Seguimiento mediante angio-resonancia magnética de aneurismas cerebrales embolizados: concordancia entre lectores

Abstract

Objective To determine the interobserver agreement in the interpretation of MR angiography (MRA) studies for surveillance of embolized intracranial aneurysms. To determine whether contrast administration improves interobserver agreement. Material and methods Two experienced neuroradiologists independently reviewed all follow-up MRA studies performed between July 2004 and December 2006 of cerebral aneurysms embolized with coils. All MRA studies included both unenhanced 3D time-of-flight (3D TOF) and contrast-enhanced MRA (CE-MRA) images. Studies were classified as: a) not assessable; b) complete occlusion; c) residual aneurysm. Interobserver agreement for unenhanced and enhanced MRA studies was determined using the kappa statistic. Kappa values were considered insignificant when<0.2, low when between 0.21 and 0.4, and moderate when between 0.41–0.6; values >0.6 were considered good agreement and >0.8 excellent agreement. Significance was set at p<0.005. Results We reviewed a total of 200 MRA studies (100 3D TOF studies and 100 CE-MRA studies) performed in 48 patients (25 women, 23 men) at 6, 12, and/or 24 months after embolization. Interobserver agreement was good in both 3D TOF and CE-MRA studies, although it was better in CE-MRA studies (κ=0.660, p<0.001 and κ=0.779, p<0.001, respectively). Conclusions Interobserver agreement is good for follow-up MRA studies of embolized intracranial aneurysms. Gadolinium administration improves interobserver agreement.