en
×

分享给微信好友或者朋友圈

使用微信“扫一扫”功能。
通讯作者:

贾中芝,E-mail: jiazhongzhi.1998@163.com

中图分类号:R814.42

文献标识码:A

文章编号:1007-4368(2024)08-1170-05

DOI:10.7655/NYDXBNSN240292

参考文献 1
ISSELBACHER E M,PREVENTZA O,HAMILTON BLACK J,3RD,et al.2022 ACC/AHA Guideline for the diagnosis and management of aortic disease:a report of the American heart Association/American College of Cardiology Joint Committee on clinical practice guidelines[J].Circulation,2022,146(24):e334-e482
参考文献 2
中国医师协会心血管外科分会大血管外科专业委员会.主动脉夹层诊断与治疗规范中国专家共识[J].中华胸心血管外科杂志,2017,33(11):641-654
参考文献 3
侯凡凡.重视造影剂所致急性肾功能衰竭的预防[J].中华内科杂志,2001,40(11):723-724
参考文献 4
MCDONALD J S,MCDONALD R J,WILLIAMSON E E,et al.Is intravenous administration of iodixanol associated with increased risk of acute kidney injury,dialysis,or mortality?A propensity score⁃adjusted study[J].Radiology,2017,285(2):414-424
参考文献 5
SEELIGER E,PERSSON P B.Kidney damage by iodinated contrast media[J].Acta Physiol(Oxf),2019,227(4):e13259
参考文献 6
FÄHLING M,SEELIGER E,PATZAK A,et al.Understanding and preventing contrast⁃induced acute kidney injury[J].Nat Rev Nephrol,2017,13(3):169-180
参考文献 7
范淑玉,张秀美,杨向红.基层医院CT造影增强检查不满意因素调查分析[J].右江民族医学院学报,2006,28(3):416-418
参考文献 8
MATSUO Y,LECUN Y,SAHANI M,et al.Deep learning,reinforcement learning,and world models[J].Neural Netw,2022,152:267-275
参考文献 9
ZHOU S K,GREENSPAN H,DAVATZIKOS C,et al.A review of deep learning in medical imaging:imaging traits,technology trends,case studies with progress highlights,and future promises[J].Proc IEEE Inst Electr Electron Eng,2021,109(5):820-838
参考文献 10
CHANDRASHEKAR A,HANDA A,LAPOLLA P,et al.A deep learning approach to visualize aortic aneurysm morphology without the use of intravenous contrast agents[J].Ann Surg,2023,277(2):e449-e459
参考文献 11
ZHOU M,LUO X,WANG X,et al.Deep learning prediction for distal aortic remodeling after thoracic endovascu-lar aortic repair in stanford type B aortic dissection[J].J Endovasc Ther,2023:15266028231160101
参考文献 12
HOLSTE G,OIKONOMOU E K,MORTAZAVI B J,et al.Severe aortic stenosis detection by deep learning applied to echocardiography[J].Eur Heart J,2023,44(43):4592-4604
参考文献 13
NISHIKIORI H,KURONUMA K,HIROTA K,et al.Deep-learning algorithm to detect fibrosing interstitial lung disease on chest radiographs[J].Eur Respir J,2023,61(2):2102269
参考文献 14
CAO K,XIA Y,YAO J,et al.Large⁃scale pancreatic cancer detection via non ⁃contrast CT and deep learning[J].Nat Med,2023,29(12):3033-3043
参考文献 15
SHKOLYAR E,JIA X,CHANG T C,et al.Augmented bladder tumor detection using deep learning[J].Eur Urol,2019,76(6):714-718
参考文献 16
