Patient Outcomes to Evaluate Machine Outputs

Abstract

We read with interest the letter to the Editor by Drs. Kellner and Urbach [1] commenting on our recently published article “Automated Perfusion Calculations vs. Visual Scoring of Collaterals and CBV-ASPECTS” [2]. First, we need to highlight that the primary aim of our work was to investigate the association of the core and penumbral estimates derived from different perfusion software, collaterals and the cerebral blood volume Alberta Stroke Program early CT score (CBV-ASPECTS) with patient functional outcomes at 3 months [2]. Exploring for potential differences in ischemic core and penumbral volume estimates between different automated software was the secondary aim of our work [2]. Second, Drs. Kellner and Urbach statement in their letter that “VEOcore (VEObrain GmbH, Freiburg im Breisgau, Germany) and syngo.via (Siemens Healthineers, Erlangen, Germany) overestimate the infarct core and Olea (OLEA medical Inc., La Ciotat, France) underestimates it, respectively” [1], constitutes a personal interpretation of the presented data, while is not in accordance with our conclusions, clearly stating that “core and penumbra estimates using automated software packages vary significantly and should therefore be used with caution” [2]. In our manuscript we presented the results of all pairwise analyses between different software and make no specific recommendation that “VEOcore software is inappropriate for treatment decisions beyond the 6 hours window”, as stated in the letter by Drs. Kellner and Urbach [1]. We express