Tomosynthesis Direct Reconstruction Pre-Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current X-ray tomosynthesis reconstruction techniques face challenges in efficiently separating overlying tissue, enhancing contrast of small structures, and minimizing artifacts due to limited data acquisition, with direct reconstruction methods offering poor image quality and iterative methods being computationally expensive.
Innovation Solution
A method for generating three-dimensional image data sets through pre-processing and direct reconstruction of projection images using X-ray tomosynthesis systems, which includes pre-processing based on acquisition geometry and direct reconstruction techniques such as backprojection and iterative updates to improve image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If direct reconstruction techniques (shift-and-add algorithms or simple backprojection) are used, then computational efficiency is improved and reconstruction speed increases, but image quality deteriorates with low contrast and significant artifacts
Solution Approach 1:
The patent applies preliminary action by performing pre-processing of projection images before reconstruction, including normalization, filtering, and correction of geometric distortions. This preparation step enhances the quality of input data, enabling direct reconstruction methods to produce acceptable image quality while maintaining computational efficiency
Solution Approach 2:
The patent employs parameter changes by adjusting reconstruction parameters such as beam hardening correction factors, filter characteristics, and geometric calibration parameters. These parameter optimizations allow direct reconstruction algorithms to achieve better image quality without sacrificing speed
2Measurement precision
If iterative reconstruction techniques (ART) are used, then image quality is improved through iterative updates, but computational cost increases substantially requiring multiple iterations
Solution Approach 1:
The patent performs preliminary pre-processing of projection images including normalization, filtering, and geometric correction before reconstruction. This preparation reduces the computational burden during iterative updates by ensuring input data is optimized, allowing fewer iterations to achieve desired image quality
Solution Approach 2:
The patent segments the reconstruction process into distinct phases: pre-processing of projection images, direct reconstruction of a three-dimensional image data set, and optional iterative refinement. This segmentation allows the computationally intensive direct reconstruction to be performed first, with subsequent iterative updates focusing only on specific regions or parameters
3Loss of time
If limited angular range data is acquired in tomosynthesis, then acquisition time and radiation dose are reduced, but reconstruction accuracy deteriorates due to incomplete data
Solution Approach 1:
The patent applies parameter changes by using advanced reconstruction algorithms that can compensate for limited angular range through mathematical modeling and iterative optimization. These algorithms adjust reconstruction parameters to maximize accuracy despite incomplete data, maintaining quality while preserving the benefits of limited-angle acquisition
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The approach provides improved image quality with reduced computational effort and artifact minimization, making it suitable for various imaging contexts including medical and industrial applications.
Implementation Method 1
an X-ray source configured to emit X-rays through a volume of interest from different locations relative to the imaged object
Implementation Method 2
The attenuated X-rays are then detected by a set of detector elements. Each detector element produces a signal based on the intensity of the attenuated X-rays
Data Source
AI summary
A technique for directly reconstructing three-dimensional images acquired in tomosynthesis imaging is provided. The method allows for pre-processing a set of projection images based upon the acquisition geometry prior to direct reconstruction of the projection images. Furthermore, the technique allows for post-processing the reconstructed image data as desired to improve image quality. If desired the direct reconstruction process may be iterated a set number of times or until a desired threshold criteria is met, such as regarding image quality.