| 摘要: |
| 本研究旨在探讨静脉溶栓时间窗外，急性脑梗死（AIS）患者入院时常规临床资料对其短期不良结局的影响。该研究通过收集我院神经内科住院治疗的溶栓时间窗外AIS患者（774例）临床资料，比较短期不同预后患者入院时临床指标，并利用机器学习分类算法建模分析患者90 d内预后不良的影响因素。采用受试者工作特征曲线（ROC）和校准图验证列线图模型的预测效能及准确度。研究结果显示：（1）3个月随访期内预后不良者占比13.95%（108例）。（2）预后不良患者年龄、血白细胞（WBC）水平、血C反应蛋白（CRP）水平、血肿甘油三酯（TG）水平、入院国立卫生研究院卒中评分（NIHSS）、梗死体积和发病-治疗时间大于或高于预后良好患者；预后不良患者入院GCS评分低于预后良好组，均差异显著（P均＜0.05）。机器学习分类算法中，极限梯度提升（XGB）方法的效果最佳（受试者工作特征曲线下面积（AUC）=0.81），且校准曲线显示模型评估预测风险与实际发生风险的一致程度高。（3）XGB算法筛选的预测变量影响权重居前6位的依次为发病-治疗时间、入院时血CRP水平、年龄、梗死体积、TG水平与NIHSS评分。上述结果说明机器学习XGB算法可用于预测静脉溶栓时间窗外AIS患者短期不良预后的影响因素。 |
| 关键词: 急性脑梗死 超溶栓时间窗 预后 机器学习 临床预测模型 |
| DOI： |
| 投稿时间：2023-10-17修订日期：2024-08-15 |
| 基金项目:广西壮族自治区卫生健康委员会科研课题 |
|
| Establishment and Validation of a Short-term Prognostic Model for Acute Ischemic Stroke Patients Outside the Intravenous Thrombolysis Time Window Based on Machine Learning |
|
Zhou Hui1, FENG Bing2, LIU Xuri1, ZOU Yong1, TANG Hui1, HUANG Qinbin2, CHEN Tingjun2
|
| (1.Translational Chinese Medical of Guiping City,Guiping;2.Guiping People’s Hospital,Guiping) |
| Abstract: |
| The objective is to investigate the impact of clinical data at admission on the short-term adverse outcomes of acute ischemic stroke (AIS) patients outside the intravenous thrombolysis time window. Clinical data of 774 AIS patients treated in the neurology department of our hospital outside the thrombolysis time window were collected. The clinical indicators at admission of patients with different short-term prognoses were compared, and machine learning classification algorithms were used to model and analyze the factors influencing poor prognosis within 90 days. The predictive performance and accuracy of the nomogram model were validated using the receiver operating characteristic (ROC) curve and calibration plot. The results showed: (1) 13.95% (108 cases) of patients had poor prognosis within the 3-month follow-up period. (2) Patients with poor prognosis had higher or greater age, blood white blood cell (WBC) level, blood C-reactive protein (CRP) level, blood triglyceride (TG) level, National Institutes of Health Stroke Scale (NIHSS) score at admission, infarct volume, and onset-to-treatment time compared to patients with good prognosis; patients with poor prognosis had lower Glasgow Coma Scale (GCS) scores at admission, with significant differences (P < 0.05). The XGB algorithm performed best (AUC = 0.81), and the calibration curve showed a high degree of consistency between the model's predicted risk and the actual occurrence of risk. (3) The top six predictive variables selected by the XGB algorithm were onset-to-treatment time, blood CRP level at admission, age, infarct volume, TG level, and NIHSS score. These results indicate that the XGB machine learning algorithm can be used to predict the factors influencing short-term adverse prognosis in AIS patients outside the intravenous thrombolysis time window. |
| Key words: Acute cerebral infarction broadened therapeutic window prognosis machine learning clinical prediction model |
