首页|期刊导航|中国脑血管病杂志|人工智能在急性缺血性卒中多模态CT病灶识别与分割中的应用研究进展

人工智能在急性缺血性卒中多模态CT病灶识别与分割中的应用研究进展OA

Advances in the application of artificial intelligence for multimodal CT-based lesion identification and segmentation in acute ischemic stroke

中文摘要英文摘要

急性缺血性卒中(AIS)起病急骤,致残与致死率高.多模态CT是AIS急诊评估中应用最广泛的影像学方法,包括平扫CT(NCCT)、CT灌注(CTP)和CT血管成像(CTA),但三者分别存在早期病灶检出敏感度有限、灌注参数定量受阈值和后处理差异影响、中远端血管闭塞识别准确性偏低等局限.近年来,基于上述3种CT模态构建的人工智能模型在AIS病灶识别与分割任务中展现出优于传统方法或人工判读的潜能.该文拟分别对基于NCCT、CTP及CTA的人工智能模型在AIS病灶识别与分割中的研究进展进行综述,以期为人工智能在AIS影像学诊断中的转化应用提供参考.

Acute ischemic stroke(AIS)presents with sudden onset and carries a high risk of disability and mortality.Multimodal CT,including non-contrast CT(NCCT),CT perfusion(CTP),and CT angiography(CTA),is the most widely used imaging approach for emergency AIS evaluation.However,these modalities have inherent limitations:NCCT has limited sensitivity for early ischemic lesions,CTP-derived quantitative parameters are affected by threshold selection and postprocessing variability,and CTA has relatively low accuracy for detecting medium-to-distal vascular occlusions.In recent years,artificial intelligence(AI)models based on these CT modalities have shown potential advantages over conventional methods or human interpretation in lesion identification and segmentation.This review summarized advances in AI-based lesion identification and segmentation using NCCT,CTP,and CTA in AIS,aiming to provide a reference for the translational application of AI in AIS imaging diagnosis.

张人元;沈文卓;梁雨昕;赵晓静;张苗;卢洁

100053 北京,首都医科大学宣武医院放射与核医学科磁共振成像脑信息学北京市重点实验室100053 北京,首都医科大学宣武医院放射与核医学科磁共振成像脑信息学北京市重点实验室100053 北京,首都医科大学宣武医院放射与核医学科磁共振成像脑信息学北京市重点实验室100053 北京,首都医科大学宣武医院放射与核医学科磁共振成像脑信息学北京市重点实验室100053 北京,首都医科大学宣武医院放射与核医学科磁共振成像脑信息学北京市重点实验室100053 北京,首都医科大学宣武医院放射与核医学科磁共振成像脑信息学北京市重点实验室

急性缺血性卒中人工智能CT病灶检测病灶分割综述

Acute ischemic strokeArtificial intelligenceCTLesion identificationLesion segmentationReview

《中国脑血管病杂志》 2026 (8)

560-570,11

首都医科大学宣武医院汇智人才项目(HZ2021ZCLJ005)

10.3969/j.issn.1672-5921.2026.08.007

评论