Medical Image Reconstruction Using Patch-Based Low Rank Regularization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current medical imaging technologies face challenges in achieving high spatial and temporal resolution while minimizing artifacts, particularly in reconstructing images of moving objects like the heart, which affects the quality of cardiac CT images.
Innovation Solution
A method and apparatus for reconstructing medical images that involve acquiring reference and reconstructed images based on ECG data, using patch-based low rank regularization to update data in unit regions by generating matrices and determining eigenvalues, and minimizing cost functions with a shrinkage operator.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional image reconstruction methods are used to achieve high spatial resolution, then image detail quality improves, but temporal resolution deteriorates and motion artifacts increase
Solution Approach 1:
The patent segments the cardiac cycle into multiple phases based on ECG data, and further segments the image into multiple patches or regions. By processing each patch independently with reference to corresponding patches from other phases, the method achieves high spatial resolution within each patch while maintaining temporal resolution through phase-specific reconstruction.
Solution Approach 2:
The patent performs preliminary actions by acquiring reference images from multiple cardiac phases before reconstruction. These reference images are pre-processed to identify corresponding patches, and reconstruction matrices are pre-calculated based on similarity metrics. This preliminary preparation enables efficient reconstruction that maintains both high spatial and temporal resolution without compromising image quality.
2Loss of time
If image reconstruction is performed to capture moving structures like the heart, then temporal resolution improves, but image quality deteriorates due to motion artifacts
Solution Approach 1:
The patent applies local quality by treating different regions of the cardiac image differently based on their motion characteristics. Each patch is reconstructed using reference patches from corresponding anatomical locations in other cardiac phases. This localized approach maintains high temporal resolution for capturing motion while preserving image quality by ensuring each region is reconstructed from appropriate reference data.
Solution Approach 2:
The patent changes parameters by using ECG-gated timing information to select appropriate reference phases for reconstruction. The similarity metric between patches is dynamically adjusted based on cardiac phase, and reconstruction weights are modified according to motion magnitude. These parameter changes enable the system to maintain both temporal resolution and image quality by adapting to the specific motion state at each moment.
3Manufacturing precision
If patch-based reconstruction with reference regions is used, then artifacts are reduced and image quality improves, but computational complexity increases
Solution Approach 1:
The patent segments the full image into multiple smaller patches, which reduces the computational burden compared to processing the entire image at once. Each patch can be independently reconstructed using reference patches from other cardiac phases. This segmentation enables artifact reduction through localized reference-based reconstruction while managing computational complexity by dividing the problem into smaller, parallelizable units.
Solution Approach 2:
The patent applies partial action by selecting only the most relevant reference patches based on similarity metrics, rather than using all available reference data. The reconstruction process focuses on patches with high similarity to reference patches, applying full reference-based reconstruction only where beneficial. This selective approach reduces computational complexity while maintaining image quality by avoiding unnecessary processing of low-similarity regions.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method of reconstructing a tomographic image in a medical imaging apparatus includes: acquiring a first image of an object and a second image of the object corresponding to an image to be reconstructed; determining, in the first image, a reference region corresponding to a unit region in the second image; and updating data in the unit region based on data contained in the reference region. Thus, a medical image having both high spatial resolution and high temporal resolution is reconstructed to have substantially reduced artifacts.