Pixon Smoothing for Tomographic Image Reconstruction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image reconstruction methods in tomographic imaging face challenges in effectively smoothing 3D image objects, particularly in medical imaging where noise and low data counts are prevalent, leading to suboptimal image quality.
Innovation Solution
The implementation of pixon smoothing, which involves reconstructing a 3D image object using a pixon map that assigns kernel functions to object points in 3D space, iteratively smoothing the object to achieve a quality within preset thresholds, and outputting a smoothed image object.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional image reconstruction methods are used, then the reconstruction process is simple and fast, but the image quality is poor due to noise and low data counts
Solution Approach 1:
The patent segments the image reconstruction process into two distinct stages: first, a preliminary reconstruction is performed using conventional methods; second, a pixon smoothing step is applied to the preliminary result. This segmentation allows each stage to be optimized independently, improving overall image quality without overwhelming computational complexity.
Solution Approach 2:
The patent performs a preliminary image reconstruction before applying pixon smoothing. This preliminary action provides an initial estimate that guides the subsequent smoothing process, allowing the algorithm to focus computational effort on refining specific regions rather than processing the entire reconstruction from scratch.
2Measurement precision
If smoothing is applied to reduce noise, then image quality improves, but lesion detection capability may deteriorate due to over-smoothing
Solution Approach 1:
The pixon smoothing algorithm applies different smoothing strengths to different regions of the image based on local characteristics. Regions with high signal variability (potential lesions) receive minimal smoothing, while homogeneous regions receive stronger smoothing. This local adaptation preserves lesion detectability while improving overall image quality.
Solution Approach 2:
The smoothing parameter in pixon smoothing is not fixed but dynamically adjusted based on the local image statistics and data quality. The algorithm adapts the smoothing strength in each region according to the uncertainty in the data, ensuring that reliable regions are smoothed more aggressively while uncertain regions retain their detail for accurate lesion detection.
3Measurement precision
If iterative pixon smoothing is performed to achieve optimal quality, then image quality improves, but computation time increases
Solution Approach 1:
The pixon smoothing algorithm incorporates feedback mechanisms that monitor image quality metrics during iteration. When the improvement in image quality falls below a threshold or when convergence criteria are met, the iteration process terminates automatically. This feedback control prevents unnecessary computations while ensuring optimal image quality is achieved.
Solution Approach 2:
The patent applies pixon smoothing selectively to regions where it provides the most benefit rather than uniformly processing the entire image. By focusing computational resources on areas with significant noise or uncertainty, the algorithm achieves near-optimal image quality with reduced computation time compared to full-image iterative processing.
Data Source
AI summary
Tomographically reconstructing a 3D image object corresponding to a data set includes reconstructing a first reconstructed object from the data set, receiving a smoothing map, smoothing the first reconstructed object based on the smoothing map thereby creating a first smoothed object, and outputting the first smoothed object as the 3D image object. Smoothing a first object thereby creating a smoothed object having a smoothed value associated to each object point in object space includes receiving the first object, determining, in a series of steps, single-kernel-smoothed objects, wherein each iteration step is associated with a kernel function and includes, determining a start object based on the first object, and smoothing the start object using the kernel function of the iteration step, thereby creating the single-kernel-smoothed object having single-kernel-smoothed values associated to each object point, and constructing the smoothed object from the single-kernel-smoothed values.