Please use this identifier to cite or link to this item:
Title: Comparison of different methods of thrombus permeability measurement and impact on recanalization in the INTERRSeCT multinational multicenter prospective cohort study
Authors: Evans, James W ;Gensicke, H.;Al-Ajlan, F.S.;Dowlatshahi, D.;Najm, M.;Calleja, A.L.;Puig, J.;Sohn, S.L.;Ahn, S.H.;Poppe, A.Y.;Mikulik, R.;Asdaghi, N.;Field, T.S.;Jin, A.;Asil, T.;Boulanger, J.M.;Hill, M.D.;Goyal, M.;Demchuk, A.M.;Menon, B.K.
Affliation: Central Coast Local Health District
Gosford Hospital
Issue Date: Nov-2019
Source: 62(3):301-306
Journal title: Neuroradiology
Department: Neuroscience
Abstract: PURPOSE: To compare the association of different measures of intracranial thrombus permeability on non-contrast computerized tomography (NCCT) and computed tomography angiography (CTA) with recanalization with or without intravenous alteplase. METHODS: Patients with anterior circulation occlusion from the INTERRSeCT study were included. Thrombus permeability was measured on non-contrast CT and CTA using the following methods: [1] automated method, mean attenuation increase on co-registered thin (< 2.5 mm) CTA/NCCT; [2] semi-automated method, maximum attenuation increase on non-registered CTA/NCCT (DeltaHUmax); [3] manual method, maximum attenuation on CTA (HUmax); and [4] visual method, residual flow grade. Primary outcome was recanalization with intravenous alteplase on the revised AOL scale (2b/3). Regression models were compared using C-statistic, Akaike (AIC), and Bayesian information criterion (BIC). RESULTS: Four hundred eighty patients were included in this analysis. Statistical models using methods 2, 3, and 4 were similar in their ability to discriminate recanalizers from non-recanalizers (C-statistic 0.667, 0.683, and 0.634, respectively); method 3 had the least information loss (AIC = 483.8; BIC = 492.2). A HUmax >/= 89 measured with method 3 provided optimal sensitivity and specificity in discriminating recanalizers from non-recanalizers [recanalization 55.4% (95%CI 46.2-64.6) when HUmax > 89 vs. 16.8% (95%CI 13.0-20.6) when HUmax </= 89]. In sensitivity analyses restricted to patients with co-registered CTA/NCCT (n = 88), methods 1-4 predicted recanalization similarly (C-statistic 0.641, 0.688, 0.640, 0.648, respectively) with Method 2 having the least information loss (AIC 104.8, BIC 109.8). CONCLUSION: Simple methods that measure thrombus permeability are as reliable as complex image processing methods in discriminating recanalizers from non-recanalizers.
DOI: 10.1007/s00234-019-02320-y
ISSN: 0028-3940
Publicaton type: Journal Article
Keywords: Neurology
Study or Trial: Prospective Cohort Study
Appears in Collections:Neurology

Show full item record

Page view(s)

checked on Jan 28, 2023

Google ScholarTM



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.