Context-Oriented Iterative CT Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current computed tomography (CT) image reconstruction methods using spatially-varying regularization parameters often result in uniform noise distribution across the entire image, which may not be desirable in clinical practice, as it can lead to over-smoothing in high-contrast regions and inadequate noise suppression in low-contrast regions, requiring multiple reconstructions for optimal results.
Innovation Solution
The method employs a context/content-oriented spatially-varying regularization parameter that varies the degree of smoothing based on the content of different organs/regions within an image, using a product of spatially-invariant, statistically-based, and context-oriented coefficients to achieve intra-organ uniformity and inter-organ diversity in noise distribution, allowing for a single reconstruction protocol that optimizes noise and resolution for multiple organs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a spatially-varying regularization parameter is used to provide uniform resolution or uniform statistical properties throughout the reconstructed image, then the image quality is improved in terms of uniformity, but the image may not have the best qualities for a particular application due to over-smoothing in high-contrast regions and inadequate noise suppression in low-contrast regions
Solution Approach 1:
The patent applies local quality by using a context-oriented spatially-varying regularization parameter that adapts to different tissue types and anatomical regions. The regularization parameter is modulated based on local image characteristics such as tissue density, contrast, and anatomical context, allowing different regions to have optimized smoothing properties. This resolves the contradiction by maintaining uniform statistical properties while also optimizing for application-specific qualities through localized adaptation.
Solution Approach 2:
The patent implements dynamics by making the regularization parameter adaptive rather than static. The parameter dynamically adjusts based on the reconstructed image content, tissue type, and application context. This dynamic adjustment allows the system to transition between uniform smoothing and application-optimized smoothing, resolving the contradiction between uniformity and application-specific quality requirements.
2Productivity
If a single regularization parameter is used for the entire image, then the reconstruction process is simple and fast, but it results in over-smoothing in high-contrast regions and inadequate noise suppression in low-contrast regions
Solution Approach 1:
The patent applies local quality by implementing a spatially-varying regularization parameter that is computed efficiently based on pre-defined tissue masks and anatomical regions. Different regularization strengths are assigned to different regions (e.g., lungs, soft tissue, bone) based on their specific imaging requirements. This allows region-specific optimization without significantly increasing computational complexity, as the spatial variation is based on pre-computed anatomical information rather than iterative optimization.
Solution Approach 2:
The patent implements preliminary action by pre-segmenting the image into anatomical regions and pre-defining tissue masks before the reconstruction process. This preliminary classification allows the system to apply appropriate regularization parameters to different regions during reconstruction, achieving region-specific quality optimization without adding significant computational burden during the actual reconstruction phase.
3Manufacturing precision
If multiple reconstructions are performed with different regularization parameters for different organs, then optimal image quality is achieved for each organ, but the clinical workflow becomes complex and time-consuming
Solution Approach 1:
The patent applies universality by creating a single multi-functional reconstruction algorithm that handles multiple organs and tissue types simultaneously. The context-oriented spatially-varying regularization parameter automatically adapts to different anatomical regions within a single reconstruction process, eliminating the need for separate reconstructions for each organ. This maintains organ-specific optimization while simplifying the clinical workflow into a single universal process.
Solution Approach 2:
The patent implements merging by combining multiple organ-specific reconstruction processes into a single unified reconstruction. The spatially-varying regularization parameter integrates information from multiple anatomical regions and tissue types, applying appropriate smoothing to each region within a single computational framework. This merges the benefits of multiple specialized reconstructions into one efficient process, reducing clinical workflow complexity.
Data Source
AI summary
A method and apparatus is provided to iteratively reconstruct a computed tomography (CT) image using a spatially-varying content-oriented regularization parameter, thereby achieving uniform statistical properties within respective organs/regions and different statistical properties (e.g., degree of smoothing and noise level) among the respective organs/regions. For example, less smoothing and sharper features/resolution can be applied within a lung region than within a soft-tissue region by using a smaller regularization parameter value in the lung region than in the soft-tissue region. This can be achieved, e.g., using a minimum intensity projection to suppress/eliminate sub-solid nodules in the lung region. The content-oriented regularization parameter can be generated by reconstructing an initial CT image, which is then segmented/classified according to organs and/or tissue type. Segmenting the image and generating the content-oriented regularization parameter can be integrated into one process by applying an HU-to-β mapping to the CT image.