ISENSEE F,JAEGER P F,KOHL S A A,et al.nnU⁃Net:a self-configuring method for deep learning-based biomedical image segmentation[J].Nat Methods,2021,18(2):203-211
参考文献 17
THEODORIS C V,XIAO L,CHOPRA A,et al.Transfer learning enables predictions in network biology[J].Nature,2023,618(7965):616-624
参考文献 18
DUNDAR A,GAO J,TAO A,et al.Fine detailed texture learning for 3D meshes with generative models[J].IEEE Trans Pattern Anal Mach Intell,2023,45(12):14563-14574
参考文献 19
HATA A,YANAGAWA M,YAMAGATA K,et al.Deep learning algorithm for detection of aortic dissection on non⁃contrast⁃enhanced CT[J].Eur Radiol,2021,31(2):1151-1159
参考文献 20
YI Y,MAO L,WANG C,et al.Advanced warning of aortic dissection on non-contrast CT:the combination of deep learning and morphological characteristics[J].Front Cardiovasc Med,2021,8:762958
参考文献 21
XIONG X,GUAN X,SUN C,et al.A cascaded deep learning framework for detecting aortic dissection using non-contrast enhanced computed tomography[J].Annu Int Conf IEEE Eng Med Biol Soc,2021,2021:2914-2917
参考文献 22
HAHN L D,MISTELBAUER G,HIGASHIGAITO K,et al.CT⁃based true-and false⁃lumen segmentation in type B aortic dissection using machine learning[J].Radiol Cardiothorac Imaging,2020,2(3):e190179
参考文献 23
YU Y,GAO Y,WEI J,et al.A three-dimensional deep convolutional neural network for automatic segmentation and diameter measurement of type B aortic dissection[J].Korean J Radiol,2021,22(2):168-178
参考文献 24
RENGIER F,WÖRZ S,GODINEZ W J,et al.Development of in vivo quantitative geometric mapping of the aortic arch for advanced endovascular aortic repair:feasibility and preliminary results[J].J Vasc Interv Radiol,2011,22(7):980-986
参考文献 25
STERNBERGH W C,MONEY S R,GREENBERG R K,et al.Influence of endograft oversizing on device migra⁃tion,endoleak,aneurysm shrinkage,and aortic neck dilation:results from the Zenith Multicenter Trial[J].J Vasc Surg,2004,39(1):20-26
参考文献 26
SIEREN M M,WIDMANN C,WEISS N,et al.Automated segmentation and quantification of the healthy and diseased aorta in CT angiographies using a dedicated deep learning approach[J].Eur Radiol,2022,32(1):690-701
参考文献 27
TALEBI S,MADANI M H,MADANI A,et al.Machine learning for endoleak detection after endovascular aortic repair[J].Sci Rep,2020,10(1):18343
参考文献 28
CHEN H,LUNDBERG S M,LEE S I.Explaining a series of models by propagating Shapley values[J].Nat Commun,2022,13(1):4512
参考文献 29
KESÄVUORI R,KASEVA T,SALLI E,et al.Deep learning-aided extraction of outer aortic surface from CT angiography scans of patients with Stanford type B aortic dissection[J].Eur Radiol Exp,2023,7(1):35
目录contents

