3277

Robust non-contrast pulmonary perfusion imaging using pseudo-continuous arterial spin labeling with background suppression

Joshua S. Greer^1,2, Yiming Wang², Limin Zhou², Tarique Hussain^1,2, and Ananth J. Madhuranthakam^2,3
¹Pediatrics, UT Southwestern Medical Center, Dallas, TX, United States, ²Radiology, UT Southwestern Medical Center, Dallas, TX, United States, ³Advanced Imaging Research Center, UT Southwestern Medical Center, Dallas, TX, United States

Synopsis

Pulmonary perfusion imaging using ASL has been demonstrated using the pCASL technique with varying degrees of reproducibility. In this study, a pCASL labeling strategy is proposed to improve the stability of perfusion signal in the lungs using cardiac triggering to reduce flow-induced signal variations, and background suppression to improve SNR. With the proposed technique, high SNR perfusion images are created with a limited number of signal averages. This optimized approach may allow full coverage of the lungs in clinically reasonable scan times.

Introduction

Non-contrast perfusion imaging using arterial spin labeled (ASL) MRI has been well established in the lungs using the pulsed-ASL FAIR technique^1,2. However, pseudo-continuous ASL (pCASL) has been shown to offer improved SNR, and is recommended for brain perfusion imaging³. The extension of pCASL to study pulmonary perfusion has been non-trivial due to the highly pulsatile blood flow and complex anatomy of the vasculature, posing a significant challenge in identifying the vessel of interest for successful labeling.
pCASL has been previously demonstrated for pulmonary perfusion imaging^4-6. However, inconsistent results were achieved due to the extended pCASL repetition time rendering the acquisition sensitive to cardiac phase⁵. Cardiac triggering the acquisition following the post-label delay can ensure consistent diastolic acquisitions, but the resulting variable post-label delay is not compatible with background suppression, which has been shown to significantly improve ASL perfusion images³.
Therefore, the purpose of this study was to implement a pCASL labeling technique to improve the stability of the ASL signal in the lungs using an effective cardiac triggering strategy that is compatible with background suppression and allows perfusion quantification.

Methods

This was a HIPAA-compliant study approved by the institutional review board. All subjects provided written informed consent prior to the participation in the study. Figure 1 shows the proposed pCASL sequence that was implemented on a 3T Ingenia scanner (Philips Healthcare, Best, The Netherlands). The saturation pulses at the start of the sequence were cardiac triggered⁴, and the label duration (LD) and post-label delay (PLD) were reduced compared to timings used in the brain and kidneys^3,7, such that the acquisition occurs during the diastolic period of the following heartbeat, and were set to 600-750ms depending on the subjects heartrate. This results in consistent diastolic acquisitions to reduce signal variations introduced by pulsatile flow^1,5. Two background suppression pulses were applied during the PLD to improve SNR and reduce physiological noise, and a single-shot turbo spin echo (SShTSE) acquisition was used to reduce susceptibility artifacts in the lungs. A guided breathing approach was used to reduce motion during the acquisition, and 4-8 signal averages were acquired in 0:48 - 1:30 minutes. 2D FAIR images were acquired in each subject for comparison, and a proton-density (M0) image was acquired for perfusion quantification^8.
The proposed labeling and cardiac triggering strategy was evaluated in 5 healthy volunteers. These images were compared with previously presented data from an additional 15 volunteers, who were scanned with various approaches applying pCASL in the lungs, which were progressively modified to improve the stability of the perfusion signal. These approaches include labeling of the IVC with the start of the sequence cardiac triggered, IVC labeling with a cardiac triggered acquisition for more consistent cardiac phase, and RPA labeling with a cardiac triggered acquisition to reduce the transit time of labeled blood. The coefficient of variation of the perfusion signal across dynamics was used to evaluate the signal stability of each of these approaches.

