Slope Constrained Cubic Interpolation for PET Image Artifacts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional interpolation methods for PET data, such as linear and cubic interpolation, degrade image resolution and introduce artifacts around high contrast objects, particularly in edge-preserving reconstructed PET data, leading to blurred edges and negative values that can confuse diagnosis.
Innovation Solution
The method employs slope constrained cubic interpolation, where the slope of the interpolation surface at each data point is limited by the data point's value, preventing negative values and preserving image resolution by calculating the interpolated data points based on surrounding reconstructed data points using a slope-limited bicubic spline.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If cubic interpolation is used to preserve image resolution, then image resolution is maintained, but artifacts (negative values) are introduced around high contrast objects
Solution Approach 1:
The patent modifies the cubic interpolation algorithm by introducing a slope constraint parameter that limits the maximum slope of the interpolation surface. By changing the parameter constraints on the interpolation function, the method maintains resolution while preventing negative values from forming around high contrast objects.
Solution Approach 2:
The patent implements a dynamic interpolation approach where the interpolation method adapts locally based on the characteristics of the data points. The slope constraint is adjusted according to the local gradient and data values, allowing the algorithm to maintain high resolution in most areas while preventing artifacts in specific regions with high contrast transitions.
2Device complexity
If linear interpolation or nearest point interpolation is used, then computational simplicity is maintained, but image resolution is degraded and edges are blurred
Solution Approach 1:
The patent enhances simple linear or nearest-neighbor interpolation by introducing slope constraint parameters that modify the interpolation behavior. This allows the method to maintain computational simplicity while improving edge preservation and preventing the blurring effect of traditional linear interpolation.
3Productivity
If traditional interpolation methods are applied to edge-preserving reconstructed PET data, then processing speed is maintained, but edge sharpness is lost and images are blurred
Solution Approach 1:
The patent implements a dynamic interpolation approach that adapts to local image characteristics. By evaluating local gradients and adjusting interpolation parameters accordingly, the method maintains processing speed while preserving edge sharpness in regions with high contrast transitions.
Solution Approach 2:
The patent modifies the interpolation parameters dynamically based on local image features. The slope constraint parameters are adjusted according to the local gradient magnitude, allowing fast processing while maintaining edge sharpness where needed.
Data Source
AI summary
Methods and systems are provided for interpolating acquired data of a tracer distribution. In one embodiment, a method comprises reconstructing the acquired data, and interpolating the reconstructed data with a surface, wherein a slope of the surface at a data point of the reconstructed data is limited by a value of the data point. In this way, interpolation artifacts around objects with high tracer density may be avoided.