Tomographic Image Reconstruction Using Virtual Rectification Support
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Tomographic reconstruction methods, especially in cone-beam geometry, require high computational intensity due to complex calculations involved in transforming measurements from different angles into reconstructed data, and existing methods like re-sampling at each angle are costly in terms of processing power.
Innovation Solution
A method and apparatus for image reconstruction that involves creating rectifications of projections by transforming data onto a virtual rectification support oriented in accordance with reconstruction calculation points, allowing for reduced computational intensity by simplifying calculations into a series of translations, summations, and magnifications, applicable to fan-beam and cone-beam geometries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional tomographic reconstruction methods are used with cone-beam geometry, then accurate 3D image reconstruction is achieved, but computational intensity becomes excessively high
Solution Approach 1:
The patent segments the reconstruction process into distinct stages: creating rectified projections from acquired projections, then using these rectified projections for reconstruction. This segmentation allows computationally intensive operations to be performed once on rectified data rather than repeatedly on original oblique projections, reducing overall computational intensity while maintaining reconstruction accuracy
Solution Approach 2:
The patent performs preliminary rectification of projections before the main reconstruction process. By pre-aligning and reformatting projections onto a standardized rectification support, the system prepares data in advance so that subsequent reconstruction operations require less computational effort, effectively performing necessary transformations beforehand
2Manufacturing precision
If re-sampling is performed at each angle of projection to align reconstruction points, then proper alignment is achieved, but calculation cost increases significantly
Solution Approach 1:
The patent creates a universal rectification support that can serve all reconstruction angles and detector positions. Instead of performing angle-specific re-sampling operations, the system transforms all projections onto a common standardized support, making the reconstruction process more efficient while maintaining proper alignment for all viewing angles
3Loss of information
If 3D reconstruction is performed from cone-beam projections, then complete volumetric information is obtained, but computational complexity increases even more
Solution Approach 1:
The patent introduces rectified projections as an intermediary representation between the acquired cone-beam projections and the final 3D reconstruction. These rectified projections serve as a simplified intermediate form that retains all necessary volumetric information while being more suitable for efficient reconstruction algorithms, reducing the complexity of the overall process
Data Source
AI summary
A method and apparatus for imaging including reconstruction of an image of an object from a set of projections acquired for different positions of an acquisition apparatus around the object to be imaged and in which a projection source is situated at a finite distance relative to the acquisition apparatus. The method includes undertaking a set of rectifications of projections acquired for different positions of the acquisition apparatus, rectifications comprising transforming the data of the projection acquired projection into projection data on a virtual rectification support placed in a preset orientation in concordance with a set of reconstruction calculation points distributed over the image to be reconstructed.