Results

The proposed scheme with cardiac triggering and background suppression provided robust pulmonary perfusion-weighted images with only 3 signal averages (Fig. 2). This approach also provided consistent perfusion-weighted images across different dynamics, that are comparable to FAIR (Fig. 3). Additionally, pCASL images demonstrated reduced vascular signal, since the inflowing perfusion bolus has a fixed duration, unlike FAIR⁹.
The average coefficient of variation across dynamics was 0.74 ± 0.18 (average ± SD) for the IVC labeling approach, 0.80 ± 0.22 for IVC labeling with the cardiac triggered acquisition, 0.62 ± 0.21 for RPA labeling with the cardiac triggered acquisition, and 0.34 ± 0.02 for RPA labeling with the shorter TR and background suppression, indicating that the proposed approach provides the most stable perfusion signal.
The average perfusion in the five subjects scanned with the proposed RPA labeling with background suppression technique was 435 ± 65 mL/100g/min, within the normal expected range².

Discussion

Previous attempts at generating pulmonary perfusion maps with pCASL have provided promising, yet inconsistent results due to variations in cardiac phase of the acquisition, limited SNR, and extended scan times. The proposed pCASL labeling and cardiac triggering strategy addressed these limitations and provided consistent perfusion signal in the lungs. The improved SNR with this technique allows fewer signal averages to be acquired and should allow full coverage of the lungs to be achieved in clinically reasonable scan times for complete assessment of pulmonary perfusion.

Acknowledgements

No acknowledgement found.

References

1 Greer, J. S. et al. Non-contrast quantitative pulmonary perfusion using flow alternating inversion recovery at 3 T: A preliminary study. Magnetic resonance imaging 46, 106-113 (2018).

2 Bolar, D. S. et al. Quantification of regional pulmonary blood flow using ASL-FAIRER. Magn Reson Med 55, 1308-1317, doi:10.1002/mrm.20891 (2006).

3 Alsop, D. C. et al. Recommended implementation of arterial spin‐labeled perfusion MRI for clinical applications: a consensus of the ISMRM perfusion study group and the European consortium for ASL in dementia. 73, 102-116 (2015).

4 Martirosian, P. et al. Measurement of Lung Perfusion Using Optimized Pseudo-Continuous Arterial Spin Labeling of Pulmonary Arteries and Fast True-FISP Imaging at 3 Tesla. International Society for Magnetic Resonance in Medicine (2017).

5 Joshua S. Greer, Xinzeng Wang, Ivan Pedrosa & Madhuranthakam, A. J. Pulmonary Perfusion using pseudo-continuous Arterial Spin Labeling. International Society for Magnetic Resonance in Medicine (2016).

6 Petros Martirosian et al. Measurement of Pulmonary Perfusion using PCASL True-FISP Imaging at 1.5 Tesla. International Society for Magnetic Resonance in Medicine (2018).

7 Robson, P. M. et al. Volumetric arterial spin-labeled perfusion imaging of the kidneys with a three-dimensional fast spin echo acquisition. Academic radiology 23, 144-154 (2016).

8 Buxton, R. B. et al. A general kinetic model for quantitative perfusion imaging with arterial spin labeling. 40, 383-396 (1998).

9 Wong, E. C., Buxton, R. B. & Frank, L. R. J. M. r. i. m. Quantitative imaging of perfusion using a single subtraction (QUIPSS and QUIPSS II). 39, 702-708 (1998).

Figures

Figure 1: Cardiac-triggered pCASL sequence with shorter repetition time to achieve acquisitions with consistent cardiac phase. The fixed post-label delay also allows for background suppression and perfusion quantification.

Figure 2: Sagittal perfusion-weighted images of the right lung in two different volunteers acquired with the proposed pCASL sequence and 3 averages, showing high signal to noise ratio with a limited number of signal averages.

Figure 3: Perfusion-weighted images shown with a single signal average in another volunteer using pCASL with proposed labeling scheme (top row) and with FAIR (bottom row), showing consistent signal between different dynamics. The quantified perfusion maps were calculated with 3 signal averages.

Figure 4: Mean, standard deviation, and coefficient of variation across dynamics of perfusion-weighted images. While each progressive modification to the sequence reduced the mean perfusion signal, the standard deviation across dynamic scans also decreased, indicating improved signal stability. The most stable signal across dynamics is achieved by the RPA labeling approach with background suppression, as indicated by the reduced coefficient of variation.

Proc. Intl. Soc. Mag. Reson. Med. 28 (2020)

3277