文献速递神经外周血管CT

深度学习重建的时间-最大强度投影与迭代重建在急性前循环缺血性卒中侧支循环评估中的对比

Deep learning-reconstructed time-maximum intensity projection versus iterative reconstruction for collateral assessment in acute anterior circulation ischemic stroke.
PubMed 文献 · Eur J Radiol · 2025-11-17

回顾性分析75例单侧前循环大血管闭塞急性缺血性卒中患者的一站式CTA-CTP扫描数据,比较深度学习重建的时间-最大强度投影(DLR t-MIP)与单期CTA、多期CTA及CTP容积参数在侧支循环评估和90天预后预测中的表现。DLR t-MIP图像质量主观评分显著高于SP-CTA组,客观噪声显著更低、血管CT值、SNR和CNR显著更高;侧支评分与最终梗死体积呈负相关,DLR t-MIP的AUC为0.84,优于SP-CTA(ΔAUC +0.113,p=0.012),与MP-CTA相当(ΔAUC +0.002,p=0.957)。

要点

  • DLR t-MIP将CTP的时序信息整合为单一高质量血管成像数据集,无需额外采集即可获得优于SP-CTA的预后判别能力。
  • DLR t-MIP侧支评分与最终梗死体积呈负相关(r=-0.73),相关性强于SP-CTA(r=-0.64),与MP-CTA(r=-0.72)相近。
  • DLR t-MIP预测90天预后的AUC为0.84,与MP-CTA(0.84)和CTP低灌注体积(0.85)相当,优于SP-CTA(0.73)。

核心结论

  • DLR t-MIP将CTP的时序信息整合为单一高质量血管成像数据集,图像质量优于SP-CTA,预后判别能力优于SP-CTA、与MP-CTA及CTP容积参数相当,且无需额外采集;该方法可能简化急性卒中分诊并辅助侧支循环指导的决策。

研究设计与样本

  • 回顾性研究,纳入75例单侧前循环大血管闭塞的急性缺血性卒中患者,均接受一站式CT血管成像和CT灌注(CTA-CTP)扫描;原始数据分别用迭代重建(IR)算法和深度学习算法重建,DLR t-MIP由动态脑扫描数据生成,轴位MIP层厚为24/4 mm;以90天改良Rankin量表(mRS)定义预后(良好≤2分 vs 不良≥3分)。

关键影像表现/征象

  • DLR t-MIP组的主观血管成像质量评分显著高于SP-CTA组(p<0.05),观察者间一致性高(kappa=0.716)。
  • DLR t-MIP的图像噪声显著低于SP-CTA(p<0.001),血管CT衰减值、SNR和CNR均显著高于SP-CTA(均p<0.001)。
  • 侧支评分与最终梗死体积(FIV)呈负相关:DLR t-MIP r=-0.73(95%CI -0.80至-0.51),SP-CTA r=-0.64(95%CI -0.78至-0.45),MP-CTA r=-0.72(95%CI -0.83至-0.57),均p<0.001。

诊断标准与关键数值

  • 侧支循环分别采用Menon单期标准(SP-CTA、DLR t-MIP)或多期标准(MP-CTA)进行分级。
  • 90天预后预测的AUC:DLR t-MIP 0.84(95%CI 0.74-0.92),SP-CTA 0.73(95%CI 0.62-0.82),MP-CTA 0.84(95%CI 0.71-0.91),缺血核心体积0.83(95%CI 0.73-0.92),低灌注体积0.85(95%CI 0.75-0.94)。
  • DeLong检验显示DLR t-MIP预测90天预后的AUC大于SP-CTA(ΔAUC +0.113;p=0.012),与MP-CTA相当(ΔAUC +0.002;p=0.957)。

临床应用与阅片要点

  • DLR t-MIP可从CTP动态数据生成,无需额外采集即可同时获得高质量血管成像和侧支循环信息。
  • DLR t-MIP的侧支评分与最终梗死体积的相关性(r=-0.73)强于SP-CTA(r=-0.64),与MP-CTA(r=-0.72)相近。
  • DLR t-MIP对90天预后的判别能力优于SP-CTA(AUC 0.84 vs 0.73),与MP-CTA(0.84)及CTP低灌注体积(0.85)相当。

局限性

  • 本文为回顾性研究。

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

PURPOSE: To evaluate the deep learning-reconstructed time-maximum intensity projection (DLR t-MIP) in acute patients with anterior circulation ischemic stroke, and to compare its ability to assess collateral circulation and predict 90-day outcomes with that of single-phase CTA (SP-CTA), multiphase CTA (MP-CTA), and CT perfusion (CTP) volumetrics.

METHODS: This retrospective study analyzed data from 75 AIS patients with unilateral anterior circulation large-vessel occlusion who underwent one-stop CT angiography and CT perfusion (CTA-CTP) scanning. The raw data were reconstructed using iterative reconstruction (IR) algorithms and deep learning algorithms. DLR t-MIPs were generated from dynamic brain scan data; axial MIPs were 24/4 mm. Subjective vessel delineation and noise (5-point scales) and objective attenuation, noise, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were compared. Collaterals were graded by Menon single-phase (SP-CTA, DLR t-MIP) or multiphase criteria (MP-CTA). Ninety-day mRS defined outcomes (good ≤ 2 vs poor ≥ 3). Prognostic performance was evaluated with a ROC curve and pairwise DeLong tests.

RESULTS: Subjective quality scores revealed that the vascular imaging quality was significantly greater in the DLR t-MIP group (p < 0.05), with high interobserver agreement (kappa = 0.716). Objectively, DLR t-MIP demonstrated a significantly lower level of image noise (p < 0.001) and significantly higher vascular CT attenuation values, SNRs, and CNRs (all p < 0.001) than SP-CTA. Collateral scores correlated inversely with the final infarct volume (FIV): DLR t-MIP r = - 0.73 (95 % CI, -0.80 to -0.51), SP-CTA r = - 0.64 (-0.78 to -0.45), and MP-CTA r = - 0.72 (-0.83 to - 0.57); all p < 0.001. The AUCs for 90-day outcomes were as follow: DLR t-MIP 0.84 (95 % CI 0.74-0.92), SP-CTA 0.73 (95 % CI 0.62-0.82), MP-CTA 0.84 (95 % CI 0.71-0.91), ischemic core volume 0.83 (95 % CI 0.73-0.92), and hypoperfusion volume 0.85 (0.75-0.94). DeLong testing showed that DLR t-MIP > SP-CTA (ΔAUC + 0.113; p = 0.012) and ≈ MP-CTA (ΔAUC + 0.002; p = 0.957).

CONCLUSIONS: DLR t-MIP consolidates temporal information from CTP into a single high-quality angiographic dataset, yielding superior image quality and better prognostic discrimination than SP-CTA, and comparable performance to MP-CTA and CTP volumetrics-without additional acquisition. This approach may streamline acute-stroke triage and collateral-guided decision-making.

来源:Eur J Radiol,查看原文。版权归原作者所有。

引用

  1. Zheng QF, Sun CF, Zhang YJ, et al. Deep learning-reconstructed time-maximum intensity projection versus iterative reconstruction for collateral assessment in acute anterior circulation ischemic stroke. Eur J Radiol. 2025. PMID: 41264986. doi:10.1016/j.ejrad.2025.112524

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。