PROPELLER MRI Multi-Level Denoising for Low-SNR Artifact Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing PROPELLER MRI imaging techniques face challenges in denoising colored noise, leading to blurriness and degraded image quality, especially in low signal-to-noise ratio scenarios, with traditional denoising methods introducing artifacts like blurring and half-pixel shifts.
Innovation Solution
A deep learning-based multi-level denoising network is employed to denoise individual blades in the image domain before gridding, followed by parameter-less phase correction and adjoint non-uniform fast Fourier transform blocks, and a lightweight artifact removal model to generate an artifact-free gridded image, trained end-to-end with combined blade and grid level losses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If traditional denoising methods are applied to colored noise in PROPELLER images, then noise is reduced, but image quality degrades with blurriness and artifacts
Solution Approach 1:
The patent divides the denoising process into multiple levels: blade-level denoising followed by grid-level denoising. This segmentation allows each level to address specific noise characteristics without introducing excessive blurriness, as the multi-level approach processes different spatial frequencies separately
Solution Approach 2:
The patent applies blade-level denoising before the gridding operation. By performing denoising preliminarily on individual blades while they still contain Gaussian noise characteristics, the method prevents colored noise formation and avoids subsequent blurriness issues
2Object-affected harmful factors
If denoising is applied to low signal-to-noise ratio data, then noise is reduced, but denoising performance deteriorates drastically
Solution Approach 1:
The patent implements a feedback mechanism where the denoised blades are gridded to create a preliminary image, which then guides the grid-level denoising process. This feedback loop allows the system to adapt to low signal-to-noise ratio conditions by using the denoised information to improve subsequent denoising steps
Solution Approach 2:
The patent transitions from processing in k-space to image domain processing, and then back to k-space for final reconstruction. This dimensional transformation allows the application of advanced denoising algorithms that are more effective in the image domain while maintaining the benefits of frequency-domain processing
3Object-affected harmful factors
If individual blade denoising is performed, then noise is reduced in each blade, but denoising-induced artifacts appear in the final image
Solution Approach 1:
The patent merges blade-level denoising results with grid-level denoising in a unified multi-level framework. By combining these two denoising approaches, the system achieves comprehensive noise reduction while the grid-level processing corrects artifacts introduced at the blade level
Solution Approach 2:
The patent employs different denoising parameters and algorithms at different levels of the processing pipeline. By adapting parameters to the specific characteristics of each processing stage (blade-level vs. grid-level), the system optimizes noise reduction while minimizing artifact generation
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
The approach significantly improves image quality by reducing artifacts and enhancing sharpness while maintaining high fidelity, even in low signal-to-noise ratio conditions.
Implementation Method 1
During magnetic resonance imaging (MRI), when a substance such as human tissue is subjected to a uniform magnetic field (polarizing field B0), the individual magnetic moments of the spins in the tissue attempt to align with this polarizing field, but precess about it in random order at their characteristic Larmor frequency. If the substance, or tissue, is subjected to a magnetic field (excitation field B1) which is in the x-y plane and which is near the Larmor frequency, the net aligned moment, or 'longitudinal magnetization ', Mz, may be rotated, or 'tipped ', into the x-y plane to produce a net transverse magnetic moment, Mt.
Implementation Method 2
utilizing a deep learning-based multi-level denoising network to denoise each blade of the plurality of blades in an image domain
Data Source
AI summary
A system and method for improving image quality of periodically rotated overlapping parallel lines with enhanced reconstruction (PROPELLER) imaging include acquiring a plurality of blades of k-space data of a region of interest in a rotational manner around a center of k-space via a magnetic resonance imaging (MRI) scanner from a coil during a PROPELLER sequence, wherein each blade of the plurality of blades of k-space data includes a plurality of parallel phase encoding lines sampled in a phase encoding order. The system and method also include utilizing a deep learning-based multi-level denoising network to denoise each blade of the plurality of blades in an image domain to generate a plurality of denoised blades, to utilize a PROPELLER reconstruction algorithm to generate a denoised-gridded image from the plurality of denoised blades, and to remove individual-based denoising-induced artifacts from the denoised-gridded image to generate a denoised, artifact-free gridded image.