Edge Preserving Regularization for Medical Image Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Iterative image reconstruction methods in medical imaging, such as PET and SPECT, are prone to introducing spurious high-intensity features that can be misinterpreted as malignant tumors, and conventional parameter tuning for edge-preserving regularization is time-consuming and lacks transparency, making it difficult to achieve optimal image quality and quantitation.
Innovation Solution
An optimized system for determining weighting and edge sensitivity parameters for edge-preserving regularization in image reconstruction, allowing for user input-based noise reduction and quantitation goals, using lookup tables or calibration curves to simplify parameter tuning and ensure physician-preferred image quality and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If iterative reconstruction is used to improve image quality, then image reconstruction quality is improved, but spurious high-intensity features are introduced that can be misinterpreted as malignant tumors
Solution Approach 1:
The patent applies edge-preserving regularization with specific parameters (β for overall weighting and γ for edge sensitivity) to modify the reconstruction process. By adjusting these parameters, the system suppresses spurious high-intensity features while preserving true edges and structures, thus improving measurement precision without introducing harmful artifacts
Solution Approach 2:
The regularization term acts as an intermediary between the iterative reconstruction process and the final image output. It mediates by introducing prior knowledge about edge preservation, which suppresses noise and spurious features while maintaining true anatomical structures, thereby resolving the contradiction between image quality and artifact introduction
2Ease of operation
If conventional parameter tuning for edge-preserving regularization is performed manually, then parameter adjustment flexibility is improved, but the process is time-consuming and lacks transparency
Solution Approach 1:
The system performs self-service by automatically determining optimal values for the regularization parameters β and γ based on the imaging data and clinical protocols. This eliminates the need for manual parameter tuning, reducing time loss while maintaining flexibility through algorithmic adaptation to different imaging scenarios
Solution Approach 2:
The patent implements automatic parameter determination through algorithms that select appropriate β and γ values based on data characteristics and clinical requirements. This automated parameter change process maintains the flexibility needed for different imaging scenarios while eliminating the time-consuming manual tuning process
3Object-generated harmful factors
If edge-preserving regularization is applied to suppress small-volume high-intensity features, then artifact reduction is improved, but the ability to detect small lesions may be reduced
Solution Approach 1:
The patent carefully selects the edge sensitivity parameter γ to control the threshold at which features are suppressed. By optimizing this parameter, the system distinguishes between spurious artifacts and true small lesions, maintaining artifact reduction while preserving the detection capability for clinically relevant small-volume features
Solution Approach 2:
The edge-preserving regularization applies different treatment to different regions of the image based on local characteristics. True edges and structures are preserved while spurious features are suppressed, allowing the system to maintain small lesion detection capability in regions where it is needed while reducing artifacts in regions where they occur
Data Source
AI summary
A non-transitory computer-readable medium stores instructions readable and executable by a workstation (18) including at least one electronic processor (20) to perform an image reconstruction method (100). The method includes: determining a weighting parameter (13) of an edge-preserving regularization or penalty of a regularized image reconstruction of an image acquisition device (12) for an imaging data set obtained by the image acquisition device; determining an edge sensitivity parameter (γ) of the edge-preserving algorithm for the imaging data set obtained by the image acquisition device; and reconstructing the imaging data set obtained by the image acquisition device to generate a reconstructed image by applying the regularized image reconstruction including the edge-preserving regularization or penalty with the determined weighting and edge sensitivity parameters to the imaging data set obtained by the image acquisition device.