Evaluation of cerebral arteriovenous shunts: a comparison of parallel imaging time-of-flight magnetic resonance angiography (TOF-MRA) and compressed sensing TOF-MRA to digital subtraction angiography

• Published in: Neuroradiology 2021-06, Vol.63 (6), p.879-887
• Main Authors: Sakata, Akihiko, Fushimi, Yasutaka, Okada, Tomohisa, Nakajima, Satoshi, Hinoda, Takuya, Speier, Peter, Schmidt, Michaela, Forman, Christoph, Yoshida, Kazumichi, Kataoka, Hiroharu, Miyamoto, Susumu, Nakamoto, Yuji
• Format: Article
• Language: English
• Subjects: Angiography; Diagnostic Neuroradiology; Diagnostic systems; Digital imaging; Imaging; Magnetic resonance; Magnetic resonance imaging; Medical imaging; Medicine; Medicine & Public Health; Neurology; Neuroradiology; Neurosciences; Neurosurgery; Radiology; Shunts; Statistical analysis
• ISSN: 0028-3940; 1432-1920
• DOI: 10.1007/s00234-020-02581-y

Purpose Time-of-flight (TOF)-MR angiography (MRA) is an important imaging sequence for the surveillance and analysis of cerebral arteriovenous shunt (AVS), including arteriovenous malformation (AVM) and arteriovenous fistula (AVF). However, this technique has the disadvantage of a relatively long scan time. The aim of this study was to compare diagnostic accuracy between compressed sensing (CS)-TOF and conventional parallel imaging (PI)-TOF-MRA for detecting and characterizing AVS. Methods This study was approved by the institutional review board for human studies. Participants comprised 56 patients who underwent both CS-TOF-MRA and PI-TOF-MRA on a 3-T MR unit with or without cerebral AVS between June 2016 and September 2018. Imaging parameters for both sequences were almost identical, except the acceleration factor of 3× for PI-TOF-MRA and 6.5× for CS-TOF-MRA, and the scan time of 5 min 19 s for PI-TOF-MRA and 2 min 26 s for CS-TOF-MRA. Two neuroradiologists assessed the accuracy of AVS detection on each sequence and analyzed AVS angioarchitecture. Concordance between CS-TOF, PI-TOF, and digital subtraction angiography was calculated using unweighted and weighted kappa statistics. Results Both CS-TOF-MRA and PI-TOF-MRA yielded excellent sensitivity and specificity for detecting intracranial AVS (reviewer 1, 97.3%, 94.7%; reviewer 2, 100%, 100%, respectively). Interrater agreement on the angioarchitectural features of intracranial AVS on CS-MRA and PI-MRA was moderate to good. Conclusion The diagnostic performance of CS-TOF-MRA is comparable to that of PI-TOF-MRA in detecting and classifying AVS with a reduced scan time under 2.5 min.