    摘要

    主动脉夹层(aortic dissection,AD)是一种临床上常见的急症,快速、精准的诊治直接关系AD患者的预后,具有一定的挑战性。近年来,基于CT图像的深度学习在辅助AD诊治方面的应用日益增多,尤其在AD的识别、分割以及预测患者预后方面,为AD的临床诊治决策提供了重要价值。为了更好地掌握基于CT图像的深度学习在AD中的应用,本综述将总结与讨论目前深度学习在AD诊治中的应用、存在的问题及潜在的优化策略。

    Abstract

    Aortic dissection(AD)is a critical emergency,where rapid and accurate diagnosis and treatment significantly influence the prognosis of AD patients. However,achieving such precise and swift diagnosis and treatment poses substantial challenges. In recent years,the application of deep learning based on CT images in assisting the diagnosis and treatment of AD has increasingly gained prominence,particularly in the identification,segmentation,and prediction of patient outcomes,thereby offering significant value for clinical decision -making in AD management. To better understand the application of deep learning based on CT images in AD,this review will summarize and discuss the current applications,existing challenges,and potential optimization strategies of deep learning in the diagnosis and treatment of AD.

    关键词

    主动脉夹层深度学习CT

    Keywords

    aortadissectiondeep learningCT

  • 主动脉夹层(aortic dissection,AD)是临床上常见的一种大血管疾病,具有发病急、死亡率高的特点[1-2]。早诊早治是改善AD患者预后的关键。AD 的诊断主要依靠影像学检查,尤其是增强CT,其可以显示AD的关键影像学特征,在AD的诊断中发挥着至关重要的作用。由于增强CT在部分情况下使用受限,如肾功能不全的患者[3-6]、碘造影剂过敏的患者,以及部分落后地区增强CT使用受限[7],因此,这类患者的AD筛查存在一定的挑战性;此外,放射科医生在解读增强CT图像时存在认知差异,因此,在AD的快速精准诊断、特征分析、预后评估等方面也存在一定的差异。基于以上背景,迫切需要寻找更高效和准确的辅助诊断方法。

  • 深度学习(deep learning,DL)是机器学习的一个分支,它通过模拟人脑处理数据的方式来分析复杂影像[8]。DL的出现为基于影像学的AD诊断提供了强大的辅助工具[9]。DL通过高效学习海量影像数据、自动识别图像中的关键特征,不但有助于AD的快速精准诊治[10],还可以预测AD患者的预后[11]

  • 近年来,DL逐渐应用于AD的临床诊治,并且取得了显著进步,文章将详细阐述DL在AD快速精准诊断、分割、预测患者预后中的应用,并探讨当前研究中存在的问题和优化策略。

  • 1 DL概述

  • DL模仿人脑的数据处理方式,专注于从复杂数据中学习和做出决策。在医学领域,DL的应用是基于其强大的识别、分析和解释复杂影像数据的能力。目前有多重 DL 算法,均在医学影像处理中发挥着重要作用,且显著提高了疾病诊断的准确性和效率[12-13]

  • 基于医学影像的 DL 的关键应用包括:①特征提取和识别,卷积神经网络(convolutional neural net⁃ work,CNN)能够有效地从医学影像图像中自动提取和识别关键特征,如肿瘤[14-15]、裂缝或其他异常结构;②图像分类和分割,U⁃Net等算法在图像分割任务中表现出色,特别是在精确分割复杂医学影像的场景中[16],如病理切片的分割。③数据增强和迁移学习,DL通过数据增强和迁移学习能够克服医学影像数据量少的限制,例如,迁移学习中的预训练 CNN减少了对大量标记数据的需求[17];④图像重建和增强,生成对抗网络被用于提高医学影像的质量,如降噪和对比度增强[18];⑤预测性分析,循环神经网络和长期记忆网络在预测疾病进展或患者预后方面发挥着重要作用。

  • 2 DL在AD诊断中的应用

  • 对于增强CT使用受限的患者,DL可以基于平扫 CT 对 AD 做出诊断。基于平扫 CT,DL 可以提取图像中 AD 的大量特征,从而实现影像学诊断。一项回顾性研究纳入了 170 例 AD 患者,基于平扫 CT 开发了一种辅助 DL 诊断工具,该算法的受试者工作特征曲线下面积(area under curve,AUC)为0.940、准确率为0.900、灵敏度为0.918、特异度为0.882,在准确率、灵敏度和特异度方面,该DL算法与放射科医生的平均水平无显著差异[19]。另有研究联合使用三维 DL 模型和高斯朴素贝叶斯算法,基于平扫 CT检测AD,通过对主动脉分割和形态特征的提取,这种集成模型对 AD 的检测表现良好,内部测试组的AUC为0.948,外部测试组AUC为0.969。该研究还邀请了3名放射科医生读片,深度集成模型的灵敏度高于所有放射科医生,在外部验证队列中差异具有统计学意义。然而,深度集成模型的特异性低于3位放射科医生,在外部测试队列中差异具有统计学意义[20]。此外,有研究团队将分割网络与条件生成对抗网络相结合,以达到使用平扫CT图像即可合成高质量增强CT图像从而诊断AD的目的,该模型较单纯生成对抗网络性能更加优越,准确率达到 0.808,灵敏度为 0.960,特异度为 0.668[21]。这些研究共同展示了DL技术在医疗影像分析特别是在AD 诊断方面的巨大潜力和实际应用价值。即使在仅有平扫 CT 图像可用的情况下,DL 也能有效诊断 AD,为无法接受增强CT的患者提供了一种新的、有效的诊断手段。

