Meeting Banner
Abstract #1029

Arterial Lesions on CT Angiography Are Not Precise Forecasters of PWI & DWI Mismatches in Acute Ischemic Stroke

Raquel da Cruz Bezerra1, William A. Copen2, Elissa McIntosh1, Izzuddin Diwan1, Priya Garg1, Steven Mocking1, Taylor W. Kimberly3, Ethem Murat Arsava1, Ruijun Ji3, Hakan Ay1, Aneesh B. Singhal3, Pamela W. Schaefer2, Ona Wu1

1Radiology, MGH, Athinoula A Martinos Center for Biomedical Imaging, Boston, MA, United States; 2Radiology, Massachusetts General Hospital, Boston, MA, United States; 3Neurology, Massachusetts General Hospital, Boston, MA, United States

We investigated the relationship between proximal arterial occlusion/stenosis on CTA and DWI-PWI mismatch in acute ischemic stroke patients (N=95). The sensitivity of proximal occlusion/stenosis in predicting DWI-PWI mismatch was 84% [95% CI: 72-92%] for MTT, and 85% [73-93%] for Tmax. The specificity of proximal occlusion/stenosis in predicting mismatch was 67% [48-81%] for MTT and 68% [49-82%] for Tmax. One third of patients without occlusion/stenosis had mismatch. Approximately 20% of patients with occlusion/stenosis did not exhibit mismatch. Therefore, a combination of vessel imaging and PWI would be pertinent for selecting patients likely to benefit from extended time-window therapies.

Abstract PDF Poster PDF

Keywords

absence acute additionally admission angiography anterior approved arrow arterial artery assist automated automatically available biomedical brain branch carotid categorical cerebral clock complementary confidence continuous critical damaged days decomposition deconvolution defined delineated demographic despite determined diffusion display divided either evaluated exhibit experimental extended fewer finding findings flair forecasters forth function functions furthermore general give goal growth hence holes hospital hour importance included index infarct input instead internal intervention intra intravenous irreversible known lancet last lately lesion lesions likely male manually maps matching median middle mismatch mismatches mocking neither neurology occlusion offer opinion parametric patients perfusion play posterior postulated potentially precise predicting predicts presence prevalence processed proximal radiology rank readily receive registered regularized relationship relying researchers retrospectively role salvageable segment selected sensitivity significantly singular specificity spurious stroke subjective subjects suggested table tailed territories therapy third threshold thresholds tissue transit treated triage type variables vascular vertebral visible visually volume volumes weber whether window years