ABSTRACT
Background: Recurrence can worsen conditions and increase mortality in ICH patients. Predicting the recurrence risk and preventing or treating these patients is a rational strategy to improve outcomes potentially. A machine learning model with improved performance is necessary to predict recurrence. Methods: We collected data from ICH patients in two hospitals for our retrospective training cohort and prospective testing cohort. The outcome was the recurrence within one year. We constructed logistic regression, support vector machine (SVM), decision trees, Voting Classifier, random forest, and XGBoost models for prediction. Results: The model included age, NIHSS score at discharge, hematoma volume at admission and discharge, PLT, AST, and CRP levels at admission, use of hypotensive drugs and history of stroke. In internal validation, logistic regression demonstrated an AUC of 0.89 and precision of 0.81, SVM showed an AUC of 0.93 and precision of 0.90, the random forest achieved an AUC of 0.95 and precision of 0.93, and XGBoost scored an AUC of 0.95 and precision of 0.92. In external validation, logistic regression achieved an AUC of 0.81 and precision of 0.79, SVM obtained an AUC of 0.87 and precision of 0.76, the random forest reached an AUC of 0.92 and precision of 0.86, and XGBoost recorded an AUC of 0.93 and precision of 0.91. Conclusion: The machine learning models performed better in predicting ICH recurrence than traditional statistical models. The XGBoost model demonstrated the best comprehensive performance for predicting ICH recurrence in the external testing cohort.
ABSTRACT
INTRODUCTION: Repeat transcranial magnetic stimulation (rTMS) demonstrates beneficial effects for stroke patients, though its efficacy varies due to the complexity of patient conditions and disease progression. Unsupervised machine learning could be the optimal solution for identifying target patients for transcranial magnetic stimulation treatment. METHODS: We collected data from ischaemic stroke patients treated with rTMS. Unsupervised machine learning methods, including K-means and Hierarchical Clustering, were used to explore the clinical characteristics of patients suitable for rTMS. We then utilized a prospective observational cohort to validate the effect of selected characteristics. For the validated cohort, outcomes included the presence of motor evoked potentials (MEP), favorable functional outcomes (FFO), and changes in the Fugl-Meyer Assessment (FMA) at 3 and 6 months. RESULTS: Hierarchical clustering methods revealed that patients in the better prognosis group were more likely to take statins. The validated cohort was grouped based on statin intake. Patients taking statins exhibited a higher rate of MEP (p = 0.006), a higher rate of FFO at 3 months (p = 0.003) and 6 months (p = 0.021), and a more significant change in FMA (p < 0.001) at both 3 and 6 months. Statin intake was associated with FFO and changes in FMA at 3 and 6 months. This relationship persisted across all subgroups for FMA changes and some FFO subgroups. CONCLUSION: Stroke patients undergoing rTMS treatment taking statins exhibited greater MEP, FFO, and changes in FMA. Statin intake was associated with a better prognosis in these patients.
