1:
The image processing method segments the intraluminal image into multiple regions based on gradient information, identifying closed regions that are likely to contain abnormal parts while excluding regions with groove positions and contour portions. This segmentation allows differential processing that improves detection sensitivity for true abnormalities while reducing false positives from normal anatomical structures.
Solution Approach 2:
The patent applies different processing criteria to different regions of the image based on local gradient characteristics. Closed regions with specific gradient properties are identified as potential abnormality locations, while regions exhibiting gradient patterns characteristic of grooves or contours are excluded. This local quality approach enables the system to maintain high detection sensitivity in abnormal regions while suppressing false positives in normal regions.
2Reliability
If simple region-based detection is used, then false positives from groove positions are reduced, but detection precision of actual abnormalities decreases
Solution Approach 1:
The method performs preliminary analysis of gradient information across the entire image before final abnormality detection. By pre-identifying closed regions and evaluating their gradient characteristics, the system prepares region-specific detection criteria that maintain high precision for actual abnormalities while excluding false positive regions. This preliminary action enables subsequent detection to focus computational resources on promising regions with appropriate sensitivity thresholds.
Solution Approach 2:
The patent changes detection parameters based on regional characteristics. Different gradient strength thresholds, curvature criteria, and region selection parameters are applied to different closed regions based on their local image properties. This parameter adaptation allows the system to maintain high detection precision for true abnormalities in each region while systematically excluding regions with gradient patterns indicative of normal anatomical structures like grooves and contours.
Data Source
AI summary
An image processing apparatus includes: a gradient information calculating unit that calculates gradient information of each of pixels, based on pixel values of an intraluminal image; a closed region creating unit that, based on the gradient information, creates a closed region satisfying a condition where the closed region does not include, on the inside thereof, any pixel of which the gradient strength is equal to or higher than a predetermined value, and also, the boundary of the closed region does not curve toward the interior of the closed region, with a curvature equal to or larger than a predetermined value; and an abnormal part detecting unit that detects an abnormal part from the inside of the closed region.