Minimizing echo and repetition times in magnetic resonance imaging using a double half‐echo k‐space acquisition and low‐rank reconstruction

Abstract

Sampling k‐space asymmetrically (ie, partial Fourier sampling) in the readout direction is a common way to reduce the echo time (TE) during magnetic resonance image acquisitions. This technique requires overlap around the center of k‐space to provide a calibration region for reconstruction, which limits the minimum fractional echo to ~60% before artifacts are observed. The present study describes a method for reconstructing images from exact half echoes using two separate acquisitions with reversed readout polarity, effectively providing a full line of k‐space without additional data around central k‐space. This approach can benefit sequences or applications that prioritize short TE, short inter‐echo spacing or short repetition time. An example of the latter is demonstrated to reduce banding artifacts in balanced steady‐state free precession.

Application value of T2 fluid-attenuated inversion recovery sequence based on deep learning in static lacunar infarction

“…In the present study, we found that there was no significant difference in the SNR of SLI lesions; however, the CNR of the images from the ACS acceleration method was significantly higher than that of the conventional T2-FLAIR images, which also indicates that the ACS technique can avoid technical defects, including the Gibbs artifact, SNR reduction, noise amplification, and residual blending caused by acceleration methods, such as HF, PI, and CS (24)(25)(26)(27)(28). Therefore, the accuracy of the diagnosis is also ensured.…”

Low‐rank reconstruction for simultaneous double half‐echo 23Na and undersampled 23Na multi‐quantum coherences MRI

“…Sampling an echo fraction below 60% of each k-space line has been shown to induce artifacts without the DHE low-rank framework. 22 Both k-space halves were jointly reconstructed using low-rank coupling and added sparsity. Two reconstructed k-space halves were obtained and combined via the sum of squares method.…”

“…This technique requires obtaining two k-space halves, which are retrospectively combined by enforcing low-rankness along the k-space rows. 22 In parallel, acceleration techniques that optimally leverage the highly multidimensional structure of 23 Na MQC MRI are warranted to reduce long acquisition times and the associated limitations in spatial resolution. Low-rank reconstruction has proven applicable to highly redundant 1 H multichannel MRI data.…”

Partial Fourier reconstruction of complex MR images using complex‐valued convolutional neural networks

“…Consequently, their accuracy of both image phase and magnitude estimations is suboptimal especially in presence of rapid local phase changes, exhibiting severe image artifacts and preventing the use of low PF fractions. 10,11 This limitation is particularly true for GRE based imaging and phase-sensitive imaging protocols. 4,5 Recently, deep learning using convolutional neural networks (CNNs) has emerged as a powerful computational tool in various aspects of MR image formation, including image reconstruction, artifacts reduction, and noise suppression.…”

“…Although these methods can often effectively reconstruct full‐resolution magnitude images, they cannot provide detailed phase information. Consequently, their accuracy of both image phase and magnitude estimations is suboptimal especially in presence of rapid local phase changes, exhibiting severe image artifacts and preventing the use of low PF fractions 10,11 . This limitation is particularly true for GRE based imaging and phase‐sensitive imaging protocols 4,5 …”