Interest of HYPR flow dynamic MRA for characterization of cerebral arteriovenous malformations: comparison with TRICKS MRA and catheter DSA

European radiology(2015)

引用 11|浏览12
暂无评分
摘要
Objective HYPR flow is a 3D dynamic contrast-enhanced MRA technique providing isotropic sub-millimetre resolution with half-second temporal resolution. We compared HYPR flow and time-resolved imaging of contrast kinetics (TRICKS) MRA for the characterization of cerebral arteriovenous malformations (cAVMs), using catheter DSA as reference. Methods Twenty-two patients underwent HYPR flow and TRICKS MRA within 15 days of DSA. HYPR flow and TRICKS datasets were reviewed separately by two readers for image quality, Spetzler–Martin grade, venous ectasia, and deep venous drainage. Results Image quality was better for HYPR flow than for TRICKS (narrower full width at half maximum; larger arterial diagnostic window; greater number of arterial frames, P ≤ 0.05). Using HYPR flow, inter-reader agreement was excellent for all cAVM characteristics. The agreement with DSA for the overall Spetzler–Martin grade was excellent for HYPR flow (ICC = 0.96 and 0.98, depending on the reader) and TRICKS (ICC = 0.82 and 0.95). In comparison to TRICKS, HYPR flow showed higher concordance with DSA for the identification of venous ectasia and deep venous drainage. Conclusion Owing to an excellent agreement with DSA with respect to depiction of the vascular architecture of cAVMs, HYPR flow could be useful for the non-invasive characterization of cAVMs. Key Points • Dynamic MRA is used for cerebral AVM depiction and follow-up • HYPR flow is a new, highly-resolved dynamic MRA sequence • HYPR flow provides whole brain coverage • HYPR flow provides excellent agreement with the Spetzler–Martin grade • Compared to TRICKS MRA, HYPR flow improves cerebral AVM characterization
更多
查看译文
关键词
Intracranial arteriovenous malformations, Time-resolved contrast-enhanced magnetic resonance angiography, Digital subtraction angiography, Brain
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要