Friederike Austein1, Christian Riedel2, Tina Kerby2, Johannes Meyne2, Andreas Binder2, Thomas Lindner2, Monika Huhndorf2, Fritz Wodarg2, Olav Jansen2. 1. Departments of Radiology and Neuroradiology (F.A., C.R., T.K., T.L., M.H., F.W., O.J.) and Neurology (J.M., A.B.), University Hospital, Schleswig-Holstein, Campus Kiel, Germany. friederike.austein@uksh.de. 2. Departments of Radiology and Neuroradiology (F.A., C.R., T.K., T.L., M.H., F.W., O.J.) and Neurology (J.M., A.B.), University Hospital, Schleswig-Holstein, Campus Kiel, Germany.
Abstract
BACKGROUND AND PURPOSE: Computed tomographic perfusion represents an interesting physiological imaging modality to select patients for reperfusion therapy in acute ischemic stroke. The purpose of our study was to determine the accuracy of different commercial perfusion CT software packages (Philips (A), Siemens (B), and RAPID (C)) to predict the final infarct volume (FIV) after mechanical thrombectomy. METHODS: Single-institutional computed tomographic perfusion data from 147 mechanically recanalized acute ischemic stroke patients were postprocessed. Ischemic core and FIV were compared about thrombolysis in cerebral infarction (TICI) score and time interval to reperfusion. FIV was measured at follow-up imaging between days 1 and 8 after stroke. RESULTS: In 118 successfully recanalized patients (TICI 2b/3), a moderately to strongly positive correlation was observed between ischemic core and FIV. The highest accuracy and best correlation are shown in early and fully recanalized patients (Pearson r for A=0.42, B=0.64, and C=0.83; P<0.001). Bland-Altman plots and boxplots demonstrate smaller ranges in package C than in A and B. Significant differences were found between the packages about over- and underestimation of the ischemic core. Package A, compared with B and C, estimated more than twice as many patients with a malignant stroke profile (P<0.001). Package C best predicted hypoperfusion volume in nonsuccessfully recanalized patients. CONCLUSIONS: Our study demonstrates best accuracy and approximation between the results of a fully automated software (RAPID) and FIV, especially in early and fully recanalized patients. Furthermore, this software package overestimated the FIV to a significantly lower degree and estimated a malignant mismatch profile less often than other software.
BACKGROUND AND PURPOSE: Computed tomographic perfusion represents an interesting physiological imaging modality to select patients for reperfusion therapy in acute ischemic stroke. The purpose of our study was to determine the accuracy of different commercial perfusion CT software packages (Philips (A), Siemens (B), and RAPID (C)) to predict the final infarct volume (FIV) after mechanical thrombectomy. METHODS: Single-institutional computed tomographic perfusion data from 147 mechanically recanalized acute ischemic strokepatients were postprocessed. Ischemic core and FIV were compared about thrombolysis in cerebral infarction (TICI) score and time interval to reperfusion. FIV was measured at follow-up imaging between days 1 and 8 after stroke. RESULTS: In 118 successfully recanalized patients (TICI 2b/3), a moderately to strongly positive correlation was observed between ischemic core and FIV. The highest accuracy and best correlation are shown in early and fully recanalized patients (Pearson r for A=0.42, B=0.64, and C=0.83; P<0.001). Bland-Altman plots and boxplots demonstrate smaller ranges in package C than in A and B. Significant differences were found between the packages about over- and underestimation of the ischemic core. Package A, compared with B and C, estimated more than twice as many patients with a malignant stroke profile (P<0.001). Package C best predicted hypoperfusion volume in nonsuccessfully recanalized patients. CONCLUSIONS: Our study demonstrates best accuracy and approximation between the results of a fully automated software (RAPID) and FIV, especially in early and fully recanalized patients. Furthermore, this software package overestimated the FIV to a significantly lower degree and estimated a malignant mismatch profile less often than other software.
Authors: Anderson Chun On Tsang; Stephanie Lenck; Christopher Hilditch; Patrick Nicholson; Waleed Brinjikji; Timo Krings; Vitor M Pereira; Frank L Silver; Joanna D Schaafsma Journal: Clin Neuroradiol Date: 2018-11-23 Impact factor: 3.649
Authors: James E Siegler; Andrew Olsen; Johannes Pulst-Korenberg; Daniel Cristancho; Jon Rosenberg; Lindsay Raab; Brett Cucchiara; Steven R Messé Journal: J Neuroimaging Date: 2019-06-14 Impact factor: 2.486
Authors: M A Almekhlafi; W G Kunz; B K Menon; R A McTaggart; M V Jayaraman; B W Baxter; D Heck; D Frei; C P Derdeyn; T Takagi; A H Aamodt; I M R Fragata; M D Hill; A M Demchuk; M Goyal Journal: AJNR Am J Neuroradiol Date: 2019-01-31 Impact factor: 3.825
Authors: Rujimas Khumtong; Timo Krings; Vitor M Pereira; Aleksandra Pikula; Joanna D Schaafsma Journal: Neuroradiology Date: 2020-01-04 Impact factor: 2.804