Functional Lung Imaging Using a Single kV CT Acquisition

摘要 静脉血栓栓塞是一个主要的全球健康和经济负担，约有1000万例发生 45岁以后的终生风险为8%。肺栓塞（PE）是一种静脉血栓。 血栓栓塞事件与高发病率和死亡率相关， 诊断或不久后。最近，COVID-19大流行导致 急性肺血栓栓塞性疾病的患者，最有可能是在感染性 累及内皮的血管炎引起局部动脉血栓形成和随后的肺梗死， 血液处于高凝状态，促进凝块形成。在这些患者中， 认识到肺灌注异常与肺实变和毛玻璃样改变有关， 阴影是预后不良的重要预测因素。目前，肺CT血管造影（CTA）已成为 诊断PE和平面肺通气/灌注（V/Q）肺动脉造影的首选方法用于以下情况： 当禁忌使用肺CTA时。一个迫切的未满足的临床需求是开发一种方法， 同时进行肺CTA和实质灌注评估，无需使用两种模式（如CTA） 和SPECT灌注。在这个项目中，一种基于成像物理学的深度学习方法将 被开发以提取先前被忽视的固有地编码在所获取的光谱中的光谱信息。 对比增强CT投影数据。由于这一突破，这种新的光谱CT成像方法， 将开发并验证用于肺灌注缺陷定量的Deep-En-Chroma 目前可用的肺CTA的实质。这将在不需要任何 昂贵的双能CT（DECT）硬件升级，已被主要CT商业化 厂商总之，在本项目完成后，新的功能性CT成像方法将具有 除了提供目前可用的肺部CTA图像外，还将检测 肺实质中的灌注缺陷，而不需要高端DECT硬件。

ABSTRACT Venous thromboembolism is a major global health and economic burden with about 10 million cases occurring every year, and a high lifetime risk of 8% after age 45 years. Pulmonary embolism (PE) is a venous thromboembolic event associated with high morbidity and mortality, with about 20% incidence of death before diagnosis or shortly thereafter. Most recently, the COVID-19 pandemic has contributed to a marked increase in patients presenting with acute pulmonary thromboembolic disease, most likely created when the infectious vasculitis involving the endothelium creates local arterial thrombosis and subsequent lung infarction, with a superimpose hypercoagulable state that promotes clot formation. In these patients, it is increasingly being recognized that pulmonary perfusion abnormalities associated with the lung consolidations and ground-glass opacities are important predictors of poor prognosis. Currently, pulmonary CT angiography (CTA) has become the preferred method for diagnosing PE and planar lung ventilation/perfusion (V/Q) scintigraphy is used in cases when pulmonary CTA is contraindicated. A compelling unmet clinical need is to develop a method for simultaneous pulmonary CTA and parenchymal perfusion assessment without the use of two modalities like CTA and SPECT perfusion in the same patient. In this project, an imaging physics-based deep learning method will be developed to extract the previously overlooked spectral information inherently encoded in the acquired contrast enhanced CT projection data. As a result of this breakthrough, this new spectral CT imaging method, referred to as Deep-En-Chroma, will be developed and validated for perfusion defect quantification in lung parenchyma from the currently available pulmonary CTA. This will be accomplished without the need for any expensive dual energy CT (DECT) hardware upgrades that have been commercialized by major CT manufacturers. In summary, upon the completion of this project, a new functional CT imaging method will have been developed, that in addition to providing the currently available pulmonary CTA images, will also detect perfusion defects in lung parenchyma without the requirement of high-end DECT hardware.