Medical Image Abnormality Detection Using Connected Blocks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting anatomical structure abnormalities in medical images struggle to accurately distinguish between normal and abnormal cases when an abnormality overlaps with a normal structure, such as a tumor or water, due to image segmentation limitations.
Innovation Solution
A medical image abnormality detection system that determines anatomical structure abnormalities based on the number of blocks in a medical image, where a single block indicates a normal structure, and multiple blocks with specific conditions (e.g., area and distance) indicate abnormalities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If image segmentation is performed to separate the region of interest and determine abnormality based on characteristics such as barycentric position, volume, shape, intensity, and density, then the detection process can identify anatomical structures, but it becomes impossible to distinguish abnormal cases when an abnormality such as a tumor or water overlaps part of a normal structure of an organ
Solution Approach 1:
The patent applies segmentation by dividing the anatomical structure into multiple connected component blocks based on pixel connectivity. Instead of treating the region of interest as a single segmented entity, the method further segments it into discrete blocks that can be individually analyzed. This allows the system to detect whether the anatomical structure is normally connected or abnormally disconnected, improving accuracy in cases where abnormalities overlap with normal structures.
2Reliability
If conventional image segmentation methods are used to detect abnormalities, then the process is relatively simple, but the accuracy is insufficient when abnormalities overlap with normal anatomical structures
Solution Approach 1:
The patent introduces a new dimension of analysis by examining the connectivity topology of the anatomical structure. Instead of relying solely on traditional characteristics like intensity and density, the method analyzes the spatial connectivity by identifying connected components and their block structures. This topological dimension provides additional information that improves detection reliability without requiring complex multi-modal imaging.
3Measurement precision
If the anatomical structure is determined to be abnormal when multiple blocks are detected, then overlapping abnormalities can be identified, but false positives may occur without additional validation conditions
Solution Approach 1:
The patent applies local quality by examining specific properties of individual blocks and their relationships. Instead of making a global determination based solely on the number of blocks, the method evaluates local characteristics such as block size, shape, intensity distribution, and spatial relationships. This allows the system to distinguish true abnormalities from false positives caused by normal anatomical variations or imaging artifacts.
Data Source
AI summary
An abnormality detection system is a system that detects, based on a medical image, whether a prespecified anatomical structure included in the medical image is abnormal, and includes: an obtainer that obtains block information indicating one or more blocks that are each a region in which pixels representing the anatomical structure are connected in the medical image; and a determiner that determines and outputs whether the anatomical structure is abnormal, based on the number of blocks indicated by the block information obtained by the obtainer, wherein the determiner determines that the anatomical structure is not abnormal when the number of blocks indicated by the block information is 1, and determines that the anatomical structure is abnormal when the number of blocks indicated by the block information is 2 or more and the two or more blocks indicated by the block information satisfy a predetermined condition.