Stratified Region Extraction for Coronary Artery Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for extracting specified regions in medical images, such as organs, are inaccurate and labor-intensive due to the complexity of organ shapes and the reliance on anatomical knowledge, leading to errors and inefficiencies, especially when applying image processing techniques like MIP processing.
Innovation Solution
A method and device that involve displaying an image, selecting a desired region, approximating partial contours using element graphics, and iteratively refining these approximations to create a first contour, with a second contour obtained based on the first, allowing for the extraction of a stratified region between the contours, enabling accurate and efficient image processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a simple graphic is used to approximate organ contours, then the operation becomes simpler, but the extraction accuracy deteriorates due to the complicated shapes of organs
Solution Approach 1:
The patent divides the organ contour approximation into multiple segments by using multiple element graphics (e.g., multiple ellipses) instead of a single simple graphic. Each element graphic approximates a specific portion of the organ contour, and these segments are combined to form the complete first contour, thereby maintaining simplicity while improving accuracy for complex organ shapes
Solution Approach 2:
The patent introduces a new dimension to the approximation process by creating a stratified region between the first contour (approximated by element graphics) and the second contour (obtained through expansion). This additional dimensional space allows for more precise region extraction while maintaining the simplicity of the base approximation operation
2Loss of information
If MIP processing is applied to extract coronary arteries, then the blood vessel information can be obtained, but the cardiac internal region creates high signal interference that makes the coronary artery invisible
Solution Approach 1:
The patent extracts and removes the cardiac internal region from the analysis by defining a first contour that approximates the organ surface and a second contour that expands from it. The stratified region between these contours specifically isolates the surface area where coronary arteries are located, effectively separating the coronary artery region from the high-signal cardiac internal region, thereby enabling clear MIP processing of the coronary arteries
3Adaptability or versatility
If manual approximation of organ contours is performed, then the extraction can be adapted to individual cases, but the operation becomes repetitive and time-consuming
Solution Approach 1:
The patent uses parameter changes by allowing the element graphics (e.g., ellipses) to be adjusted in terms of position, size, and shape parameters to fit different organ contours. This enables the system to adapt to various organ shapes and sizes automatically, maintaining flexibility while reducing the time required compared to manual approximation, as the parameters can be modified through automated algorithms rather than repetitive manual operations
Data Source
AI summary
A region-extraction method for extracting a contour of a desired region is provided.The method includes: a step (a) for displaying an image; a step (b) for selecting a desired region in the image; a step (c) for selecting an element graphic corresponding at least a partial contour in a partial region in the desired region; a step (d) for approximating the contour of a part of the element image to at least a partial contour in the partial region; a step (e) for repeating the steps (c) to (d) at least twice; and a step (f) for making a first contour by combining at least a part of elements after the approximation.