Interventional MRI - Devices

1354694 - bSSFP cine MR imaging in the presence of cardiac devices: value of interleaved radial linear combination acquisition with partial dephasing

Friday, January 27, 2023

12:20 PM – 1:20 PM

Location: Pacific Jewel

Presenting Author(s)

Dana C. Peters, PhD, FSCMR

Professor of Radiology and Biomedical Imaging
Yale University, Connecticut, United States

Primary Author(s)

JX

Jie Xiang, BA

PhD student
Yale University, Connecticut, United States

Co-Author(s)

JL

Jérôme Lamy, PhD

Postdoctoral associate
Yale University
New haven, Connecticut, United States
RL

Rachel Lampert, MD

Professor
Yale School of Medicine, Connecticut, United States
Dana C. Peters, PhD, FSCMR

Professor of Radiology and Biomedical Imaging
Yale University, Connecticut, United States

Background: The high SNR and excellent blood-myocardium contrast make balanced steady state free precession (bSSFP) (1) a workhorse in cardiac MRI, but bSSFP suffers from interpretation-limiting artifacts in patients with cardiac implants. Linear combination bSSFP (lcSSFP) (2, 3) is an established way to reduce banding artifacts due to the off-resonance induced by the devices, but the immediate problem is that the increased scan time (usually four-fold) makes it impractical for cardiac cine imaging. Furthermore, the near band hyperintensity due to flow in bSSFP can also impact image quality. We recently developed an interleaved radial lcSSFP (RLC-SSFP), combined with partial dephasing (PD) (4) (using unbalanced slice-select pre-phaser) for improved cardiac cine imaging when cardiovascular electronic devices are present.

Methods: Based on simulations and phantom experiments with and without flow, an optimized protocol, including PD extent, was identified. Ten healthy subjects were then imaged on a 3T MRI (Siemens), in this IRB approved study, with an ICD positioned extracorporeally on the left chest in the pre-pectoral area below the clavicle, mimicking clinical placement. The four-phase cycling with interleaved radials (Figure 1) was acquired during a single breath-hold with retrospective gating using the following scan parameters: TR/TE/θ = 3.4 ms/1.7 ms/35-40°, FOV = 380 mm, slice thickness = 8 mm, 60 radial projections per dynamic, 12 readouts per segment, 21 heartbeat breath-hold. The slice selection prephaser was unbalanced by a factor of 16.7% to mitigate flow-related artifacts. Pre-whitening, complex-summation of the four dynamics, gridding and coil selection were performed before the GPU-accelerated iterative reconstruction NLINV (5). The image quality of RLC-SSFP with PD was evaluated by two independent graders (1=worst, 5=best), compared with the standard Cartesian bSSFP. Additionally, Cartesian GRE and Cartesian lcSSFP were acquired for comparison.

Results: For the in-vivo experiments, RLC-SSFP showed improved cine image quality compared to bSSFP and maintained bSSFP contrast compared to GRE (Figure 2). Specifically, the proposed method significantly reduced banding artifacts (4.7±0.5 vs 2.4±0.7) vs. bSSFP; the interleaved radials produced fewer streak artifacts (3.7±0.6 vs 2.6±0.7) vs. no interleaving; application of PD reduced transient flow-induced hyperintense artifacts (4.4±0.4 vs 3.8±0.6) vs no PD, thus enabled better cavity visibility (3.7±0.46 vs 2.7±0.94) and overall image quality (3.4±0.64 vs 2.2±0.70) compared to the standard Cartesian bSSFP(Figure 3).

Conclusion: The proposed RLC-SSFP method with PD may improve cine image quality in subjects with cardiac devices while maintaining the acquisition time within a feasible single breath-hold of 17 seconds.