Deep Learning MRI Reconstruction with Extended FOV Coil Sensitivity Calibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Deep learning-based MRI reconstruction methods face challenges with limited field of view (FOV) scans, leading to tissue wrapping effects and degraded image quality due to problematic sensitivity estimation, particularly when iterative reconstruction is used, and these methods can also increase computation time.
Innovation Solution
The approach involves external coil sensitivity calibration using an extended FOV calibration scan to provide sensitivity maps and interleaved k-space data to a neural network, which reconstructs images with an extended FOV, minimizing aliasing artifacts while maintaining image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If self-calibration strategy (C3 calibration) is used to estimate sensitivity maps, then the reconstruction process can be simplified, but tissue wrapping effects are amplified leading to degraded image quality
Solution Approach 1:
The patent applies preliminary action by performing sensitivity calibration with an extended FOV before the actual limited FOV scan. The sensitivity maps are pre-computed using a calibration scan that covers a larger field of view, then these pre-computed sensitivity maps are reused for the accelerated limited FOV reconstruction, avoiding the need to perform self-calibration during the actual scan and preventing aliasing artifacts.
2Loss of time
If iterative deep learning reconstruction is used to accelerate MR scans, then scan time is reduced, but tissue wrapping effects are amplified resulting in degraded final reconstructed images
Solution Approach 1:
The patent performs sensitivity calibration in advance with an extended FOV calibration scan, storing the sensitivity maps for later use. During the actual accelerated scan with limited FOV, these pre-computed sensitivity maps are used to guide the iterative deep learning reconstruction, preventing the amplification of tissue wrapping effects while maintaining the speed benefits of accelerated scanning.
3Productivity
If limited FOV is used during acquisition, then scan coverage is focused on region of interest, but sensitivity estimation becomes problematic leading to degraded images
Solution Approach 1:
The patent performs sensitivity calibration in advance with an extended FOV calibration scan, storing the sensitivity maps for later use. During the actual accelerated scan with limited FOV, these pre-computed sensitivity maps are used to guide the iterative deep learning reconstruction, preventing the amplification of tissue wrapping effects while maintaining the speed benefits of accelerated scanning.
Solution Approach 2:
The patent extends the calibration FOV beyond the imaging FOV in the phase encoding direction, using a larger field of view for sensitivity calibration than what is needed for the actual image acquisition. This dimensional extension allows sensitivity maps to be computed without aliasing artifacts from the limited FOV scan.
4Measurement precision
If extended FOV calibration scan is performed, then sensitivity maps can be accurately estimated, but additional scan time is required
Solution Approach 1:
The patent performs sensitivity calibration in advance with an extended FOV calibration scan, storing the sensitivity maps for later use. During the actual accelerated scan with limited FOV, these pre-computed sensitivity maps are used to guide the iterative deep learning reconstruction, preventing the amplification of tissue wrapping effects while maintaining the speed benefits of accelerated scanning.
Solution Approach 2:
The extended FOV calibration scan serves dual purposes: it provides accurate sensitivity maps for the accelerated reconstruction and can itself be reconstructed using the same deep learning framework, making the calibration process more efficient and reducing the net additional time required.
Data Source
AI summary
Image reconstruction systems and methods include providing sensitivity maps for coils of a magnetic resonance imaging (MRI) system to a neural network. The systems and methods also include providing interleaved k-space data to the neural network, wherein the interleaved k-space data includes partial k-space data interleaved with zeros, or synthesized k-space data, to provide an extended field of view (FOV) different from a FOV utilized during acquisition of the partial k-space data, wherein the partial k-space data were obtained during a scan of a region of interest with the MRI system. The systems and methods further include outputting, from the neural network, a final reconstructed MR image based at least on the sensitivity maps and the interleaved k-space data, wherein the final reconstructed MR image includes the FOV utilized during the acquisition of the partial k-space data.