Highly accelerated free-breathing real-time 2D flow imaging using compressed sensing and shared velocity encoding

• Published in: European radiology 2024-03, Vol.34 (3), p.1692-1703
• Main Authors: Xiong, Fei, Emrich, Tilman, Schoepf, U. Joseph, Jin, Ning, Hall, SarahRose, Ruddy, Jean Marie, Giese, Daniel, Lautenschlager, Carla, Emrich, Anna Lena, Varga-Szemes, Akos
• Format: Article
• Language: English
• Subjects: Aorta; Arrhythmia; Atrial Fibrillation; Blood Flow Velocity; Breathing; Coding; Correlation; Diagnostic Radiology; Dissection; Flow distribution; Flow velocity; Humans; Image quality; Imaging; Internal Medicine; Interventional Radiology; Magnetic Resonance; Magnetic resonance imaging; Magnetic Resonance Imaging - methods; Medical imaging; Medicine; Medicine & Public Health; Neuroradiology; Phantoms, Imaging; Prototypes; Radiology; Real time; Reproducibility of Results; Respiration; Sequences; Temporal resolution; Two dimensional flow; Ultrasound; Velocity
• ISSN: 1432-1084; 0938-7994; 1432-1084
• DOI: 10.1007/s00330-023-10157-6

Objectives 2D real-time (RT) phase-contrast (PC) MRI is a promising alternative to conventional PC MRI, which overcomes problems due to irregular heartbeats or poor respiratory control. This study aims to evaluate a prototype compressed sensing (CS)-accelerated 2D RT-PC MRI technique with shared velocity encoding (SVE) for accurate beat-to-beat flow measurements. Methods The CS RT-PC technique was implemented using a single-shot fast RF-spoiled gradient echo with SVE by symmetric velocity encoding, and acquired with a temporal resolution of 51–56.5 ms in 1–5 heartbeats. Both aortic dissection phantom ( n  = 8) and volunteer ( n  = 7) studies were conducted using the prototype CS RT (CS, R  = 8), the conventional (GRAPPA, R  = 2), and the fully sampled PC sequences on a 3T clinical system. Flow parameters including peak velocity, peak flow rate, net flow rate, and maximum velocity were calculated to compare the performance between different methods using linear regression, intraclass correlation (ICC), and Bland–Altman analyses. Results Comparisons of the flow measurements at all locations in the phantoms demonstrated an excellent correlation (all R 2  ≥ 0.93) and agreement (all ICC ≥ 0.97) with negligible means of differences. In healthy volunteers, a similarly good correlation (all R 2  ≥ 0.80) and agreement (all ICC ≥ 0.90) were observed; however, CS RT slightly underestimated the maximum velocities and flow rates (~ 12%). Conclusion The highly accelerated CS RT-PC technique is feasible for the evaluation of flow patterns without requiring breath-holding, and it allows for rapid flow assessment in patients with arrhythmia or poor breath-hold capacity. Clinical relevance statement The free-breathing real-time flow MRI technique offers improved spatial and temporal resolutions, as well as the ability to image individual cardiac cycles, resulting in superior image quality compared to the conventional PC technique when imaging patients with arrhythmias, especially those with atrial fibrillation. Key Points • The highly accelerated prototype CS RT-PC MRI technique with improved temporal resolution by the concept of SVE is feasible for beat-to-beat flow evaluation without requiring breath-holding. • The results of the phantom and in vivo quantitative flow evaluation show the ability of the prototype CS RT-PC technique to obtain reliable flow measurements similarly to the conventional PC MRI. • With less than 12% underestimation, excellent agreements between th