High spatial resolution spiral first‐pass myocardial perfusion imaging with whole‐heart coverage at 3 T

Wang, Junyu; Yang, Ya; Weller, Daniel S.; Zhou, Ruixi; Houten, Matthew Van; Sun, Changyu; Epstein, Frederick H.; Meyer, Craig H.; Kramer, Christopher M.; Salerno, Michael

doi:10.1002/mrm.28701

Cited by 11 publications

(37 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The interleaved SS and SMS images can be reconstructed using a motion‐compensated technique, namely (SMS‐slice‐)L1‐SPIRiT, 6,8,9 where the through‐plane operator was incorporated in the proposed SMS‐slice‐L1‐SPRIiT reconstruction for spiral perfusion imaging to reduce the slice leakage and improve the reconstruction performance, which can be expressed as follows:

\underset{x}{argmin} {‖ΦR F_{u} x - y‖}_{2}^{2} goodbreak+ λ_{1} {‖(G_{italicSPIRiT} - I) x‖}_{2}^{2} goodbreak+ λ_{2} {‖italicΨx‖}_{1} goodbreak+ λ_{3} {‖(G_{italicTP} - I) x‖}_{2}^{2},

where

x

are motion‐corrected multichannel dynamic perfusion images to be reconstructed for each slice,

y

are the acquired spiral SS or SMS data that might contain motion,

F_{u}

is an inverse Fourier gridding operator (NUFFT) that transforms from Cartesian image space to spiral k‐space, 31

R

is the motion‐correction operator mapping the motion‐corrected perfusion image series to the motion‐corrupted image series, 32

G_{SPIRiT}

is an image‐space SPIRiT operator for each slice, 33

G_{TP}

is an image‐space through‐plane operator that is calibrated by enforcing consistency on the desired slice and blocking the signal from the interfering slices to reduce slice leakage artifacts, 7,34,35

Ψ

is the finite time difference transform that operates on each individual coil separately enforcing sparsity in the temporal domain of perfusion image series,

I

is the identity matrix,

λ_{1}

λ_{2}

and

λ_{3}

are parameters that balance the data acquisition consistency with SPIRiT calibration consistency, temporal sparsity and slice consistency, and

Φ

is the S...…”

Section: Methodsmentioning

confidence: 99%

“…For SS acquisition, the same equation can be utilized by setting

Φ

I

and setting

λ_{3} = 0

, which results in the SS L1‐SPIRiT reconstruction.

λ_{1} = 1

and

λ_{2} = 0.4 M_{0}

, where

M_{0}

was the maximal magnitude value of the NUFFT images, were chosen for SS reconstructions, as in our SMS‐slice‐L1‐SPIRiT paper 7 . Similarly, for SMS reconstructions,

λ_{1} = 1

λ_{2} = 0.5 M_{0}

, and

λ_{3} = 1

were selected, as in our prior publication 7 .…”

Section: Methodsmentioning

confidence: 99%

“…

λ_{1} = 1

and

λ_{2} = 0.4 M_{0}

, where

M_{0}

was the maximal magnitude value of the NUFFT images, were chosen for SS reconstructions, as in our SMS‐slice‐L1‐SPIRiT paper 7 . Similarly, for SMS reconstructions,

λ_{1} = 1

λ_{2} = 0.5 M_{0}

, and

λ_{3} = 1

were selected, as in our prior publication 7 . These parameters were selected based on the results of a retrospective experiment that balanced image quality and temporal fidelity, as first described by Feng et al 36 The same regularization parameters were used for all reconstructions.…”

Section: Methodsmentioning

confidence: 99%

“…Recently, we developed a rapid interleaved single‐slice (SS) spiral perfusion pulse sequence and a corresponding motion‐compensated L1‐SPIRiT image reconstruction technique 5–7 . We have also developed spiral simultaneous multislice (SMS) techniques, including motion‐compensated SMS‐L1‐SPIRiT reconstruction and SMS‐slice‐L1‐SPIRiT reconstruction, which includes in‐plane and through‐plane kernels that can further reduce the through‐plane aliasing artifacts 7–10 . However, these advanced image reconstruction techniques rely on iterative algorithms such as nonlinear conjugate gradient to recover the images, which is time‐consuming.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

