4D-CT Image Registration Using Reference Mediators for Large Organ Integration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
4D-CT devices face challenges in integrating images of larger organs without sufficient overlap between frame images, leading to precision issues in registration and integration.
Innovation Solution
An image processing apparatus and method that acquire and register frame images with corresponding phases from different imaging ranges, using reference images and registration information to align images even without overlap, by calculating registration information between frame images and reference images, and combining them based on this information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If multiple 4D-CT images are captured and integrated to observe large organs, then the imaging range is extended, but registration precision deteriorates due to insufficient overlap regions between frame images
Solution Approach 1:
The patent introduces reference images as intermediary elements that have overlap regions with both the first and second frame images. These reference images serve as mediators to establish registration relationships between frame images that may not directly overlap, thereby maintaining registration precision while extending the imaging range.
Solution Approach 2:
The patent extends registration from direct frame-to-frame comparison to a multi-dimensional approach involving frame images, reference images, and multiple phases. By introducing the dimension of reference images and utilizing overlap regions in different phases, the system achieves precise registration across larger imaging ranges.
2Device complexity
If frame images without overlap regions are registered directly, then processing is simplified, but registration precision deteriorates
Solution Approach 1:
Reference images are introduced as intermediaries to facilitate registration between frame images without direct overlap. This approach increases processing complexity but significantly improves registration precision by providing reliable reference points and overlap regions for accurate alignment.
Data Source
AI summary
An image processing apparatus configured to acquire first registration information being registration information between a first image of interest, which is the first frame image in the frame image pair, and the first reference image; to acquire second registration information being registration information between a second image of interest, which is the second frame image in the frame image pair, and the second reference image; to acquire reference registration information being registration information between the first reference image and the second reference image; and to acquire third registration information being registration information between the first image of interest and the second image of interest, based on the first registration information, the second registration information and the reference registration information.