Deviceless and Autonomous Prospective Cardiac CT Triggering

项目总结/摘要 冠心病（CHD）是世界范围内的主要死亡原因。估计有380万男性和340万女性 每年有100万妇女死于冠心病。心脏CT是一种安全、准确、无创的成像方式，用于 诊断CHD和计划治疗干预。心脏CT检查仍然具有挑战性 由于跳动的心脏以及需要基于心动时相和基于何时 达到峰值碘对比增强。整个考试持续时间和执行这些考试的复杂性 检查（与有限的报销水平相比）限制了患者获得心脏CT的机会， 医院和专门的心脏成像中心。与其他CT检查相比，心脏CT检查需要 额外的患者准备时间，额外的CT扫描以跟踪团注，以及额外的造影剂以避免 错过了峰值增强。 该项目的目标是开发一种智能心脏CT扫描仪， 扫描时间间隔，无ECG、传统推注跟踪或定时推注。初步结果显示， 为了在诊断CT扫描之前从几个CT投影测量中提取心脏门控信息， 没有重建。这是通过快速X射线管脉冲和深脉冲的创新组合实现的。 学习原始数据分析。该项目建立在GE Research在心脏CT技术方面的经验基础上， 学习算法和X射线管物理学，以及大学强大的临床心脏CT专业知识 位于加州圣地亚哥。 该项目的成果将是自主触发方法的临床可行性研究， 有可能简化和增加患者对心脏CT的访问，同时减少检查时间，减少并发症， 可靠的造影剂体积，并确保稳健的图像质量。

PROJECT SUMMARY/ABSTRACT Coronary heart disease (CHD) is the leading cause of death worldwide. An estimated 3.8 million men and 3.4 million women die each year from CHD. Cardiac CT is a safe, accurate, non-invasive imaging modality used for diagnosing CHD and for planning therapeutic interventions. Cardiac CT exams are still challenging to perform due to the beating heart and the need to carefully time the scan based on cardiac phase and based on when the peak iodine contrast enhancement is reached. The overall exam duration and the complexity of performing these exams (contrasted with limited reimbursement levels) have limited patient access to cardiac CT to academic hospitals and specialized cardiac imaging centers. As compared to other CT exams, cardiac CT exams require additional patient preparation time, additional CT scans to track the bolus, and additional contrast agent to avoid missing the peak enhancement. The goal of this project is to develop a smart cardiac CT scanner that autonomously determines the optimal scan time interval without ECG, traditional bolus tracking or timing bolus. Initial results show that it is possible to extract cardiac gating information from a few CT projection measurements prior to the diagnostic CT scan, without reconstruction. This is made possible by an innovative combination of fast X-ray tube pulsing and deep learning raw data analysis. This project builds on GE Research's experience with cardiac CT technologies, deep learning algorithms and X-ray tube physics, as well as the strong clinical cardiac CT expertise at the University of California San Diego. The outcome of this project will be a clinical feasibility study of the autonomous triggering approach, which has the potential to simplify and increase patient access to cardiac CT, while reducing exam time, reducing con- trast agent volume, and ensuring robust image quality.