  • 3 DL在AD的分割与直径测量中的应用

  • DL 在 AD 分割和直径测量中的应用对于指导 AD的手术治疗具有重要意义。首先,DL通过提供更精确和一致的分割结果,帮助临床医生更好地掌握病变的范围和结构,从而做出更合适的手术决策。有研究纳入了45例B型AD患者,基于CT血管造影图像数据,选择主动脉最长路径自动检测主动脉中心线,并根据主动脉中心线生成垂直于中心线的多平面重建(multiplanar reconstruction,MPR)图像,使用CNN模型对MPR图像进行真假腔的分割,结果显示:真腔的Dice相似系数平均值为0.873,假腔为 0.894,证明了该方法在自动分割和测量 B 型 AD真假腔方面的有效性[22]。另一项研究中同样采用了 CNN 来自动分割 B 型 AD,共纳入 139 例 B 型 AD患者,训练后得到的DL算法在主动脉、真腔和假腔的平均 Dice 相似系数分别为 0.958、0.961 和 0.932,显示出较高的准确性。该研究邀请2位放射科医生测量直径,并将2位的测量结果取平均值作为参考标准,测量结果差异较大时,由第3名放射科医生达成共识,研究团队将 DL 分割与手动分割进行比较,结果表明,手动分割和DL分割的测量结果均与参考标准呈线性关系,且在主动脉、真腔和假腔的测量上,对比参考标准DL的平均误差低于手动方法[23]。此外,与手动分割方法相比,DL方法在缩短测量时间及减少观察者之间的差异方面展现了明显优势。因此,DL在图像分割和直径测量的应用展现出显著的潜力和价值。其次,研究表明:人工测量主动脉直径的误差可能大于不同类型支架直径的间隔值(2 mm),这体现了准确的直径测量在选择合适支架方面的重要性[2224-25]。最后,DL技术的应用,尤其是自动化分割技术,显著缩短了术前准备时间,使临床医生能够在紧急情况下更迅速地做出反应。

  • 从单张图像的分割到更复杂的 3D 图像处理, DL 的应用范围正在不断扩大,一些研究探索了3D 图像处理技术在AD分割和测量中的应用。研究人员回顾性收集了191例AD患者的CT血管造影(CT angiography,CTA)图像,并使用先进的3D U⁃Net(一种DL算法,是U⁃Net 的一个变种,能够处理三维数据)手动分割每张图像,这意味着它不仅可以处理单个2D图像,还可以处理一系列连续图像,从而构建出整个主动脉的三维形态。该研究得到的模型的Dice相似系数为0.95,平均表面距离为0.76,显示出不同节段间的高度一致性[26]。这种方法可以更准确地捕获和量化主动脉的复杂结构,特别是存在病变时,3D比2D分割更能精确地反映实际的解剖结构。通过这种方式,能够在三维空间中精确地定位和量化健康及病变的主动脉,从而提供比2D图像分割更丰富的空间信息。

  • 基于以上研究,DL技术通过精确的图像分割和直径测量,为主动脉疾病的手术治疗提供了关键数据支持,使得手术规划更为精确,对指导AD患者的临床治疗决策有着积极意义。

  • 4 DL在检测AD术后内漏发生与预测预后中的应用

  • DL在检测AD腔内修复(endovascular aneurysm repair,EVAR)术后内漏方面也显示出极大的潜力。有研究纳入了50例术后出现内漏的患者和20例术后未出现内漏患者的 CTA 图像,使用 DL 方法开发了一个用于检测内漏的模型,该模型的性能显著优于普通放射科医生,且与放射科专科医生相当[27]。该研究表明:DL可能在临床实践中辅助临床医生解读 EVAR术后的CTA图像,提高对内漏诊断的准确性和效率。

