Confidence Map for Limited Angle CT Artifact Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
C-arm imaging systems with limited angular ranges in orthopedic interventions face challenges in data acquisition, leading to incomplete trajectories and severe image quality issues due to limited angle artifacts in image reconstruction.
Innovation Solution
An image processing system that includes a visualizer to display a confidence map and a directional indicator, utilizing a machine learning algorithm to distinguish border regions and indicate missing projection directions, thereby improving image reconstruction accuracy in limited angle tomography.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Volume of moving object
If the angular range of rotation is reduced to achieve a slim design and small footprint, then the system becomes more compact and easier to handle, but limited angle artifacts occur and image quality deteriorates
Solution Approach 1:
The patent introduces a confidence map as an intermediary visual aid that does not physically alter the C-arm system but mediates between the limited angle scan constraints and the image quality requirements. The confidence map provides additional information to users about reconstruction reliability, allowing them to interpret images more accurately despite the angular limitations.
Solution Approach 2:
The confidence map provides feedback to users about the reliability of different regions in the reconstructed image. By visualizing uncertainty levels, the system feedbacks information about which areas should be interpreted with caution, effectively closing the loop between the limited angle acquisition constraints and the final image interpretation.
2Length of moving object
If the angular range of rotation is reduced below 180 degrees, then the system design becomes more compact, but incomplete trajectory leads to limited angle artifacts in image reconstruction
Solution Approach 1:
The confidence map serves as an intermediary that bridges the gap between the incomplete trajectory data and the reconstructed image. It provides a visual representation of where the reconstruction may be unreliable due to the limited angular range, without requiring the physical system to be changed.
Solution Approach 2:
The patent changes the parameter representation by introducing a confidence metric that quantifies reconstruction reliability. This parameter change allows the system to maintain the compact angular range while providing information about reconstruction quality through the confidence map visualization.
3Loss of information
If border regions are reconstructed from limited angle data, then the complete anatomical structure is visualized, but the border portions may be incorrectly represented due to missing projections
Solution Approach 1:
The confidence map acts as an intermediary layer between the reconstructed image and the user interpretation. It provides additional information about border region reliability without altering the anatomical coverage achieved through the limited angle scan.
Solution Approach 2:
The confidence map applies local quality assessment by differentiating between reliable and unreliable regions within the image. Border regions with missing projections are marked with low confidence values, allowing users to understand that different parts of the image have different levels of accuracy.
4Manufacturing precision
If a machine learning estimator is used to improve reconstruction from limited angle data, then image quality improves, but the system complexity increases
Solution Approach 1:
The patent replaces traditional mechanical/image processing reconstruction methods with a machine learning estimator. This substitution allows the system to achieve better reconstruction accuracy from limited angle data by using learned patterns rather than conventional algorithms.
Solution Approach 2:
The confidence map serves as an intermediary that simplifies the interpretation of complex machine learning outputs. While the estimator adds complexity, the confidence map provides a straightforward visual representation that helps users understand the reliability of different image regions.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Image processing system (IPS), comprising an input interface (IN) for receiving an input image (IM) based on projection data (π) collected in a limited angle scan along different projection directions by an imaging apparatus (IA). A directional analyzer (DA) to compute a direction component for border regions in the input image. A directional discriminator (DD) is to discriminate border regions based on whether or not their direction component is along at least one of the projection directions.