DSA Image Subtraction Using ROI-Based Mask Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing DSA techniques face challenges in reducing artifacts due to positional shifts caused by body movement during subtraction processing between images before and after contrast medium injection, particularly in patients who cannot hold their breath, such as elderly or unconscious individuals.
Innovation Solution
An X-ray diagnostic apparatus and image processing system that selects an optimal mask image by correlating with non-contrast regions to minimize positional shifts, using a combination of manual, semi-automatic, and automatic modes for mask image selection, and performs subtraction processing to reduce artifacts from body movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional DSA subtraction processing is performed between mask images and contrast images, then blood vessels can be enhanced, but artifacts due to positional shifts caused by body movement occur
Solution Approach 1:
The patent segments the image into multiple regions of interest (ROIs) including a first ROI for blood vessels and a second ROI for surrounding tissues. By performing subtraction processing separately in each ROI and selectively combining results, the method isolates the blood vessel enhancement function from the body movement artifact problem, allowing precise handling of each region's specific characteristics
Solution Approach 2:
The patent applies different processing strategies to different regions: in the first ROI (blood vessels), standard subtraction is performed to enhance vessels, while in the second ROI (surrounding tissues), the selected mask image is used to suppress artifacts. This local differentiation allows each region to be optimized for its specific function, improving overall image quality while minimizing artifacts
2Object-affected harmful factors
If an optimal mask image is selected by comparing with contrast image upon localization to region with much movement, then motion artifacts can be suppressed, but great shift may occur due to blood vessel coincidence with bone structure
Solution Approach 1:
The patent divides the image into multiple ROIs with different purposes: the first ROI focuses on blood vessels where contrast enhancement is prioritized, while the second ROI focuses on surrounding tissues where motion artifact suppression is prioritized. This segmentation allows the system to select mask images optimized for each specific region without compromising overall accuracy
Solution Approach 2:
Different mask selection and subtraction strategies are applied locally to different ROIs. The first ROI uses processing optimized for vessel enhancement, while the second ROI uses processing optimized for artifact suppression. This local optimization prevents the great shift problem by ensuring that mask selection in tissue regions does not compromise blood vessel positioning accuracy
3Measurement precision
If multiple mask images are acquired and optimal mask is selected for each contrast image, then DSA image quality can be improved, but processing complexity increases
Solution Approach 1:
The patent segments both the image and the processing workflow into distinct ROIs and processing steps. By dividing the large-scale optimization problem into smaller regional sub-problems, the system can efficiently select optimal mask images for each ROI independently, reducing the overall computational complexity while maintaining high image quality
Solution Approach 2:
The patent performs preliminary processing steps including acquiring multiple mask images, dividing them into ROIs, and selecting optimal mask images before performing the final subtraction processing. This preliminary organization and selection reduces the complexity of the main processing task by pre-establishing the optimal processing path for each region
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An X-ray diagnostic apparatus includes an image generating unit (2) which generates a plurality of X-ray images by repeatedly radiographing a subject before and after injection of a contrast medium, a region detecting unit (20) which detects a non-contrast region from a plurality of mask images before injection of the contrast medium and a plurality of contrast images after injection of the contrast medium, which constitute the plurality of X-ray images, a mask selecting unit (19) which separately selects one mask image with respect to each of the contrast images on the basis of a correlation between the contrast image and the mask image upon localization to the non-contrast region, and a subtraction processing unit (21) which generates a subtraction image by performing subtraction between the contrast image and the selected mask image.