ABSTRACT
Objective To explore the correlation between the total burden of cerebral small vessel disease and poor prognosis of branch atheromatous disease(BAD)in elderly patients.Methods A total of 114 BAD patients admitted to Shanghai Eighth People's Hospital between January 2021 and March 2023 were enrolled,and according to mRS score at 90 d after onset,they were divided into a good prognosis group(mRS score ≤2,67 cases)and a poor prognosis group(mRS score＞2,47 cases).The clinical and imaging characteristics were analyzed,and the relationship between total cerebral small vessel disease burden and clinical prognosis of BAD was investigated using lo-gistic regression analysis.ROC curve analysis was used to determine the threshold of the total cere-bral small vessel disease burden for predicting adverse outcomes and to evaluate its sensitivity and specificity.Results The good prognosis group had younger age,smaller proportion of diabetes,lower SBP,NIHSS score at admission and white matter hyperintensities,and reduced ratio of cerebral microbleeds than the poor prognosis group(P＜0.05,P＜0.01).Statistical difference was observed in the total cerebral small vessel disease burden between the two groups(P＜0.01).Binary logistic regression analysis showed that the total cerebral small vessel disease burden score and NIHSS score at admission were independent predicators of poor prognosis in BAD patients(OR=3.350,95％CI:1.439-7.798,P=0.005;OR=2.814,95％CI:1.586-4.993,P=0.001).ROC curve analysis indicated that the total cerebral small vessel disease burden had a cut-off val-ue of 1.5,and the sensitivity and specificity for predicting poor prognosis was 63.8％and 86.6％,respectively,for BAD patients.Conclusion The total cerebral small vessel disease burden is an in-dependent predictor for poor prognosis of BAD patients.
ABSTRACT
Objective: Neurological outcome prediction in patients with ischemic stroke is very critical in treatment strategy and post-stroke management. Machine learning techniques with high accuracy are increasingly being developed in the medical field. We studied the application of machine learning models to predict long-term neurological outcomes in patients with after intravenous thrombolysis. Methods: A retrospective cohort study was performed to review all stroke patients with intravenous thrombolysis. Patients with modified Rankin Score (mRs) less than two at three months post-thrombolysis were considered as good outcome. The clinical features between stroke patients with good and with poor outcomes were compared using three different machine learning models (Random Forest, Support Vector Machine and Logistic Regression) to identify which performed best. Two datasets from the other stroke center were included accordingly for external verification and performed with explainable AI models. Results: Of the 488 patients enrolled in this study, and 374 (76.6%) patients had favorable outcomes. Patients with higher mRs at 3 months had increased systolic pressure, blood glucose, cholesterol (TC), and 7-day National Institute of Health Stroke Scale (NIHSS) score compared to those with lower mRs. The predictability and the areas under the curves (AUC) for the random forest model was relatively higher than support vector machine and LR models. These findings were further validated in the external dataset and similar results were obtained. The explainable AI model identified the risk factors as well. Conclusion: Explainable AI model is able to identify NIHSS_Day7 is independently efficient in predicting neurological outcomes in patients with ischemic stroke after intravenous thrombolysis.
ABSTRACT
Background and purpose: The prediction of neurological outcomes in ischemic stroke patients is very useful in treatment choices, as well as in post-stroke management. This study is to develop a convenient nomogram for the bedside evaluation of stroke patients with intravenous thrombolysis. Materials and methods: We reviewed all enrolled stroke patients with intravenous thrombolysis retrospectively. Favorable outcome was defined as modified Rankin Score (mRs) less than 2 at 90 days post thrombolysis. We compared the clinical characteristics between patients with favorable outcome and poor outcome. Then, we applied logistic regression models and compared their predictability. Results: A total of 918 patients were enrolled in this study, 448 patients from one hospital were included to develop a nomogram, whereas 470 patients from the other hospital were used for the external validation. Associated risk factors were identified by multivariate logistic regression. The nomogram was validated by the area under the receiver operating characteristic curve (AUC). A nomogram was developed with baseline NIHSS, blood sugar, blood cholesterol level, part-and full anterior circulation infarction (OCSP type). The AUC was 0.767 (95% CI 0.653-0.772) and 0.836 (95% CI 0.697-0.847) in the derivation and external validation cohorts, respectively. The calibration plot for the probability of severe neurological outcome showed an optimal agreement between the prediction by nomogram and actual observation in both derivation and validation cohorts. Conclusion: A convenient outcome evaluation nomogram for patients with intravenous thrombolysis was developed, which could be used by physicians in making clinical decisions and predicting patients' prognosis.