文献速递神经外周血管CT

自动化CTA衍生侧支循环分级与形态学指标用于增强卒中后结局预测

Automated CTA-Derived Collateral Grading and Morphologic Metrics for Enhanced Prediction of Poststroke Outcomes.
PubMed 文献 · AJNR Am J Neuroradiol · 2026-09-02

研究基于230例2019至2023年接受血管内取栓的急性缺血性卒中患者,用深度学习U-Net从CTA自动提取定量侧支指数(qCI)及血管形态指标。qCI与专家评分一致性高(准确率0.863,κ=0.786),CTA-only模型预测90天mRS优于CTP-only模型(AUROC 0.730 vs 0.645),联合模型最佳(AUROC 0.781)。

要点

  • 自动qCI与专家侧支评分高度一致(准确率0.863,κ=0.786)。
  • CTA-only模型预测90天mRS优于CTP-only(AUROC 0.730 vs 0.645)。
  • CTA+CTP联合模型预测效能最佳(AUROC 0.781)。

核心结论

  • 自动CTA衍生qCI及脑血管形态学可快速、客观、可重复地评估侧支循环,与专家分级一致,在多项预测任务中优于灌注指标,与CTP联合可进一步提升预后准确性。

研究设计与样本

  • 回顾性分析2019至2023年前瞻性收集的230例接受血管内取栓的急性缺血性卒中患者数据,使用经验证的神经网络血管提取模型分割CTA生成3D血管网络并计算形态指标,通过半球比较导出全自动qCI。

诊断标准与关键数值

  • 自动qCI(0-3级)与专家评分一致性:准确率0.863,Pearson R=0.880,Cohen κ=0.786。
  • 二分类侧支状态准确率0.938,AUROC=0.945。
  • 90天mRS预测:CTA-only模型AUROC 0.730 vs CTP-only模型AUROC 0.645,Brier评分0.178 vs 0.295。
  • 联合CTA+CTP模型AUROC 0.781。

临床应用与阅片要点

  • qCI可作为一种快速、客观、可重复的侧支循环评估工具,与专家评分高度一致。
  • 在CTP不可用的机构,CTA衍生特征可替代灌注指标用于预后预测。
  • 将CTA形态学指标加入灌注模型可显著改善风险重分类。

局限性

  • 原文未报告局限性。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

英文原文(摘要)· 点击展开

BACKGROUND AND PURPOSE: Collateral circulation is a key determinant of treatment response and outcomes in acute ischemic stroke (AIS), yet its assessment in clinical practice remains limited and subjective. While CTP offers insight into tissue viability, its restricted availability and susceptibility to artifacts reduce its practical utility, particularly in smaller centers. As an accessible alternative, we developed and validated an automated quantitative collateral index (qCI) derived from CTA using a deep learning U-Net segmentation framework and evaluated the ability of CTA-based features to predict poststroke recovery and functional outcomes.

MATERIALS AND METHODS: We retrospectively analyzed prospectively collected data from 230 patients with AIS who underwent endovascular thrombectomy (EVT) between 2019 and 2023. CTA scans were segmented using a validated neural network-based vascular extraction model to generate 3D vessel networks and compute morphology metrics (vessel length, branching, fractal dimension, tortuosity). A fully automated qCI was derived through hemispheric comparison of vascular features following spatial registration. Agreement of qCI with clinician grading was quantified. Gradient boosted decision tree models were trained to predict early neurologic deterioration, early neurologic improvement, and 90-day mRS using the following: 1) CTP-only (core, penumbra, mismatch), 2) CTA-only (qCI + morphology), and 3) combined CTA + CTP features.

RESULTS: Automated qCI (grades 0-3) showed strong concordance with expert scoring (accuracy 0.863; Pearson R = 0.880; Cohen κ = 0.786). Dichotomized collateral status achieved 0.938 accuracy (area under the receiver operating curve [AUROC] = 0.945). For 90-day mRS prediction, the CTA-only model outperformed the CTP-only model (AUROC, 0.730 versus 0.645) with better calibration (Brier score 0.178 versus 0.295). The combined CTA + CTP model performed best overall (AUROC 0.781), with similar improvements observed for early neurologic deterioration. CTA-derived features led to significant reclassification gains when added to perfusion-based models.

CONCLUSIONS: Automated CTA-derived qCI and cerebrovascular morphology provide rapid, objective, and reproducible collateral assessment with agreement with expert grading. These features outperform perfusion metrics in several predictive tasks and further enhance prognostic accuracy when combined with CTP. Because CTA is widely available, qCI offers a scalable, clinically practical tool for improving prediction of stroke outcome, particularly in settings where CTP is unavailable.

来源:AJNR Am J Neuroradiol,查看原文。版权归原作者所有。

引用

  1. Deshpande A, Wang J, Bochenek KM, et al. Automated CTA-Derived Collateral Grading and Morphologic Metrics for Enhanced Prediction of Poststroke Outcomes. AJNR Am J Neuroradiol. 2026. PMID: 41826063. doi:10.3174/ajnr.A9289

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。