B Jiang1, G Zhu1, Y Xie1, J J Heit1, H Chen1, Y Li1, V Ding2, A Eskandari3, P Michel3, G Zaharchuk1, M Wintermark4. 1. From the Department of Radiology, Neuroradiology Section (B.J., G.Z., Y.X., J.J.H., H.C., Y.L., G.Z., M.W.), Stanford University School of Medicine, Palo Alto, California. 2. Department of Medicine (V.D.), Quantitative Sciences Unit, Stanford University, Stanford, California. 3. Neurology Service (A.E., P.M.), Centre Hospitalier Universitaire Vaudois and Lausanne University, Lausanne, Switzerland. 4. From the Department of Radiology, Neuroradiology Section (B.J., G.Z., Y.X., J.J.H., H.C., Y.L., G.Z., M.W.), Stanford University School of Medicine, Palo Alto, California max.wintermark@gmail.com.
Abstract
BACKGROUND AND PURPOSE: Traditional statistical models and pretreatment scoring systems have been used to predict the outcome for acute ischemic stroke patients (AIS). Our aim was to select the most relevant features in terms of outcome prediction on the basis of machine learning algorithms for patients with acute ischemic stroke and to compare the performance between multiple models and the Stroke Prognostication Using Age and National Institutes of Health Stroke Scale (SPAN-100) index model. MATERIALS AND METHODS: A retrospective multicenter cohort of 1431 patients with acute ischemic stroke was subdivided into recanalized and nonrecanalized patients. Extreme Gradient Boosting machine learning models were built to predict the mRS score at 90 days using clinical, imaging, combined, and best-performing features. Feature selection was performed using the relative weight and frequency of occurrence in the models. The model with the best performance was compared with the SPAN-100 index model using area under the receiver operating curve analysis. RESULTS: In 3 groups of patients, the baseline NIHSS was the most significant predictor of outcome among all the parameters, with relative weights of 0.36∼0.69; ischemic core volume on CTP ranked as the most important imaging biomarker with relative weights of 0.29∼0.47. The model with the best-performing features had a better performance than the other machine learning models. The area under the curve of the model with the best-performing features was higher than SPAN-100 model and reached statistical significance for the total (P < .05) and the nonrecanalized patients (P < .001). CONCLUSIONS: Machine learning-based feature selection can identify parameters with higher performance in outcome prediction. Machine learning models with the best-performing features, especially advanced CTP data, had superior performance of the recovery outcome prediction for patients with stroke at admission in comparison with SPAN-100.
BACKGROUND AND PURPOSE: Traditional statistical models and pretreatment scoring systems have been used to predict the outcome for acute ischemic stroke patients (AIS). Our aim was to select the most relevant features in terms of outcome prediction on the basis of machine learning algorithms for patients with acute ischemic stroke and to compare the performance between multiple models and the Stroke Prognostication Using Age and National Institutes of Health Stroke Scale (SPAN-100) index model. MATERIALS AND METHODS: A retrospective multicenter cohort of 1431 patients with acute ischemic stroke was subdivided into recanalized and nonrecanalized patients. Extreme Gradient Boosting machine learning models were built to predict the mRS score at 90 days using clinical, imaging, combined, and best-performing features. Feature selection was performed using the relative weight and frequency of occurrence in the models. The model with the best performance was compared with the SPAN-100 index model using area under the receiver operating curve analysis. RESULTS: In 3 groups of patients, the baseline NIHSS was the most significant predictor of outcome among all the parameters, with relative weights of 0.36∼0.69; ischemic core volume on CTP ranked as the most important imaging biomarker with relative weights of 0.29∼0.47. The model with the best-performing features had a better performance than the other machine learning models. The area under the curve of the model with the best-performing features was higher than SPAN-100 model and reached statistical significance for the total (P < .05) and the nonrecanalized patients (P < .001). CONCLUSIONS: Machine learning-based feature selection can identify parameters with higher performance in outcome prediction. Machine learning models with the best-performing features, especially advanced CTP data, had superior performance of the recovery outcome prediction for patients with stroke at admission in comparison with SPAN-100.
Authors: Raul G Nogueira; Ashutosh P Jadhav; Diogo C Haussen; Alain Bonafe; Ronald F Budzik; Parita Bhuva; Dileep R Yavagal; Marc Ribo; Christophe Cognard; Ricardo A Hanel; Cathy A Sila; Ameer E Hassan; Monica Millan; Elad I Levy; Peter Mitchell; Michael Chen; Joey D English; Qaisar A Shah; Frank L Silver; Vitor M Pereira; Brijesh P Mehta; Blaise W Baxter; Michael G Abraham; Pedro Cardona; Erol Veznedaroglu; Frank R Hellinger; Lei Feng; Jawad F Kirmani; Demetrius K Lopes; Brian T Jankowitz; Michael R Frankel; Vincent Costalat; Nirav A Vora; Albert J Yoo; Amer M Malik; Anthony J Furlan; Marta Rubiera; Amin Aghaebrahim; Jean-Marc Olivot; Wondwossen G Tekle; Ryan Shields; Todd Graves; Roger J Lewis; Wade S Smith; David S Liebeskind; Jeffrey L Saver; Tudor G Jovin Journal: N Engl J Med Date: 2017-11-11 Impact factor: 91.245
Authors: Charith Cooray; Michael Mazya; Matteo Bottai; Laura Dorado; Ondrej Skoda; Danilo Toni; Gary A Ford; Nils Wahlgren; Niaz Ahmed Journal: Stroke Date: 2016-05-12 Impact factor: 7.914
Authors: H P Adams; P H Davis; E C Leira; K C Chang; B H Bendixen; W R Clarke; R F Woolson; M D Hansen Journal: Neurology Date: 1999-07-13 Impact factor: 9.910
Authors: I Y L Tan; A M Demchuk; J Hopyan; L Zhang; D Gladstone; K Wong; M Martin; S P Symons; A J Fox; R I Aviv Journal: AJNR Am J Neuroradiol Date: 2009-01-15 Impact factor: 3.825
Authors: Stefan Winzeck; Arsany Hakim; Richard McKinley; José A A D S R Pinto; Victor Alves; Carlos Silva; Maxim Pisov; Egor Krivov; Mikhail Belyaev; Miguel Monteiro; Arlindo Oliveira; Youngwon Choi; Myunghee Cho Paik; Yongchan Kwon; Hanbyul Lee; Beom Joon Kim; Joong-Ho Won; Mobarakol Islam; Hongliang Ren; David Robben; Paul Suetens; Enhao Gong; Yilin Niu; Junshen Xu; John M Pauly; Christian Lucas; Mattias P Heinrich; Luis C Rivera; Laura S Castillo; Laura A Daza; Andrew L Beers; Pablo Arbelaezs; Oskar Maier; Ken Chang; James M Brown; Jayashree Kalpathy-Cramer; Greg Zaharchuk; Roland Wiest; Mauricio Reyes Journal: Front Neurol Date: 2018-09-13 Impact factor: 4.003