Research Performance of e-ASPECTS software in comparison to that of stroke physicians on assessing CT scans of acute ischemic stroke patients Christian Herweh 1 , Peter A Ringleb 2 , Geraldine Rauch 3 , Steven Gerry 4 , Lars Behrens 1 , Markus Mo ¨ hlenbruch 1 , Rebecca Gottorf 2 , Daniel Richter 2 , Simon Schieber 2 and Simon Nagel 2 Abstract Background: The Alberta Stroke Program Early CT score (ASPECTS) is an established 10-point quantitative topo- graphic computed tomography scan score to assess early ischemic changes. We compared the performance of the e-ASPECTS software with those of stroke physicians at different professional levels. Methods: The baseline computed tomography scans of acute stroke patients, in whom computed tomography and diffusion-weighted imaging scans were obtained less than two hours apart, were retrospectively scored by e-ASPECTS as well as by three stroke experts and three neurology trainees blinded to any clinical information. The ground truth was defined as the ASPECTS on diffusion-weighted imaging scored by another two non-blinded independent experts on consensus basis. Sensitivity and specificity in an ASPECTS region-based and an ASPECTS score-based analysis as well as receiver-operating characteristic curves, Bland–Altman plots with mean score error, and Matthews correlation coeffi- cients were calculated. Comparisons were made between the human scorers and e-ASPECTS with diffusion-weighted imaging being the ground truth. Two methods for clustered data were used to estimate sensitivity and specificity in the region-based analysis. Results: In total, 34 patients were included and 680 (34  20) ASPECTS regions were scored. Mean time from onset to computed tomography was 172  135 min and mean time difference between computed tomographyand magnetic res- onance imaging was 41  31 min. The region-based sensitivity (46.46% [CI: 30.8;62.1]) of e-ASPECTS was better than three trainees and one expert (p  0.01) and not statistically different from another two experts. Specificity (94.15% [CI: 91.7;96.6]) was lower than one expert and one trainee (p < 0.01) and not statistically different to the other four physicians. e-ASPECTS had the best Matthews correlation coefficient of 0.44 (experts: 0.38  0.08 and trainees: 0.19  0.05) and the lowest mean score error of 0.56 (experts: 1.44  1.79 and trainees: 1.97  2.12). Conclusion: e-ASPECTS showed a similar performance to that of stroke experts in the assessment of brain computed tomographys of acute ischemic stroke patients with the Alberta Stroke Program Early CT score method. Keywords Alberta Stroke Program Early CT score, computed tomography, ischemic stroke, machine learning Received: 20 May 2015; accepted: 12 November 2015 Introduction The interpretation of a non-contrast-enhanced com- puted tomography (CT) brain scan of a patient with an acute ischemic stroke before thrombolysis or thrombectomy is a challenging task. A quantitative evaluation of early focal ischemic damage based on a 1 Department of Neuroradiology, University Hospital Heidelberg, Germany 2 Department of Neurology, University Hospital Heidelberg, Germany 3 Institute of Medical Biometry and Informatics, University of Heidelberg, Germany 4 Centre for Statistics in Medicine, University of Oxford, United Kingdom Corresponding author: Simon Nagel, Department of Neurology, University Hospital Heidelberg, INF 400, 69120 Heidelberg, Germany. Email: simon.nagel@med.uni-heidelberg.de International Journal of Stroke, 11(4) International Journal of Stroke 2016, Vol. 11(4) 438–445 ! 2016 World Stroke Organization Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/1747493016632244 wso.sagepub.com