AUTHOR=Hu Ping , Li Yuntao , Liu Yangfan , Guo Geng , Gao Xu , Su Zhongzhou , Wang Long , Deng Gang , Yang Shuang , Qi Yangzhi , Xu Yang , Ye Liguo , Sun Qian , Nie Xiaohu , Sun Yanqi , Li Mingchang , Zhang Hongbo , Chen Qianxue 

TITLE=Comparison of Conventional Logistic Regression and Machine Learning Methods for Predicting Delayed Cerebral Ischemia After Aneurysmal Subarachnoid Hemorrhage: A Multicentric Observational Cohort Study

JOURNAL=Frontiers in Aging Neuroscience

VOLUME=Volume 14 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/aging-neuroscience/articles/10.3389/fnagi.2022.857521

DOI=10.3389/fnagi.2022.857521

ISSN=1663-4365

ABSTRACT=Abstract
Background: Timely and accurate prediction of delayed cerebral ischemia is critical for improving the prognosis of patients with aneurysmal subarachnoid hemorrhage. Machine learning (ML) algorithms are increasingly regarded as having higher prediction power than conventional logistic regression (LR). This study aims to construct LR and ML models and compare their prediction power on delayed cerebral ischemia (DCI) after aneurysmal subarachnoid hemorrhage (aSAH). 
Methods: This was a multicenter, retrospective, observational cohort study that enrolled patients with aneurysmal subarachnoid hemorrhage from five hospitals in China. A total of 404 aSAH patients were prospectively enrolled. We randomly divided the patients as training (N=303) and validation cohorts (N=101) according to a ratio of 75% to 25%. One LR and six popular ML algorithms were used to construct models. Area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, specificity, calibration curve, and Hosmer-Lemeshow test were used to assess and compare the model performance. 
Results: A total of 112 (27.7%) patients developed DCI. Our results showed that conventional LR with an AUC value of 0.824 (95%CI: 0.73-0.91) in external validation outperformed k-nearest neighbor, decision tree, support vector machine, and extreme gradient boosting model with the AUCs of 0.792 (95%CI: 0.68-0.9, P=0.03), 0.675 (95%CI: 0.56-0.79, P<0.01), 0.677 (95%CI: 0.57-0.77, P<0.01), and 0.78 (95%CI: 0.68-0.87, P=0.02). However, random forest (RF) and artificial neural network model with a same AUC (0.858, 95%CI: 0.78-0.93, P=0.03) performed well than the LR. The accuracy of the RF was 20.8% higher than the latter, and it also showed good calibration in validation cohort (Hosmer-Lemeshow: P =0.203). We found that the CT value of subarachnoid hemorrhage, WBC count, neutrophil count, CT value of cerebral edema, and monocyte count were the five most important features for DCI prediction. We then developed an online prediction tool (https://dynamic-nomogram.shinyapps.io/DynNomapp-DCI/) based on RF algorithm to calculate DCI risk precisely.
Conclusions: In this multicenter study, we found that several ML methods, particularly RF, outperformed conventional LR. Furthermore, an online prediction tool based on the RF model was developed to identify patients at high risk for DCI after SAH and facilitate timely interventions.
Registration-URL: http://www.chictr.org.cn; Unique identifier: ChiCTR2100044448.