GRAPPA Kernel Estimation Using Tile-All-Frame ACS Lines
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The signal-to-noise ratio (SNR) of GRAPPA reconstruction in dynamic MRI is paradoxically lowered by higher SNR autocalibration signal (ACS) lines due to high condition numbers in the GRAPPA kernel encoding equations, leading to corrupted kernel estimates and reduced image quality.
Innovation Solution
The tile-all-frame (TAF) method is introduced to acquire ACS lines, which provides a greater number of linear equations for estimating the k-space convolution kernel, improving the SNR of reconstructed images by using all k-space data without increasing the SNR excessively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If higher SNR ACS lines are used in GRAPPA, then the accuracy of kernel estimation is improved, but the condition number of the encoding equations becomes too high, causing the estimated kernel to be corrupted by random noise and reducing the SNR of reconstructed images
Solution Approach 1:
The patent changes the parameter of ACS line selection by using temporally interleaved k-spaces with low-pass filtered images instead of fully-sampled high-SNR ACS lines. This parameter change reduces the condition number of the encoding equations while maintaining sufficient kernel estimation accuracy, thereby avoiding noise amplification in the reconstructed images.
Solution Approach 2:
The patent uses partial k-space sampling in a temporally interleaved manner rather than full k-space sampling for ACS lines. By using only the necessary k-space data from multiple time points, the method achieves sufficient kernel estimation without creating an ill-conditioned system, thus avoiding the SNR penalty associated with high-SNR but ill-conditioned ACS lines.
2Ease of operation
If the average-all-frame method is used to acquire ACS lines in TGRAPPA, then the processing is simplified, but the SNR of reconstructed images is reduced due to high condition number
Solution Approach 1:
The patent introduces temporal interleaving where k-space data is acquired dynamically across multiple time points rather than statically in a single frame. This dynamic acquisition pattern allows the system to adaptively select ACS lines from different time points, reducing the condition number while maintaining operational simplicity through automated temporal sorting and filtering.
Solution Approach 2:
The patent uses periodic temporal interleaving where k-space lines are acquired in alternating time points, creating a periodic acquisition pattern. This periodic structure enables systematic organization of ACS line selection across multiple frames, simplifying the acquisition process while producing well-conditioned encoding equations that improve reconstructed image SNR.
Data Source
AI summary
A method for improving the signal-to-noise ratio (SNR) of TGRAPPA. The SNR of the ACS lines is proportional to the condition number of the GRAPPA kernel encoding equations. Therefore, the GRAPPA kernel estimated from higher SNR ACS lines amplifies the random noise in GRAPPA reconstruction. In TGRAPPA reconstruction of dynamic image series, a widely used method to acquire ACS lines is to average-all-frame (AAF). The present disclosure utilizes a tile-all-frame (TAF) as ACS lines to improve the SNR of the reconstructed images.