DEep learning‐based rapid Spiral Image REconstruction (DESIRE) for high‐resolution spiral first‐pass myocardial perfusion imaging

et al. 2021

Self Cite

View full text Add to dashboard Cite

The objective of the current study was to develop and evaluate a DEep learning‐based rapid Spiral Image REconstruction (DESIRE) technique for high‐resolution spiral first‐pass myocardial perfusion imaging with whole‐heart coverage, to provide fast and accurate image reconstruction for both single‐slice (SS) and simultaneous multislice (SMS) acquisitions. Three‐dimensional U‐Net–based image enhancement architectures were evaluated for high‐resolution spiral perfusion imaging at 3 T. The SS and SMS MB = 2 networks were trained on SS perfusion images from 156 slices from 20 subjects. Structural similarity index (SSIM), peak signal‐to‐noise ratio (PSNR), and normalized root mean square error (NRMSE) were assessed, and prospective images were blindly graded by two experienced cardiologists (5: excellent; 1: poor). Excellent performance was demonstrated for the proposed technique. For SS, SSIM, PSNR, and NRMSE were 0.977 [0.972, 0.982], 42.113 [40.174, 43.493] dB, and 0.102 [0.080, 0.125], respectively, for the best network. For SMS MB = 2 retrospective data, SSIM, PSNR, and NRMSE were 0.961 [0.950, 0.969], 40.834 [39.619, 42.004] dB, and 0.107 [0.086, 0.133], respectively, for the best network. The image quality scores were 4.5 [4.1, 4.8], 4.5 [4.3, 4.6], 3.5 [3.3, 4], and 3.5 [3.3, 3.8] for SS DESIRE, SS L1‐SPIRiT, MB = 2 DESIRE, and MB = 2 SMS‐slice‐L1‐SPIRiT, respectively, showing no statistically significant difference (p = 1 and p = 1 for SS and SMS, respectively) between L1‐SPIRiT and the proposed DESIRE technique. The network inference time was ~100 ms per dynamic perfusion series with DESIRE, while the reconstruction time of L1‐SPIRiT with GPU acceleration was ~ 30 min. It was concluded that DESIRE enabled fast and high‐quality image reconstruction for both SS and SMS MB = 2 whole‐heart high‐resolution spiral perfusion imaging.

show abstract

\underset{x}{argmin} {‖ΦR F_{u} x - y‖}_{2}^{2} goodbreak+ λ_{1} {‖(G_{italicSPIRiT} - I) x‖}_{2}^{2} goodbreak+ λ_{2} {‖italicΨx‖}_{1} goodbreak+ λ_{3} {‖(G_{italicTP} - I) x‖}_{2}^{2},

where

x

are motion‐corrected multichannel dynamic perfusion images to be reconstructed for each slice,

y

are the acquired spiral SS or SMS data that might contain motion,

F_{u}

is an inverse Fourier gridding operator (NUFFT) that transforms from Cartesian image space to spiral k‐space, 31

R

is the motion‐correction operator mapping the motion‐corrected perfusion image series to the motion‐corrupted image series, 32

G_{SPIRiT}

is an image‐space SPIRiT operator for each slice, 33

G_{TP}

Ψ

is the finite time difference transform that operates on each individual coil separately enforcing sparsity in the temporal domain of perfusion image series,

I

is the identity matrix,

λ_{1}

λ_{2}

and

λ_{3}

are parameters that balance the data acquisition consistency with SPIRiT calibration consistency, temporal sparsity and slice consistency, and

Φ

is the S...…”

Section: Methodsmentioning

confidence: 99%

“…For SS acquisition, the same equation can be utilized by setting

Φ

I

and setting

λ_{3} = 0

, which results in the SS L1‐SPIRiT reconstruction.

λ_{1} = 1

and

λ_{2} = 0.4 M_{0}

, where

M_{0}

was the maximal magnitude value of the NUFFT images, were chosen for SS reconstructions, as in our SMS‐slice‐L1‐SPIRiT paper 7 . Similarly, for SMS reconstructions,

λ_{1} = 1

λ_{2} = 0.5 M_{0}

, and

λ_{3} = 1

were selected, as in our prior publication 7 .…”

Section: Methodsmentioning

confidence: 99%

“…

λ_{1} = 1

and

λ_{2} = 0.4 M_{0}

, where

M_{0}

was the maximal magnitude value of the NUFFT images, were chosen for SS reconstructions, as in our SMS‐slice‐L1‐SPIRiT paper 7 . Similarly, for SMS reconstructions,

λ_{1} = 1

λ_{2} = 0.5 M_{0}

, and

λ_{3} = 1

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

DEep learning‐based rapid Spiral Image REconstruction (DESIRE) for high‐resolution spiral first‐pass myocardial perfusion imaging

et al. 2021

Self Cite

View full text Add to dashboard Cite

show abstract

“…Non-Cartesian sampling schemes may be preferred because they are less sensitive to motion (123,151,163,169,171,172,181,198,202,210,224,230,237,(242)(243)(244)(245)(246)(247)(248)(249)(250)(251)(252)(253), due to a densely sampled low-frequency range and the repeated sampling of the k-space center enables the extraction of motion signals (selfnavigation). Unfortunately, non-Cartesian sampling requires resampling of the acquired data onto a Cartesian grid, which is computationally expensive.…”

Section: Cmr Trajectories: It Is That Samplementioning

confidence: 99%

Cardiac MR: From Theory to Practice

Ismail

Strugnell

Coletti

et al. 2022

Front. Cardiovasc. Med.

