Variable Filter Length Local Tomography for Reduced Radiation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image reconstruction methods in medical applications, such as CT scans, require irradiating the entire patient cross-section, leading to high radiation doses and difficulty in differentiating tissue types and detecting contrast, especially in regions of interest like the cardiac area, while also being computationally intensive and artifact-prone.
Innovation Solution
The method employs variable filter length local tomography, using truncated cone-beam projection data and a novel filtering and backprojection algorithm to reconstruct images with improved spatial and contrast resolution, reducing radiation exposure and computational power, and allowing for real-time image reconstruction within a region of interest.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If conventional CT reconstruction is used to reconstruct the entire patient cross-section, then complete anatomical information is obtained, but radiation dose increases significantly
Solution Approach 1:
The patent divides the reconstruction space into a region of interest (ROI) and surrounding regions. Only x-rays intersecting the ROI are transmitted through the patient, blocking x-rays that would otherwise irradiate the entire cross-section. This segmentation allows selective reconstruction of only the necessary anatomical region, reducing radiation exposure while maintaining complete anatomical information within the ROI.
Solution Approach 2:
The patent implements local tomography by transmitting x-rays only through paths that intersect the ROI. This creates a locally optimized reconstruction where the ROI receives full imaging information while surrounding tissues receive minimal or no radiation. The local quality principle ensures that the imaging process is tailored specifically to the anatomical region requiring diagnostic evaluation.
2Object-affected harmful factors
If local tomography is used to reduce radiation dose by transmitting only x-rays intersecting the ROI, then radiation dose is reduced, but the ability to differentiate tissue types and detect contrast is lost
Solution Approach 1:
The patent employs adaptive iterative reconstruction that dynamically adjusts reconstruction parameters based on the specific imaging task and ROI characteristics. The algorithm iteratively refines the reconstruction to optimize both radiation dose reduction and tissue differentiation capability, adapting the reconstruction process rather than using a fixed approach. This dynamic adjustment maintains measurement precision while preserving dose reduction benefits.
Solution Approach 2:
The patent changes reconstruction parameters including filter length, iteration count, and regularization strength to optimize the balance between dose reduction and tissue differentiation. By adjusting these parameters adaptively, the system maintains the ability to differentiate tissue types and detect contrast agents even when using truncated x-ray data from local tomography, thus preserving measurement precision.
3Measurement precision
If conventional iterative reconstruction is used to improve image quality, then diagnostic accuracy improves, but computational power and time requirements increase
Solution Approach 1:
The patent applies partial iterative reconstruction by performing a limited number of iterations focused specifically on the ROI rather than reconstructing the entire patient cross-section. This partial action approach achieves sufficient image quality for diagnostic purposes while dramatically reducing computational requirements compared to full iterative reconstruction of the complete anatomy. The excessive action principle is applied by using more iterations than minimally required, optimized for the specific ROI to ensure diagnostic accuracy.
4Object-affected harmful factors
If local tomography transmits only x-rays through the ROI, then radiation dose is reduced, but artifacts are introduced in the reconstructed image
Solution Approach 1:
The patent applies preliminary filtering and preprocessing to the truncated x-ray data before reconstruction to minimize artifact formation. By preparing the data in advance with appropriate filtering operations, the system reduces the introduction of artifacts that would otherwise compromise image accuracy. This preliminary action ensures that the local tomography reconstruction maintains reliability despite the truncated data acquisition.
Solution Approach 2:
The patent introduces regularization terms and filtering operations as intermediary processes between the truncated x-ray data and the final reconstruction. These intermediaries smooth the reconstruction process and suppress artifact formation while preserving the dose-reduction benefits of local tomography. The intermediary operations mediate between the limited data acquisition and the need for reliable, artifact-free images.
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 enables differentiation of tissue types, reduces radiation dose by approximately 50%, and provides clear contrast visibility, while being computationally efficient and minimizing artifacts, thus enhancing diagnostic accuracy and reducing radiation exposure.
Implementation Method 1
transmitting only those X-rays through the patient that intersect the Region of Interest (ROI) inside the patient
Data Source
AI summary
Methods, processes and systems of image reconstruction using variable filter length local tomography, for reconstructing internal body images in medical applications, and the like. Algorithm uses less radiation, less computer power than prior art, and without using iteration algorithms so that all target sizes from large to small can be reconstructed.