  • 除此之外,有研究回顾性评估了接受EVAR的 147 例 B 型 AD 患者,采用点云神经网络处理AD 患者的3D图像数据,可以更好地捕捉和展示AD的3D 结构的复杂性[11]。通过预测术后远端主动脉的重构情况,临床医生可更好地评估患者的长期风险和治疗效果,从而进行更具针对性的监测和干预。DL 为AD患者EVAR术后并发症的检测及预测预后提供了新的可能性。

  • 这些研究表明DL技术在EVAR 术后内漏检测和评估中的应用不仅能提高诊断的准确性和效率,还为临床医生提供了更多关于AD结构和功能的深入洞察,从而有助于改善患者的长期管理和治疗结果。随着技术的进一步发展和优化,DL有望在心血管疾病的临床诊断和治疗中扮演更加重要的角色。

  • 5 DL在AD中的应用挑战与前景

  • DL在AD中的应用展现出巨大潜力,但同时也面临着一定的挑战。这些挑战不仅涉及技术层面,还包括数据处理和临床实践方面。一方面,DL算法的“黑箱”性质使其在医疗应用中的透明度和可解释性受到限制。这在 AD 的诊断和治疗中尤为关键,因为临床医生需要理解模型是如何进行决策的。为此,研究者们正在探索新的方法来提高 DL 模型的解释能力。目前,研究人员通过计算 SHAP (shapley additive explanations)值[28]、简化算法、建立决策树等方法提高模型的可解释性。另一方面,由于 AD 相关数据往往来源于单一医疗中心,且样本量有限,这限制了模型的泛化能力和准确性。有研究展示了半自动分割方法在处理这一问题时的有效性[29],通过减少人为分割差异和加速分割过程,提高了模型的性能。

  • 对于DL在AD中的未来发展,有几个关键研究方向值得关注。首先,提高算法的可解释性是一个重点。随着医疗决策对精确度和可靠性的要求不断提高,临床医师越来越需要理解模型的决策过程。其次,低辐射检测技术的开发将是未来研究的重要方向。在AD诊断中减少对造影剂的依赖并降低辐射,将有助于降低患者的疾病风险。再者,个性化治疗和风险预测将成为AD治疗和管理的重要研究方向。DL算法可以通过分析大量患者数据,包括遗传信息、生活方式和病史等,来提供更精确的疾病风险评估和治疗效果预测。除此之外,迁移学习也是值得关注的研究方向之一,通过迁移学习不仅可以限制模型过拟合的问题,还可以提升模型在不同人群或不同医疗影像设备产生的数据上的适应性,这种方法提升了 DL 技术更广泛地应用于临床实践的可能性。最后,跨模态数据融合将是未来的一个研究热点。结合不同来源和类型的数据,如影像数据、临床数据和基因数据,将使DL模型能够提供更全面、精确的诊断信息。

  • 总之,尽管 DL 在 AD 的应用中面临着挑战,但随着技术的不断进步和研究的深入,这些问题将逐渐得到解决,DL有助于AD的快速精准诊断、分割、预测预后,具有广阔的应用前景。

  • 参考文献

    • [1] ISSELBACHER E M,PREVENTZA O,HAMILTON BLACK J,3RD,et al.2022 ACC/AHA Guideline for the diagnosis and management of aortic disease:a report of the American heart Association/American College of Cardiology Joint Committee on clinical practice guidelines[J].Circulation,2022,146(24):e334-e482

    • [2] 中国医师协会心血管外科分会大血管外科专业委员会.主动脉夹层诊断与治疗规范中国专家共识[J].中华胸心血管外科杂志,2017,33(11):641-654

