A Deep Learning Approach to Visualize Aortic Aneurysm Morphology Without the Use of Intravenous Contrast Agents.
A Deep Learning Approach to Visualize Aortic Aneurysm Morphology Without the Use of Intravenous Contrast Agents.
复制标题
DOI：
10.1097/sla.0000000000004835
复制
发表时间：
2023-02-01
影响因子：
9
通讯作者：
Lee R
中科院分区：
文献类型：
--
作者：
Chandrashekar A;Handa A;Lapolla P;Shivakumar N;Uberoi R;Grau V;Lee R
静脉内造影剂通常用于CT成像，以使血管内病变可视化，例如腹主动脉瘤。然而，碘过敏患者禁用该注射剂，且与肾脏并发症相关。在这项研究中，我们调查，如果从一个noncontrast CT图像采集的原始数据包含足够的信息来区分血液和其他软组织成分。开发了由生成对抗网络支持的深度学习管道，以使用非造影CT模拟造影增强CTA图像。两个生成模型（循环和条件）采用来自75例腹主动脉瘤患者（共11，243对图像）的成对非造影和造影增强CT进行训练，采用3重交叉验证方法，训练/测试分为50：25例患者。随后，在200名患者的独立验证队列（总共29，468对图像）上评估模型。两种深度学习生成模型都能够执行此图像转换任务，其中循环生成对抗网络（GAN）模型的性能优于条件GAN模型，如动脉瘤管腔分割准确性所测量的（Cycle-GAN：86.1% ± 12.2% vs Con-GAN：85.7% ± 10.4%）和血栓空间形态分类准确性（Cycle-GAN：93.5% vs Con-GAN：85.7%）。该管道实现了深度学习方法，可以从非造影图像生成CTA，而无需注射造影剂，与地面实况具有很强的一致性，并能够评估重要的临床指标。我们的管道准备破坏需要静脉造影剂的临床途径。
Intravenous contrast agents are routinely used in CT imaging to enable the visualization of intravascular pathology, such as with abdominal aortic aneurysms. However, the injection is contraindicated in patients with iodine allergy and is associated with renal complications. In this study, we investigate if the raw data acquired from a noncontrast CT image contains sufficient information to differentiate blood and other soft tissue components. A deep learning pipeline underpinned by generative adversarial networks was developed to simulate contrast enhanced CTA images using noncontrast CTs. Two generative models (cycle- and conditional) are trained with paired noncontrast and contrast enhanced CTs from seventy-five patients (total of 11,243 pairs of images) with abdominal aortic aneurysms in a 3-fold cross-validation approach with a training/testing split of 50:25 patients. Subsequently, models were evaluated on an independent validation cohort of 200 patients (total of 29,468 pairs of images). Both deep learning generative models are able to perform this image transformation task with the Cycle-generative adversarial network (GAN) model outperforming the Conditional-GAN model as measured by aneurysm lumen segmentation accuracy (Cycle-GAN: 86.1% ± 12.2% vs Con-GAN: 85.7% ± 10.4%) and thrombus spatial morphology classification accuracy (Cycle-GAN: 93.5% vs Con-GAN: 85.7%). This pipeline implements deep learning methods to generate CTAs from noncontrast images, without the need of contrast injection, that bear strong concordance to the ground truth and enable the assessment of important clinical metrics. Our pipeline is poised to disrupt clinical pathways requiring intravenous contrast.