Prior Image Constrained Compressed Sensing for Low-Dose CT
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current medical imaging techniques face challenges in maintaining high signal-to-noise ratio (SNR) while reducing radiation dose and scan time, often resulting in decreased image quality and increased noise.
Innovation Solution
The method involves reconstructing images using a prior image with high SNR, formed by averaging data along a slice direction, and applying weighting values to the acquired data to filter it orthogonally, followed by a PICCS reconstruction algorithm to preserve high spatial resolution and SNR, applicable to various imaging modalities like CT, MRI, and PET.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If radiation dose is reduced, then patient safety is improved, but signal-to-noise ratio decreases
Solution Approach 1:
The patent applies preliminary action by reconstructing a prior image from incomplete data before using it to constrain the final image reconstruction. This prior image, created from averaged data along slice directions, serves as a template that guides the iterative reconstruction process, allowing the system to achieve high signal-to-noise ratio images from reduced radiation dose data by leveraging the structural information already present in the prior image.
Solution Approach 2:
The patent uses an intermediary approach by introducing a prior image as a mediator between the incomplete measured data and the final reconstructed image. This prior image, derived from averaged projections, acts as a constraint that guides the iterative reconstruction algorithm, enabling the system to recover high-quality images from low-dose data by continuously comparing and refining against the prior image structure.
2Productivity
If scan time is reduced, then productivity is improved, but signal-to-noise ratio decreases
Solution Approach 1:
The patent applies preliminary action by pre-reconstructing a prior image from averaged data before the final reconstruction. This prior image is created from incomplete data that has been averaged along slice directions, providing a structural template that accelerates the final reconstruction process. By having this preliminary structure, the system can achieve high signal-to-noise ratio images faster than conventional methods would require full data acquisition.
Solution Approach 2:
The patent uses an intermediary approach by introducing a prior image as a mediator between the incomplete measured data and the final reconstructed image. This prior image, derived from averaged projections, acts as a constraint that guides the iterative reconstruction algorithm, enabling the system to recover high-quality images from reduced scan time data by continuously comparing and refining against the prior image structure.
3Measurement precision
If image resolution is increased, then measurement precision is improved, but radiation dose increases
Solution Approach 1:
The patent applies preliminary action by reconstructing a prior image from incomplete data before using it to constrain the final image reconstruction. This prior image, created from averaged data along slice directions, provides high spatial resolution structure that guides the iterative reconstruction process, allowing the system to achieve fine detail images from reduced radiation dose data by leveraging the structural information already present in the prior image.
Solution Approach 2:
The patent uses an intermediary approach by introducing a prior image as a mediator between the incomplete measured data and the final reconstructed image. This prior image, derived from averaged projections, acts as a constraint that guides the iterative reconstruction algorithm, enabling the system to recover high-resolution images from low-dose data by continuously comparing and refining against the prior image structure.
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 achievable SNR without significant decreases in spatial resolution, allowing for trade-offs in radiation dose and scan time, improving image quality and clinical value.
Implementation Method 1
a prior image is formed by averaging acquired image data along a slice direction
Implementation Method 2
Weighted image data is formed by applying the calculated plurality of weightings to the acquired image data such that the acquired image data is filtered along a direction orthogonal to an orientation of the plurality of image slices
Implementation Method 3
PICCS image reconstruction is then performed to reconstruct a plurality of target images of the subject
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method for reconstructing an image of a subject with a medical imaging system is provided. Image data is acquired with the medical imaging system, typically in an undersampled manner. A prior image constrained compressed sensing (PICCS] image reconstruction method is then implemented to reconstruct images of the subject being imaged. The prior image used in such a method is produced from so-called averaged image data. The averaged image data is produced by effectively averaging signal information associated with a plurality of different slice locations along a direction orthogonal to the plane parallel to the slice locations. Weightings are calculated from images reconstructed in a conventional manner from the acquired image data, and these weightings are employed to produce the averaged image data. By producing a prior image in this manner, a higher signal-to-noise ratio is achievable, allowing trade¬ offs with factors such as radiation dose in x-ray imaging.