    • [3] 侯凡凡.重视造影剂所致急性肾功能衰竭的预防[J].中华内科杂志,2001,40(11):723-724

    • [4] MCDONALD J S,MCDONALD R J,WILLIAMSON E E,et al.Is intravenous administration of iodixanol associated with increased risk of acute kidney injury,dialysis,or mortality?A propensity score⁃adjusted study[J].Radiology,2017,285(2):414-424

    • [5] SEELIGER E,PERSSON P B.Kidney damage by iodinated contrast media[J].Acta Physiol(Oxf),2019,227(4):e13259

    • [6] FÄHLING M,SEELIGER E,PATZAK A,et al.Understanding and preventing contrast⁃induced acute kidney injury[J].Nat Rev Nephrol,2017,13(3):169-180

    • [7] 范淑玉,张秀美,杨向红.基层医院CT造影增强检查不满意因素调查分析[J].右江民族医学院学报,2006,28(3):416-418

    • [8] MATSUO Y,LECUN Y,SAHANI M,et al.Deep learning,reinforcement learning,and world models[J].Neural Netw,2022,152:267-275

    • [9] ZHOU S K,GREENSPAN H,DAVATZIKOS C,et al.A review of deep learning in medical imaging:imaging traits,technology trends,case studies with progress highlights,and future promises[J].Proc IEEE Inst Electr Electron Eng,2021,109(5):820-838

    • [10] CHANDRASHEKAR A,HANDA A,LAPOLLA P,et al.A deep learning approach to visualize aortic aneurysm morphology without the use of intravenous contrast agents[J].Ann Surg,2023,277(2):e449-e459

    • [11] ZHOU M,LUO X,WANG X,et al.Deep learning prediction for distal aortic remodeling after thoracic endovascu-lar aortic repair in stanford type B aortic dissection[J].J Endovasc Ther,2023:15266028231160101

    • [12] HOLSTE G,OIKONOMOU E K,MORTAZAVI B J,et al.Severe aortic stenosis detection by deep learning applied to echocardiography[J].Eur Heart J,2023,44(43):4592-4604

    • [13] NISHIKIORI H,KURONUMA K,HIROTA K,et al.Deep-learning algorithm to detect fibrosing interstitial lung disease on chest radiographs[J].Eur Respir J,2023,61(2):2102269

    • [14] CAO K,XIA Y,YAO J,et al.Large⁃scale pancreatic cancer detection via non ⁃contrast CT and deep learning[J].Nat Med,2023,29(12):3033-3043

    • [15] SHKOLYAR E,JIA X,CHANG T C,et al.Augmented bladder tumor detection using deep learning[J].Eur Urol,2019,76(6):714-718

    • [16] ISENSEE F,JAEGER P F,KOHL S A A,et al.nnU⁃Net:a self-configuring method for deep learning-based biomedical image segmentation[J].Nat Methods,2021,18(2):203-211

    • [17] THEODORIS C V,XIAO L,CHOPRA A,et al.Transfer learning enables predictions in network biology[J].Nature,2023,618(7965):616-624

    • [18] DUNDAR A,GAO J,TAO A,et al.Fine detailed texture learning for 3D meshes with generative models[J].IEEE Trans Pattern Anal Mach Intell,2023,45(12):14563-14574

    • [19] HATA A,YANAGAWA M,YAMAGATA K,et al.Deep learning algorithm for detection of aortic dissection on non⁃contrast⁃enhanced CT[J].Eur Radiol,2021,31(2):1151-1159

    • [20] YI Y,MAO L,WANG C,et al.Advanced warning of aortic dissection on non-contrast CT:the combination of deep learning and morphological characteristics[J].Front Cardiovasc Med,2021,8:762958

    • [21] XIONG X,GUAN X,SUN C,et al.A cascaded deep learning framework for detecting aortic dissection using non-contrast enhanced computed tomography[J].Annu Int Conf IEEE Eng Med Biol Soc,2021,2021:2914-2917

    • [22] HAHN L D,MISTELBAUER G,HIGASHIGAITO K,et al.CT⁃based true-and false⁃lumen segmentation in type B aortic dissection using machine learning[J].Radiol Cardiothorac Imaging,2020,2(3):e190179

