Axially Varying Truncation for PET/MR Attenuation Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
PET/MR systems face challenges with MR-based attenuation correction due to image distortion at the edges of the magnetic resonance (MR) field-of-view, leading to truncated images and potential inclusion of distorted data, which is not adequately addressed by existing methods.
Innovation Solution
A method and system that perform truncation completion and attenuation correction by defining multiple transverse slices offset from the station centerline, using a center and annulus with varying inner and outer radii for PET and MR data integration, and adjusting the effective inner radius based on the axial distance to mitigate distortion and enhance data accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If MR-based attenuation correction is used in PET/MR systems, then anatomical imaging accuracy is improved, but image distortion occurs at the edges of the MR field-of-view
Solution Approach 1:
The patent divides the image processing into three distinct regions: a center region using only MR data, an annular region using both MR and PET data, and an outer region using only PET data. This segmentation allows each region to be processed with the most appropriate data source, resolving the contradiction between using MR for accurate anatomy and avoiding distorted edge regions.
Solution Approach 2:
The patent applies different data integration strategies to different spatial locations: central regions use high-quality MR anatomical data, while peripheral regions use PET data where MR distortion occurs. This local differentiation resolves the contradiction by optimizing the quality metric for each specific location rather than applying a uniform approach.
2Area of stationary object
If the MR field-of-view is extended to cover the entire PET field-of-view, then complete anatomical coverage is achieved, but image distortion increases at the edges
Solution Approach 1:
The patent segments the radial distance from the center into three zones defined by inner and outer radii. The annular region between these radii is where MR and PET data are combined, allowing the effective FOV to extend to PET boundaries while maintaining quality through selective data usage in the annular transition zone.
Solution Approach 2:
Instead of extending MR FOV and accepting distortion, the patent inverts the approach by using PET data to complete the attenuation correction in regions where MR data is distorted, thereby achieving full coverage without the harmful distortion effects.
3Productivity
If truncated MR images are used for attenuation correction, then processing speed is maintained, but accuracy is reduced due to missing peripheral data
Solution Approach 1:
The patent performs preliminary segmentation of the image space into center and annular regions before attenuation correction. By pre-defining where MR and PET data will be combined, the system prepares the framework for accurate correction without requiring computationally intensive post-processing, thus maintaining processing speed while improving accuracy.
Solution Approach 2:
The patent introduces an intermediate annular region that serves as a transition zone between the pure MR center region and the pure PET outer region. This intermediary zone allows smooth integration of both data types, improving accuracy by utilizing complementary information from both modalities without requiring full reprocessing.
4Measurement precision
If PET data is used to complete truncated MR images, then attenuation correction accuracy is improved, but data integration complexity increases
Solution Approach 1:
The patent simplifies the integration complexity by segmenting the problem into three clear radial zones with predetermined data source assignments. This segmentation provides a straightforward algorithmic approach that reduces the complexity of deciding how and where to integrate PET and MR data, making the process more manageable despite the multi-modal integration.
Data Source
AI summary
An apparatus and method for a PET/MR system having a PET scanner and an MR scanner. A patient may be advanced through the system in sequential stations, with multiple transverse slices defined within at least one of the stations in which each slice is offset a distance Δz from the station centerline. The method may comprise the steps of: defining a center and an annulus thereabout for each slice, wherein the annulus has an inner radius Reff and an outer radius Rout; conducting attenuation correction of the PET images using only MR data from the region within the inner radius, only PET data from the region outside the outer radius, and a combination of PET and MR data from the region of the annulus; and varying Reff as a function of Δz for selected slices within the station.