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

Abstract

BackgroundTimely 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 a 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).MethodsThis 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 into training (N = 303) and validation cohorts (N = 101) according to a ratio of 75–25%. One LR and six popular ML algorithms were used to construct models. The area under the receiver operating characteristic curve (AUC), accuracy, balanced accuracy, confusion matrix, sensitivity, specificity, calibration curve, and Hosmer–Lemeshow test were used to assess and compare the model performance. Finally, we calculated each feature of importance.ResultsA 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 the validation cohort 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 = 0.46), 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 = 0.50). However, random forest (RF) and artificial neural network model with the same AUC (0.858, 95%CI: 0.78–0.93, P = 0.26) were better than the LR. The accuracy and the balanced accuracy of the RF were 20.8% and 11% higher than the latter, and the RF also showed good calibration in the 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 in the RF model. We then developed an online prediction tool (https://dynamic-nomogram.shinyapps.io/DynNomapp-DCI/) based on important features to calculate DCI risk precisely.ConclusionsIn 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.Clinical Trial Registrationhttp://www.chictr.org.cn, Unique identifier: ChiCTR2100044448.

Prognostic model for aneurysmal subarachnoid hemorrhage patients requiring mechanical ventilation

“…Hu et al. developed an online prediction tool based on a random forest model to identify patients at a high risk of delayed cerebral ischemia after aSAH 30 . Fang et al.…”

“…29 Hu et al developed an online prediction tool based on a random forest model to identify patients at a high risk of delayed cerebral ischemia after aSAH. 30 Fang et al reported that elevated D-dimer levels on admission were associated with short-and long-term mortality. 31 Wi sniewski et al found that glucose-6phosphate dehydrogenase and 8-iso-prostaglandin F2a are potential predictors of delayed cerebral ischemia after aSAH.…”

Artificial Intelligence and Machine Learning Applications in Critically Ill Brain Injured Patients

“…153 Supervised ML models using broader EHR and imaging data confirmed known DCI risk factors like SAH volume and initial GCS, though also identified new predictors such as leukocytosis and neutrophilia, indicating a potential neuroinflammatory component in the disease process. 154,155 Similarly, an integrative ML model showed superior predictive power over logistic regression (AUC: 0.74 vs. 0.63) and identified potentially modifiable risk factors like treatment modality and timing from ictus to imaging by employing a Local Interpretable Model-agnostic Explanation (LIME) program that provides clear interpretation of each model prediction decision. 156 Furthermore, ML has also been instrumental in discovering novel biomarkers that could enhance DCI and clinical outcome models and provide insights into the molecular and physiological changes in this complex condition.…”

Time-Series Modeling and Forecasting of Cerebral Pressure–Flow Physiology: A Scoping Systematic Review of the Human and Animal Literature

“…Similarly, work including transfer function analysis (TFA), the continuous wavelet transform (CWT), empirical mode decomposition (EMD), fast Fourier transform (FFT), cross-spectral analysis, wavelet analysis, Granger causality, autoregressive (AR) models, and so on [ 9 ] have been described. In addition to the statistical time-series analysis techniques, there are various machine learning algorithms used for cerebral physiologic signal modeling [ 10 , 11 ] as well as for the prediction task [ 12 ], including models such as linear regression, artificial neural networks (ANNs), convolutional neural networks (CNNs), extreme gradient boosting (XGBoost), and decision trees. These algorithms offer adaptability and data-driven capabilities that can uncover intricate patterns within the data, particularly in cases where complexities demand more flexible modeling approaches [ 13 ].…”