    • [23] YU Y,GAO Y,WEI J,et al.A three-dimensional deep convolutional neural network for automatic segmentation and diameter measurement of type B aortic dissection[J].Korean J Radiol,2021,22(2):168-178

    • [24] RENGIER F,WÖRZ S,GODINEZ W J,et al.Development of in vivo quantitative geometric mapping of the aortic arch for advanced endovascular aortic repair:feasibility and preliminary results[J].J Vasc Interv Radiol,2011,22(7):980-986

    • [25] STERNBERGH W C,MONEY S R,GREENBERG R K,et al.Influence of endograft oversizing on device migra⁃tion,endoleak,aneurysm shrinkage,and aortic neck dilation:results from the Zenith Multicenter Trial[J].J Vasc Surg,2004,39(1):20-26

    • [26] SIEREN M M,WIDMANN C,WEISS N,et al.Automated segmentation and quantification of the healthy and diseased aorta in CT angiographies using a dedicated deep learning approach[J].Eur Radiol,2022,32(1):690-701

    • [27] TALEBI S,MADANI M H,MADANI A,et al.Machine learning for endoleak detection after endovascular aortic repair[J].Sci Rep,2020,10(1):18343

    • [28] CHEN H,LUNDBERG S M,LEE S I.Explaining a series of models by propagating Shapley values[J].Nat Commun,2022,13(1):4512

    • [29] KESÄVUORI R,KASEVA T,SALLI E,et al.Deep learning-aided extraction of outer aortic surface from CT angiography scans of patients with Stanford type B aortic dissection[J].Eur Radiol Exp,2023,7(1):35

  • 参考文献

    • [1] ISSELBACHER E M,PREVENTZA O,HAMILTON BLACK J,3RD,et al.2022 ACC/AHA Guideline for the diagnosis and management of aortic disease:a report of the American heart Association/American College of Cardiology Joint Committee on clinical practice guidelines[J].Circulation,2022,146(24):e334-e482

    • [2] 中国医师协会心血管外科分会大血管外科专业委员会.主动脉夹层诊断与治疗规范中国专家共识[J].中华胸心血管外科杂志,2017,33(11):641-654

    • [3] 侯凡凡.重视造影剂所致急性肾功能衰竭的预防[J].中华内科杂志,2001,40(11):723-724

    • [4] MCDONALD J S,MCDONALD R J,WILLIAMSON E E,et al.Is intravenous administration of iodixanol associated with increased risk of acute kidney injury,dialysis,or mortality?A propensity score⁃adjusted study[J].Radiology,2017,285(2):414-424

    • [5] SEELIGER E,PERSSON P B.Kidney damage by iodinated contrast media[J].Acta Physiol(Oxf),2019,227(4):e13259

    • [6] FÄHLING M,SEELIGER E,PATZAK A,et al.Understanding and preventing contrast⁃induced acute kidney injury[J].Nat Rev Nephrol,2017,13(3):169-180

    • [7] 范淑玉,张秀美,杨向红.基层医院CT造影增强检查不满意因素调查分析[J].右江民族医学院学报,2006,28(3):416-418

    • [8] MATSUO Y,LECUN Y,SAHANI M,et al.Deep learning,reinforcement learning,and world models[J].Neural Netw,2022,152:267-275

    • [9] ZHOU S K,GREENSPAN H,DAVATZIKOS C,et al.A review of deep learning in medical imaging:imaging traits,technology trends,case studies with progress highlights,and future promises[J].Proc IEEE Inst Electr Electron Eng,2021,109(5):820-838

    • [10] CHANDRASHEKAR A,HANDA A,LAPOLLA P,et al.A deep learning approach to visualize aortic aneurysm morphology without the use of intravenous contrast agents[J].Ann Surg,2023,277(2):e449-e459

    • [11] ZHOU M,LUO X,WANG X,et al.Deep learning prediction for distal aortic remodeling after thoracic endovascu-lar aortic repair in stanford type B aortic dissection[J].J Endovasc Ther,2023:15266028231160101

