Radiographic Imaging Apparatus Feature Point Distance Weighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing radiographic imaging apparatuses face challenges in generating clear enhanced images of moving stent objects during intravascular interventions due to increased processing load when weighting X-ray images based on similarity information, leading to blurred sequences.
Innovation Solution
A radiographic imaging apparatus that synthesizes radiation images in real time by weighting them based on feature point distances between reference and other images, aligning with heartbeat phases to enhance the clarity of the enhancement target object while reducing processing load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If X-ray images are weighted based on similarity information, then image enhancement quality improves, but processing load increases
Solution Approach 1:
The patent segments the image enhancement process by dividing it into two distinct stages: (1) a registration stage that aligns images using only feature point distances without weighting, and (2) a weighting stage that applies similarity-based weighting only to the registered images. This segmentation allows the computationally intensive weighting operation to be performed on fewer, pre-aligned images, thereby reducing overall processing load while maintaining enhancement quality.
Solution Approach 2:
The patent performs preliminary image registration using feature point distance calculations before applying the computationally expensive similarity-based weighting. By pre-aligning the images through registration, the system prepares the data in advance so that subsequent weighting operations can be performed more efficiently, reducing the total processing load while preserving image enhancement quality.
2Productivity
If X-ray images are simply temporally integrated, then processing load is reduced, but image clarity deteriorates due to stent shape changes
Solution Approach 1:
The patent performs preliminary registration of X-ray images based on feature point distances before temporal integration. This preliminary alignment ensures that images captured at different heartbeat phases are properly positioned relative to each other, preventing the blurring that would otherwise occur due to stent shape changes. As a result, the system can achieve clear enhanced images without requiring computationally intensive similarity-based weighting for every image.
Solution Approach 2:
The patent introduces feature point distance-based registration as an intermediary step between simple temporal integration and full similarity-based weighting. This intermediary approach provides sufficient alignment to maintain image clarity during temporal integration while avoiding the excessive processing load of applying full similarity weighting to all images, thus bridging the gap between processing efficiency and image quality.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach allows for accurate weighting and generation of enhanced images with reduced processing load, ensuring the enhancement target object is clearly enhanced even when its shape changes with heartbeats, thereby improving image clarity and efficiency.
Implementation Method 1
an image generation unit that generates a radiation image based on radiation transmitted through a subject
Data Source
AI summary
A radiographic imaging apparatus includes an image generation unit that generates radiation images, and an image processing unit that synthesizes a plurality of radiation images to generate an enhanced image. The image processing unit is configured to synthesize the plurality of radiation images by weighting each of the plurality of radiation images in accordance with a feature point distance.