流体衰减反转恢复
医学
核医学
神经组阅片室
放射科
梅尼埃病
磁共振成像
神经学
外科
眩晕
精神科
作者
Anja Bernaerts,Nick Janssen,Floris L. Wuyts,Cathérine Blaivie,Robby Vanspauwen,Joost van Dinther,Andrzej Zarowski,Erwin Offeciers,F. Deckers,Jan Casselman,Bert De Foer
出处
期刊:Neuroradiology
[Springer Science+Business Media]
日期:2022-02-12
卷期号:64 (5): 1011-1020
被引量:11
标识
DOI:10.1007/s00234-022-02913-0
摘要
Heavily T2-weighted 3D FLAIR (hT2w-3D-FLAIR) sequence with constant flip angle (CFA) has been reported as being more sensitive to low concentrations of gadolinium (Gd) enabling endolymphatic hydrops (EH) visualization. The purpose of this study was to compare signal-to-noise (SNR) ratio, detection rate of EH, and increased perilymphatic enhancement (PE) as well as diagnostic accuracy in diagnosing definite Menière's disease (MD), using 3D-SPACE FLAIR versus conventional 3D-TSE FLAIR.This retrospective study included 29 definite MD patients who underwent a 4-h delayed intravenous (IV) Gd-enhanced 3D-TSE FLAIR and 3D-SPACE FLAIR MRI between February 2019 and February 2020. MR images were qualitatively and quantitatively analyzed twice by 2 experienced head and neck radiologists. Qualitative assessment included grading of cochlear and vestibular EH and visual comparison of PE. Quantitative assessment of PE was performed by placing a region of interest (ROI) and ratio calculation in the basal turn of the cochlea and the brainstem.The intra- and inter-reader reliability for grading of EH and PE was excellent (0.7 < kappa < 0.9) for 3D-SPACE FLAIR and exceeded the values for 3D-TSE FLAIR (0.5 < kappa < 0.9) The combination of EH and visual assessment of PE has the highest diagnostic accuracy in diagnosing definite MD on 3D-SPACE FLAIR with a sensitivity of 0.91 and a specificity of 0.98 resulting in a sensitivity raise of 6% compared to 3D-TSE FLAIR.Four-hour delayed IV Gd-enhanced 3D-SPACE FLAIR sequence has a higher sensitivity and reproducibility than 3D-TSE FLAIR for the visualization of EH and increased PE in definite MD patients.
科研通智能强力驱动
Strongly Powered by AbleSci AI