Introduction

First-pass cardiac magnetic resonance (CMR) perfusion imaging is widely used for non-invasive assessment of coronary artery disease (CAD). However, conventional CMR perfusion sequences at 1.5T are limited in spatial coverage and resolution, limiting the assessment of subendocardial perfusion defects and transmural perfusion gradients and precluding assessment of total myocardial ischemic burden. We have previously demonstrated that simultaneous multi-slice (SMS) bSSFP with a multiband factor of 2 and in-plane acceleration of 5.5 can be combined with compressed sensing reconstruction to acquire a 6-slice CMR perfusion sequence at 1.5T with high spatial resolution¹. However, a 6-slice acquisition is not sufficient for full left-ventricular coverage in most patients. SMS-bSSFP has recently been extended to a multiband (MB) factor of 3², enabling a further increase in spatial coverage. The aim of the current study is to use SMS-bSSFP with MB3 to achieve high resolution 1.5T SMS-bSSFP perfusion imaging with whole heart coverage.

Methods

A prototype SMS-bSSFP sequence with CAIPIRINHA encoding³, GC-LOLA correction⁴ and pseudo-random undersampling was used (see Figure 1). The pseudorandom undersampling scheme for SMS-bSSFP MB3 is analogous to that proposed for MB21. To extend this algorithm to MB3, the fully sampled central region is maintained, and the under-sampled (peripheral) k-space lines are sorted into 3 bins which correspond to the 3 phase-cycling steps for each slice (slice 1: 0, 2π/3, 4π/3, slice 2: 0, 0, 0, slice 3: 0, 4π/3, 2π/3). An equal number of k-space lines are randomly selected from each bin and the selected lines are acquired sequentially from each bin to maintain the required SMS-bSSFP phase cycling and CAIPIRINHA phase ramp in k-space. Two first-pass rest perfusion acquisitions (proposed and conventional), separated by 10 minutes and with a contrast dose of 0.075 mmol/kg each, were acquired in a random order in 6 patients (2M, mean age 53±10) undergoing routine clinical CMR studies at 1.5T (MAGNETOM Aera, Siemens Healthcare, Erlangen, Germany). The patients were instructed to hold their breath during the first-pass of the contrast agent. To enable whole-heart coverage and high in-plane spatial resolution, the total acceleration of proposed SMS-bSSFP sequence is 16 (in-plane acceleration: 5.5, MB factor 3). The in-plane acceleration of the conventional 3-slice (single-band) sequence is 3. Other imaging parameters are matched as closely as possible: TR/TE/α: 2.97ms/1.26ms/40° (proposed), TR/TE/α: 2.73ms/1.37ms/40° (conventional), voxel size: 1.4x1.4x10mm³ (proposed), 1.9x1.9x10mm³ (conventional), typical FOV: 360x360mm², TS 99ms (proposed) 94ms (conventional), total acquisition duration: 558ms (proposed), 567ms (conventional), bandwidth 1015Hz/Px. The conventional sequence was reconstructed using a standard GRAPPA reconstruction, while the proposed sequence was reconstructed using an inline SMS-compatible⁵ prototype compressed sensing reconstruction⁶. The sharpness was measured as previously described^1,7 at the septal blood-myocardium interface on the dynamic frame corresponding to peak blood pool signal intensity. The measurement was performed on all slices acquired with the conventional sequence, and on the 3 SMS slices with matched cardiac phase and position to the conventional acquisition.

Results

Whole heart coverage was achieved in all patients using the proposed SMS-bSSFP sequence. Figures 2-3 show examples of first-pass perfusion acquisitions in two patients with no suspected CAD, using the proposed SMS-bSSFP perfusion sequence with 9 slices and high in-plane spatial resolution (1.4x1.4mm²) and a conventional 3-slice sequence with in-plane resolution = 1.9 x 1.9 mm². All images are shown at peak myocardial signal enhancement. Sharpness measurements across 3 slices and 6 patients were higher using the proposed 9-slice SMS sequence vs. a conventional 3-slice sequence (0.31 ± 0.06 mm^-1 vs. 0.38 ± 0.09 mm^-1, p = 0.002, Figure 5).

Discussion

This study demonstrates that SMS-bSSFP with a MB factor of 3 can be used to acquire whole-heart CMR perfusion imaging at 1.5 T with high in-plane spatial resolution. The temporal regularisation employed in the iterative reconstruction results in blurring and artefacts outside of the first-pass (breath-hold) section of the acquisition. Development of a motion-compensated acquisition-reconstruction framework for SMS-bSSFP is warranted to ensure the image quality is not degraded in the case of a sub-optimal breath-hold or free-breathing acquisition. The current study was carried out at rest in patients without suspected CAD, therefore further validation of the sequence in patients with CAD and in stress conditions is required to evaluate the diagnostic value of this sequence.

Conclusion

SMS-bSSFP with a total acceleration factor of 16 and compressed sensing reconstruction enables whole-heart CMR perfusion imaging at 1.5 T with high in-plane spatial resolution (1.4 x 1.4 mm²).

Acknowledgements

This work was supported by the Engineering and Physical Sciences Research Council (EPSRC) grant (EP/R010935/1), the British Heart Foundation (BHF) (PG/19/11/34243), the Wellcome EPSRC Centre for Medical Engineering at King’s College London (WT 203148/Z/16/Z), the National Institute for Health Research (NIHR) Biomedical Research Centre based at Guy’s and St Thomas’ National Health Service (NHS) Foundation Trust and King’s College London, and Siemens Healthineers. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health.