机器学习在前循环急性大血管闭塞行机械取栓预后及不良事件发生风险预测中的应用进展  

Research advances in machine learning for prognosis and risk of adverse event prediction after mechanical thrombectomy in acute anterior circulation large vessel occlusion

作　　者：李晨威 杨珂珂 王小军 郭伟花 冯志恒 彭慧渊 Li Chenwei;Yang Keke;Wang Xiaojun;Guo Weihua;Feng Zhiheng;Peng Huiyuan(Department of Neurology,Zhongshan Hospital of Traditional Chinese Medicine(Zhongshan Hospital of Traditional Chinese Medicine Affiliated to Guangzhou University of Traditional Chinese Medicine),Zhongshan,Guangdong 528400,China)

机构地区：[1]中山市中医院(广州中医药大学附属中山中医院)神经内科,528400

出　　处：《中国脑血管病杂志》2025年第3期210-217,共8页Chinese Journal of Cerebrovascular Diseases

基　　金：国家中医药传承创新发展示范试点项目(中山办字[2023]39号);中山市社会公益与基础研究项目(2021B1065)。

摘　　要：前循环急性大血管闭塞性卒中(ALVOS)的临床表现严重,并具有较高的致残率、病死率,机械取栓是目前主要的治疗手段。然而行机械取栓术后该类患者预后差异较大,且预后不良比例较高。机器学习为目前医学领域的研究热点,其可综合分析繁杂数据,提取特异性标志物,辅助临床预测疾病预后及不良事件发生风险。作者就机器学习在前循环ALVOS行机械取栓术后患者的预后、无效再通及出血转化、恶性脑水肿等不良事件发生风险预测的研究进展进行了综述,以期为前循环ALVOS患者制定个体化诊疗方案,改善患者临床结局提供依据。Acute large vessel occlusion stroke(ALVOS)of anterior circulation is associated with severe clinical manifestations and high rates of disability and mortality.Mechanical thrombectomy has emerged as the primary therapeutic intervention.However,post-procedural outcomes remain highly variable,and patients continue to face elevated risks of poor prognosis.Machine learning,a transformative tool in medical research,enables comprehensive analysis of multimodal data to identify specific biomarkers and improve the accuracy of predictions for clinical outcomes and adverse events.This review summarized the latest developments in machine learning applications aim at predicting post-thrombectomy prognosis and risk of adverse event,including futile recanalization,hemorrhagic transformation,and malignant cerebral edema in patients with anterior circulation ALVOS in order to provide a basis for developing personalized treatment plan and improve their clinical prognosis.

关 键 词：急性缺血性卒中 机械取栓 机器学习 预后预测 综述 

分 类 号：R743.3[医药卫生—神经病学与精神病学]