    • [12] HOLSTE G,OIKONOMOU E K,MORTAZAVI B J,et al.Severe aortic stenosis detection by deep learning applied to echocardiography[J].Eur Heart J,2023,44(43):4592-4604

    • [13] NISHIKIORI H,KURONUMA K,HIROTA K,et al.Deep-learning algorithm to detect fibrosing interstitial lung disease on chest radiographs[J].Eur Respir J,2023,61(2):2102269

    • [14] CAO K,XIA Y,YAO J,et al.Large⁃scale pancreatic cancer detection via non ⁃contrast CT and deep learning[J].Nat Med,2023,29(12):3033-3043

    • [15] SHKOLYAR E,JIA X,CHANG T C,et al.Augmented bladder tumor detection using deep learning[J].Eur Urol,2019,76(6):714-718

    • [16] ISENSEE F,JAEGER P F,KOHL S A A,et al.nnU⁃Net:a self-configuring method for deep learning-based biomedical image segmentation[J].Nat Methods,2021,18(2):203-211

    • [17] THEODORIS C V,XIAO L,CHOPRA A,et al.Transfer learning enables predictions in network biology[J].Nature,2023,618(7965):616-624

    • [18] DUNDAR A,GAO J,TAO A,et al.Fine detailed texture learning for 3D meshes with generative models[J].IEEE Trans Pattern Anal Mach Intell,2023,45(12):14563-14574

    • [19] HATA A,YANAGAWA M,YAMAGATA K,et al.Deep learning algorithm for detection of aortic dissection on non⁃contrast⁃enhanced CT[J].Eur Radiol,2021,31(2):1151-1159

    • [20] YI Y,MAO L,WANG C,et al.Advanced warning of aortic dissection on non-contrast CT:the combination of deep learning and morphological characteristics[J].Front Cardiovasc Med,2021,8:762958

    • [21] XIONG X,GUAN X,SUN C,et al.A cascaded deep learning framework for detecting aortic dissection using non-contrast enhanced computed tomography[J].Annu Int Conf IEEE Eng Med Biol Soc,2021,2021:2914-2917

    • [22] HAHN L D,MISTELBAUER G,HIGASHIGAITO K,et al.CT⁃based true-and false⁃lumen segmentation in type B aortic dissection using machine learning[J].Radiol Cardiothorac Imaging,2020,2(3):e190179

    • [23] YU Y,GAO Y,WEI J,et al.A three-dimensional deep convolutional neural network for automatic segmentation and diameter measurement of type B aortic dissection[J].Korean J Radiol,2021,22(2):168-178

    • [24] RENGIER F,WÖRZ S,GODINEZ W J,et al.Development of in vivo quantitative geometric mapping of the aortic arch for advanced endovascular aortic repair:feasibility and preliminary results[J].J Vasc Interv Radiol,2011,22(7):980-986

    • [25] STERNBERGH W C,MONEY S R,GREENBERG R K,et al.Influence of endograft oversizing on device migra⁃tion,endoleak,aneurysm shrinkage,and aortic neck dilation:results from the Zenith Multicenter Trial[J].J Vasc Surg,2004,39(1):20-26

    • [26] SIEREN M M,WIDMANN C,WEISS N,et al.Automated segmentation and quantification of the healthy and diseased aorta in CT angiographies using a dedicated deep learning approach[J].Eur Radiol,2022,32(1):690-701

    • [27] TALEBI S,MADANI M H,MADANI A,et al.Machine learning for endoleak detection after endovascular aortic repair[J].Sci Rep,2020,10(1):18343

    • [28] CHEN H,LUNDBERG S M,LEE S I.Explaining a series of models by propagating Shapley values[J].Nat Commun,2022,13(1):4512

    • [29] KESÄVUORI R,KASEVA T,SALLI E,et al.Deep learning-aided extraction of outer aortic surface from CT angiography scans of patients with Stanford type B aortic dissection[J].Eur Radiol Exp,2023,7(1):35