2018
DOI: 10.1016/j.jmr.2018.03.006
|View full text |Cite
|
Sign up to set email alerts
|

Using a local low rank plus sparse reconstruction to accelerate dynamic hyperpolarized 13 C imaging using the bSSFP sequence

Abstract: Acceleration of dynamic 2D (T Mapping) and 3D hyperpolarized C MRI acquisitions using the balanced steady-state free precession sequence was achieved with a specialized reconstruction method, based on the combination of low rank plus sparse and local low rank reconstructions. Methods were validated using both retrospectively and prospectively undersampled in vivo data from normal rats and tumor-bearing mice. Four-fold acceleration of 1-2 mm isotropic 3D dynamic acquisitions with 2-5 s temporal resolution and t… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
13
0

Year Published

2018
2018
2023
2023

Publication Types

Select...
5
1

Relationship

1
5

Authors

Journals

citations
Cited by 10 publications
(13 citation statements)
references
References 44 publications
0
13
0
Order By: Relevance
“…An alternating center frequency scheme was used in both sets of studies, with each metabolite acquired every 1.8 s. Each 3D acquisition was undersampled along the time dimension by using a different undersampling pattern for each time‐point, which resulted in 70% undersampling for rat studies and 75% undersampling for mouse studies. A local low rank plus sparse (LLR+S) algorithm was used for reconstruction, as described previously . Briefly, the LLR+S algorithm enforces both low rank and sparse constraints via iterative soft thresholding on singular values and sparse coefficients to reconstruct undersampled dynamic MRI.…”
Section: Methodsmentioning
confidence: 99%
“…An alternating center frequency scheme was used in both sets of studies, with each metabolite acquired every 1.8 s. Each 3D acquisition was undersampled along the time dimension by using a different undersampling pattern for each time‐point, which resulted in 70% undersampling for rat studies and 75% undersampling for mouse studies. A local low rank plus sparse (LLR+S) algorithm was used for reconstruction, as described previously . Briefly, the LLR+S algorithm enforces both low rank and sparse constraints via iterative soft thresholding on singular values and sparse coefficients to reconstruct undersampled dynamic MRI.…”
Section: Methodsmentioning
confidence: 99%
“…300 6.4 | Spectral-spatial excitation Finally, another common strategy to accelerate metabolic imaging is to use prior knowledge about the resonance frequency of spectrally well separated signals for frequency-selective excitation. This is applied especially to hyperpolarized 13 C-MRI by combining spectral-spatial pulses with fast MRI readouts 286,[301][302][303][304][305][306][307][308][309] but has been shown also for 31 P-MRI. 310,311 Using a narrow-band spectral-spatial pulse to excite only a single metabolite resonance line makes the need for chemical shift encoding obsolete and enables conventional fast MRI readouts such as SSFP, echo-planar or spiral MRI that can be further accelerated via coherent/incoherent undersampling.…”
Section: Super-resolution Reconstructionmentioning
confidence: 99%
“…Following an initial application in denoising of MRSI data, 284 low‐rank filtering was integrated with B 0 inhomogeneity correction 283 and additional image prior knowledge on tissue type boundaries 282 for a more robust reconstruction of fully sampled Cartesian MRSI. Shortly afterwards, low‐rank approximation was combined with balanced‐SSFP or dynamic spiral hyperpolarized 13 C‐MRSI with incoherent k ‐space sampling 285,286 and 1 H‐FID‐MRSI accelerated by short T R and a combined SENSE and CS acceleration 58 . Low‐rank reconstruction was also used together with a spatiotemporal lipid prior that assumes orthogonality of spatiotemporal metabolite versus lipid signals and applied to lipid‐unsuppressed dual‐density spiral 1 H‐MRSI 168 …”
Section: Using Prior Knowledgementioning
confidence: 99%
“…Furthermore, most hyperpolarized signal is non-renewable, and is thus often not suitable for use in B 1 calibration [73], with the exceptions of 3 He and 129 Xe lung imaging [62,87]. Instead, a separate high concentration gas or liquid thermally polarized phantom is normally used to calibrate B 1 before most X-nuclei hyperpolarized measurements [106][107][108]. For coils with subject-dependent loading, such a phantom may be placed adjacent to the subject, near the region or structure of interest to ensure similar coil loading and local B 1 , and may be removed or left in place for the subsequent hyperpolarized measurement.…”
Section: B 1 Calibration For X-nucleimentioning
confidence: 99%
“…Further applications of bSSFP in hyperpolarized imaging include non-frequency selective hyperpolarized gas lung imaging in both 2D [284] and 3D [295,296], single [150] and multi-compound [297] hyperpolarized 13 C angiography, rapid imaging of the heart [298], and numerous metabolic imaging approaches [108,134,202,288].…”
Section: Balanced Steady-state Free Precessionmentioning
confidence: 99%