Hulin Kuang1, Wu Qiu1, Anna M Boers2, Scott Brown3, Keith Muir4, Charles B L M Majoie5,6, Diederik W J Dippel7, Phil White8, Jonathan Epstein9, Peter J Mitchell10, Antoni Dávalos11, Serge Bracard12,13, Bruce Campbell14, Jeffrey L Saver15, Tudor G Jovin16, Marta Rubiera17, Alexander V Khaw18, Jai J Shankar19, Enrico Fainardi20, Michael D Hill1,21,22, Andrew M Demchuk1,21,22, Mayank Goyal1,21,22, Bijoy K Menon1,21,22. 1. Department of Clinical Neurosciences (H.K., W.Q., M.D.H., A.M.D., M.G., B.K.M.), University of Calgary. 2. Department of Biomedical Engineering and Physics (A.M.B.), Amsterdam University Medical Centre. 3. Altair Biostatistics, Mooresville, NC (S.B.). 4. Institute of Neuroscience and Psychology, University of Glasgow, Queen Elizabeth University Hospital (K.M.). 5. Department of Radiology and Nuclear Medicine (C.B.L.M.M.), Amsterdam University Medical Centre. 6. Department of Radiology and Nuclear Medicine, Amsterdam University Medical Centre (C.B.L.M.M.). 7. Department of Neurology, Erasmus University Medical Center (D.W.J.D.). 8. Institute of Neuroscience, Newcastle University (P.W.). 9. Centre Hospitalier Régional et Universitaire de Nancy, Université de Lorraine (J.E.). 10. Department of Radiology, Royal Melbourne Hospital, University of Melbourne (P.J.M.). 11. Department of Neuroscience, Hospital Germans Trias i Pujol, Universitat Autònoma de Barcelona (A.D.). 12. IADI, Inserm, Université de Lorraine, CHRU-Nancy, France (S.B.). 13. Department of Diagnostic and Interventional Neuroradiology, Nancy, France (S.B.). 14. Department of Medicine and Neurology, Royal Melbourne Hospital, University of Melbourne (B.C.). 15. David Geffen School of Medicine, University of Los Angeles (J.L.S.). 16. Cooper Neurological Institute (T.G.J.). 17. Department of Neurology, Hospital Vall d'Hebron, Ps. Vall d'Hebron, Barcelona, Spain (M.R.). 18. Department of Clinical Neurosciences, University of Western Ontario, London, Canada (A.V.K.). 19. Department of Radiology, University of Manitoba, Winnipeg, Canada (J.J.S.). 20. Department of Experimental and Clinical Biomedical Sciences "Mario Serio", University of Florence, Italy (E.F.). 21. Department of Radiology (M.D.H., A.M.D., M.G., B.K.M.), University of Calgary. 22. Hotchkiss Brain Institute (M.D.H., A.M.D., M.G., B.K.M.), University of Calgary.
Abstract
BACKGROUND AND PURPOSE: Prediction of infarct extent among patients with acute ischemic stroke using computed tomography perfusion is defined by predefined discrete computed tomography perfusion thresholds. Our objective is to develop a threshold-free computed tomography perfusion-based machine learning (ML) model to predict follow-up infarct in patients with acute ischemic stroke. METHODS: Sixty-eight patients from the PRoveIT study (Measuring Collaterals With Multi-Phase CT Angiography in Patients With Ischemic Stroke) were used to derive a ML model using random forest to predict follow-up infarction voxel by voxel, and 137 patients from the HERMES study (Highly Effective Reperfusion Evaluated in Multiple Endovascular Stroke Trials) were used to test the derived ML model. Average map, Tmax, cerebral blood flow, cerebral blood volume, and time variables including stroke onset-to-imaging and imaging-to-reperfusion time, were used as features to train the ML model. Spatial and volumetric agreement between the ML model predicted follow-up infarct and actual follow-up infarct were assessed. Relative cerebral blood flow <0.3 threshold using RAPID software and time-dependent Tmax thresholds were compared with the ML model. RESULTS: In the test cohort (137 patients), median follow-up infarct volume predicted by the ML model was 30.9 mL (interquartile range, 16.4-54.3 mL), compared with a median 29.6 mL (interquartile range, 11.1-70.9 mL) of actual follow-up infarct volume. The Pearson correlation coefficient between 2 measurements was 0.80 (95% CI, 0.74-0.86, P<0.001) while the volumetric difference was -3.2 mL (interquartile range, -16.7 to 6.1 mL). Volumetric difference with the ML model was smaller versus the relative cerebral blood flow <0.3 threshold and the time-dependent Tmax threshold (P<0.001). CONCLUSIONS: A ML using computed tomography perfusion data and time estimates follow-up infarction in patients with acute ischemic stroke better than current methods.
BACKGROUND AND PURPOSE: Prediction of infarct extent among patients with acute ischemic stroke using computed tomography perfusion is defined by predefined discrete computed tomography perfusion thresholds. Our objective is to develop a threshold-free computed tomography perfusion-based machine learning (ML) model to predict follow-up infarct in patients with acute ischemic stroke. METHODS: Sixty-eight patients from the PRoveIT study (Measuring Collaterals With Multi-Phase CT Angiography in Patients With Ischemic Stroke) were used to derive a ML model using random forest to predict follow-up infarction voxel by voxel, and 137 patients from the HERMES study (Highly Effective Reperfusion Evaluated in Multiple Endovascular Stroke Trials) were used to test the derived ML model. Average map, Tmax, cerebral blood flow, cerebral blood volume, and time variables including stroke onset-to-imaging and imaging-to-reperfusion time, were used as features to train the ML model. Spatial and volumetric agreement between the ML model predicted follow-up infarct and actual follow-up infarct were assessed. Relative cerebral blood flow <0.3 threshold using RAPID software and time-dependent Tmax thresholds were compared with the ML model. RESULTS: In the test cohort (137 patients), median follow-up infarct volume predicted by the ML model was 30.9 mL (interquartile range, 16.4-54.3 mL), compared with a median 29.6 mL (interquartile range, 11.1-70.9 mL) of actual follow-up infarct volume. The Pearson correlation coefficient between 2 measurements was 0.80 (95% CI, 0.74-0.86, P<0.001) while the volumetric difference was -3.2 mL (interquartile range, -16.7 to 6.1 mL). Volumetric difference with the ML model was smaller versus the relative cerebral blood flow <0.3 threshold and the time-dependent Tmax threshold (P<0.001). CONCLUSIONS: A ML using computed tomography perfusion data and time estimates follow-up infarction in patients with acute ischemic stroke better than current methods.
Authors: Nima Kashani; Petra Cimflova; Johanna M Ospel; Manon Kappelhof; Nishita Singh; Rosalie V McDonough; Mohammed A Almekhlafi; Michael Chen; Nobuyuki Sakai; Jens Fiehler; Uzair Ahmed; Lissa Peeling; Michael Kelly; Mayank Goyal Journal: Clin Neuroradiol Date: 2022-07-19 Impact factor: 3.156
Authors: Anke Wouters; David Robben; Soren Christensen; Henk A Marquering; Yvo B W E M Roos; Robert J van Oostenbrugge; Wim H van Zwam; Diederik W J Dippel; Charles B L M Majoie; Wouter J Schonewille; Aad van der Lugt; Maarten Lansberg; Gregory W Albers; Paul Suetens; Robin Lemmens Journal: Stroke Date: 2021-09-30 Impact factor: 7.914