Local Non-Uniformity Assessment for Nuclear Imaging Artifacts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In nuclear medical imaging, reconstructed images often contain artifacts due to motion or poor calibration, which can be difficult to recognize and correct, especially with traditional global histogram analysis that is not sensitive to local artifacts.
Innovation Solution
The method assesses reconstruction quality by calculating local non-uniformity in the modified cumulative distribution function (MCDF) across pixels or voxels, providing a quantitative measure that aids radiologists in identifying and addressing image degradation, and recommending alternative reconstruction approaches.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If global histogram distribution (MCDF) is used for artifact detection, then overall reconstruction quality can be assessed, but local artifacts and image degradations cannot be detected
Solution Approach 1:
The patent divides the global histogram analysis into local region analyses. Instead of computing a single MCDF for the entire image, the method segments the image into multiple local regions and computes separate MCDFs for each region. This segmentation enables detection of local artifacts that would be invisible in global histogram analysis, directly resolving the contradiction between detection sensitivity and analysis complexity.
Solution Approach 2:
The patent transitions from analyzing only the distribution values in the MCDF to analyzing the spatial distribution of these values across the image. By adding the spatial dimension to the histogram analysis, local variations in artifact presence become detectable. This dimensional expansion allows the system to maintain relatively simple histogram computation while gaining the ability to detect local artifacts through spatial mapping of MCDF values.
2Reliability
If different reconstruction methods are applied to correct artifacts, then image quality may be improved, but it becomes difficult to recognize when artifacts exist or correction is needed
Solution Approach 1:
The patent implements a feedback mechanism where local MCDF values are computed and mapped back to the image space, providing visual feedback to radiologists about regions with potential artifacts. This feedback loop allows radiologists to easily identify problematic regions without requiring expert knowledge of histogram analysis, making artifact detection straightforward and enabling informed decisions about whether alternative reconstruction methods are needed.
Solution Approach 2:
The patent introduces an intermediary visualization layer that translates complex MCDF statistical measures into an easily interpretable spatial map. This intermediary representation serves as a bridge between the complex reconstruction process and the radiologist's visual assessment, making artifact detection accessible without requiring specialized expertise in histogram analysis or reconstruction algorithms.
3Measurement precision
If MCDF rendering is provided for expert observation, then artifacts may be recognized by experienced observers, but other radiologists have difficulty recognizing artifacts even in rendered MCDF
Solution Approach 1:
The patent applies local quality by computing and displaying MCDF values specifically for local regions rather than providing only a global histogram. This allows the visualization to highlight areas with local artifacts while maintaining overall image context. The local MCDF analysis provides detailed information about specific problem regions without overwhelming less experienced radiologists with global statistical complexity, thereby improving both detection accuracy and usability.
Data Source
AI summary
Reconstruction quality is assessed in medical imaging. An amount of local non-uniformity in a distribution of a statistical measure (e.g., MCDF) is determined. The amount indicates a level of reconstruction quality. A more easily understood amount rather than a rendering of MCDF and/or the amount being a function of local artifacts aids a radiologist in recognizing reconstruction quality and determining whether different reconstruction is warranted.