Multi-Source IGCT Rebinning for Artifact Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional CT scanners face challenges with field of view limitations and artifacts in multi-slice and helical cone-beam geometries, particularly due to incomplete data sets and non-uniform radiation distribution, which affect image quality and efficiency.
Innovation Solution
The implementation of a multi-source inverse geometry computed tomography (IGCT) system with a small detector and distributed x-ray sources, using rebinning techniques such as z-rebinning, trans-axial rebinning, and feathering to process projection data into a format suitable for conventional 3D cone-beam reconstruction, along with the application of TOM-windowing and Hilbert Transform methods for improved image reconstruction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If a large detector is used to increase field of view in conventional CT systems, then the field of view increases, but the device complexity and cost increase significantly
Solution Approach 1:
The invention divides the x-ray source into multiple discrete sources arranged in an array, where each source can be independently activated. This segmentation allows the system to achieve a large field of view by selectively activating only the necessary sources for each projection angle, rather than requiring a physically large detector to capture all angles simultaneously.
Solution Approach 2:
The invention transitions from a single-source, single-detector configuration to a multi-source array configuration. By adding the dimension of multiple sources arranged in a specific geometry (e.g., on a circle or arc), the system achieves enhanced field of view and sampling capability without proportionally increasing detector size or complexity.
2Productivity
If fan-beam or cone-beam geometry is used in multi-slice CT, then scan speed increases, but artifacts occur due to incomplete data and non-planar datasets
Solution Approach 1:
The projection data acquisition is segmented into multiple discrete angular views, with each view obtained by activating a specific subset of x-ray sources. This segmentation ensures that complete projection data is collected at each angle, avoiding the data incompleteness that causes artifacts in continuous fan-beam or cone-beam geometries.
Solution Approach 2:
The invention changes the geometric parameters of the x-ray beam arrangement by using multiple discrete sources at different angular positions rather than a continuous fan or cone beam. This parameter change allows the system to maintain complete data sampling while achieving multi-slice imaging capability, thereby eliminating artifacts related to incomplete data.
3Area of stationary object
If the detector size is increased to cover more field of view, then more areas are covered, but technical difficulty and cost increase
Solution Approach 1:
Instead of increasing detector size to cover a larger field of view, the invention inverts the approach by using multiple small, discrete x-ray sources arranged in an array. Each source projects beams at different angles, and by selectively activating sources, the system achieves wide coverage without requiring a large detector, thereby reducing technical difficulty and cost.
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
This approach increases the field of view, achieves uniform cone angles and radiation distribution, and reduces artifacts, resulting in improved image quality and efficiency without sacrificing scan coverage.
Implementation Method 1
The Hilbert transform is a mathematical operation used in signal processing and image reconstruction. In this patent, it is applied to projection data from multiple x-ray sources to reconstruct the scanned object.
Implementation Method 2
a first plurality of x-ray sources (103) arranged in a first trans-axial pattern... a second plurality of x-ray sources (103) arranged in a second trans-axial pattern
Data Source
AI summary
Disclosed are embodiments of methods for reconstructing x-ray projection data (e.g., one or more sinograms) acquired using a multi-source, inverse-geometry computed tomography (“IGCT”) scanner. One embodiment of a first method processes an IGCT sinogram by rebinning first in “z” and then in “xy,” with feathering applied during the “xy” rebinning. This produces an equivalent of a multi-axial 3rd generation sinogram, which may be further processed using a parallel derivative and/or Hilbert transform. A TOM-window (with feathering) technique and a combines backprojection technique may also be applied to produce a reconstructed volume. An embodiment of a second method processes an IGCT sinogram using a parallel derivative and/or redundancy weighting. The second method may also use signum weighting, TOM-windowing (with feathering), backprojection, and a Hilbert Inversion to produce another reconstructed volume.