View full text Add to dashboard Cite

Cardiovascular disease (CVD) is the leading single cause of morbidity and mortality, causing over 17. 9 million deaths worldwide per year with associated costs of over $800 billion. Improving prevention, diagnosis, and treatment of CVD is therefore a global priority. Cardiovascular magnetic resonance (CMR) has emerged as a clinically important technique for the assessment of cardiovascular anatomy, function, perfusion, and viability. However, diversity and complexity of imaging, reconstruction and analysis methods pose some limitations to the widespread use of CMR. Especially in view of recent developments in the field of machine learning that provide novel solutions to address existing problems, it is necessary to bridge the gap between the clinical and scientific communities. This review covers five essential aspects of CMR to provide a comprehensive overview ranging from CVDs to CMR pulse sequence design, acquisition protocols, motion handling, image reconstruction and quantitative analysis of the obtained data. (1) The basic MR physics of CMR is introduced. Basic pulse sequence building blocks that are commonly used in CMR imaging are presented. Sequences containing these building blocks are formed for parametric mapping and functional imaging techniques. Commonly perceived artifacts and potential countermeasures are discussed for these methods. (2) CMR methods for identifying CVDs are illustrated. Basic anatomy and functional processes are described to understand the cardiac pathologies and how they can be captured by CMR imaging. (3) The planning and conduct of a complete CMR exam which is targeted for the respective pathology is shown. Building blocks are illustrated to create an efficient and patient-centered workflow. Further strategies to cope with challenging patients are discussed. (4) Imaging acceleration and reconstruction techniques are presented that enable acquisition of spatial, temporal, and parametric dynamics of the cardiac cycle. The handling of respiratory and cardiac motion strategies as well as their integration into the reconstruction processes is showcased. (5) Recent advances on deep learning-based reconstructions for this purpose are summarized. Furthermore, an overview of novel deep learning image segmentation and analysis methods is provided with a focus on automatic, fast and reliable extraction of biomarkers and parameters of clinical relevance.

show abstract

Simultaneous multislice steady‐state free precession myocardial perfusion with full left ventricular coverage and high resolution at 1.5 T

McElroy

Ferrazzi

Nazir

et al. 2022

Magnetic Resonance in Med

View full text Add to dashboard Cite

To implement and evaluate a simultaneous multi-slice balanced SSFP (SMS-bSSFP) perfusion sequence and compressed sensing reconstruction for cardiac MR perfusion imaging with full left ventricular (LV) coverage (nine slices/heartbeat) and high spatial resolution (1.4 × 1.4 mm 2 ) at 1.5T. Methods:A preliminary study was performed to evaluate the performance of blipped controlled aliasing in parallel imaging (CAIPI) and RF-CAIPI with gradient-controlled local Larmor adjustment (GC-LOLA) in the presence of fat. A nine-slice SMS-bSSFP sequence using RF-CAIPI with GC-LOLA with high spatial resolution (1.4 × 1.4 mm 2 ) and a conventional three-slice sequence with conventional spatial resolution (1.9 × 1.9 mm 2 ) were then acquired in 10 patients under rest conditions. Qualitative assessment was performed to assess image quality and perceived signal-to-noise ratio (SNR) on a 4-point scale (0: poor image quality/low SNR; 3: excellent image quality/high SNR), and the number of myocardial segments with diagnostic image quality was recorded. Quantitative measurements of myocardial sharpness and upslope index were performed.Results: Fat signal leakage was significantly higher for blipped CAIPI than for RF-CAIPI with GC-LOLA (7.9% vs. 1.2%, p = 0.010). All 10 SMS-bSSFP perfusion datasets resulted in 16/16 diagnostic myocardial segments. There were no significant differences between the SMS and conventional acquisitions in terms of image quality (2.6 ± 0.6 vs. 2.7 ± 0.2, p = 0.8) or perceived SNR (2.8 ± 0.3 vs. 2.7 ± 0.3, p = 0.3). Inter-reader variability was good for both image quality (ICC = 0.84) and perceived SNR (ICC = 0.70). Myocardial sharpness was improved using the SMS sequence compared to the conventional Sarah McElroy and Giulio Ferrazzi contributed equally to this work.

show abstract

High spatial resolution spiral first‐pass myocardial perfusion imaging with whole‐heart coverage at 3 T

Cited by 11 publications

References 49 publications

DEep learning‐based rapid Spiral Image REconstruction (DESIRE) for high‐resolution spiral first‐pass myocardial perfusion imaging

DEep learning‐based rapid Spiral Image REconstruction (DESIRE) for high‐resolution spiral first‐pass myocardial perfusion imaging

Cardiac MR: From Theory to Practice

Simultaneous multislice steady‐state free precession myocardial perfusion with full left ventricular coverage and high resolution at 1.5 T

Contact Info

Product

Resources

About