PET Super-Resolution via Cone-Beam LOR Reconfiguration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current PET image reconfiguration algorithms have low resolution, require extensive calculations and time, and are inefficient in measuring the point spread function (PSF) at all voxel locations, leading to image degradation and increased complexity.
Innovation Solution
An image-based super-resolution method using cone-beam-based line-of-response (LOR) reconfiguration, which includes blur modeling, LOR acquisition, cone-beam sinogram conversion, and high-resolution image induction, along with non-rigid registration for interpolating PSF across voxel locations, enabling parallel processing and reduced computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional PET image reconfiguration algorithms are used, then the imaging process is simple, but the resolution is low and processing time is long
Solution Approach 1:
The patent segments the reconfiguration process into multiple stages: acquiring LOR data, converting to cone-beam sinogram, performing iterative reconfiguration, and generating super-resolution images. This segmentation allows complex operations to be broken down into manageable steps, improving resolution while controlling overall complexity through structured processing phases.
Solution Approach 2:
The patent transforms the traditional 2D sinogram representation into a 3D cone-beam sinogram format, adding a spatial dimension that enables better modeling of the point spread function at each voxel location. This dimensional transformation facilitates super-resolution by capturing spatial variations in blur characteristics throughout the imaging volume.
2Measurement precision
If multiple reconfiguration processes are performed to achieve super-resolution, then image quality improves, but calculation time and processing complexity increase significantly
Solution Approach 1:
The patent performs preliminary actions by pre-calculating and storing the cone-beam sinogram representation and point spread function characteristics before the main iterative reconfiguration process. This preparation reduces the computational burden during subsequent processing steps, achieving super-resolution while minimizing total processing time through advance computation of reusable data structures.
Solution Approach 2:
The patent creates a virtual copy of the imaging process by simulating multiple reconfiguration scenarios through iterative algorithms that process cone-beam sinograms. This virtual copying allows the system to explore multiple resolution levels and reconstruction paths without requiring multiple physical scans, reducing actual processing time while maintaining super-resolution quality.
3Device complexity
If preprocessing such as single slice rebinning and arc correction is applied, then the reconfiguration algorithm is simplified, but blur occurs in sinogram and image quality degrades
Solution Approach 1:
Instead of applying conventional preprocessing steps that simplify the algorithm at the cost of introducing blur, the patent inverts the approach by using the raw LOR data and cone-beam sinogram representation directly in the iterative reconfiguration process. This inversion allows the system to maintain image quality by avoiding premature simplification while still achieving algorithmic manageability through the structured multi-stage processing approach.
4Measurement precision
If PSF is measured at all voxel locations using Monte-Carlo simulation, then accurate blur modeling is achieved, but calculation time and memory requirements increase significantly
Solution Approach 1:
The patent applies local quality by measuring and modeling the point spread function specifically at each voxel location where it is most needed, rather than uniformly across the entire imaging volume. The cone-beam sinogram representation enables efficient local PSF characterization by capturing spatial variations in blur characteristics only at relevant positions, reducing overall computation time while maintaining accurate blur modeling where required.
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 enhances PET image resolution, reduces processing time and complexity, and provides accurate super-resolution images by modeling blur at each location and interpolating PSF efficiently across all voxel locations.
Implementation Method 1
A positron emitted from a radioactive isotope consumes all of the self-kinetic energy during a very short period of time after its emission and is coupled with a neighboring electron, thereby becoming extinct. In thisinstance, two annihilation radiations, for example, gamma rays are emitted at an angle of 180 degrees.
Data Source
AI summary
An image-based super-resolution method using a cone-beam-based line-of-response (LOR) reconfiguration in a positron emission tomography (PET) image is provided. That is, an apparatus and method for reconfiguring a super-resolution PET image using a cone-beam-based LOR reconfiguration is provided.