Enhanced x-ray angiography analysis and interpretation using deep learning

使用深度学习增强 X 射线血管造影分析和解释

基本信息

批准号：
10000961
负责人：
Ricardo Henao Giraldo
金额：
$ 70.25万
依托单位：
VIGILANT MEDICAL, INC.
依托单位国家：
美国
项目类别：
财政年份：
2018
资助国家：
美国
起止时间：
2018-09-07 至 2023-07-30
项目状态：
已结题

来源：
https://reporter.nih.gov/project-details/10000961
关键词：
3-Dimensional Address Anatomy Angiography Anterior Architecture Area Cephalic Characteristics Chest Pain Classification Clinical Computer Vision Systems Computers Coronary Coronary Arteriosclerosis Coronary Stenosis Coronary Vessels Cost of Illness Data Data Set Development Diagnosis Diagnostic radiologic examination Engineering Evaluation Evaluation Studies Goals Gold Healthcare Systems Heart Image Institutional Review Boards Intervention Intraobserver Variability Lateral Lead Learning Left Lesion Methodology Modeling Morbidity - disease rate National Heart, Lung, and Blood Institute Network-based Neural Network Simulation Observational Study Outcome Pathway Analysis Patient-Focused Outcomes Patients Performance Phase Physicians Procedures Process Reader Reading Reporting Research Roentgen Rays Self-Help Devices Severities Specificity Stenosis Structure Systems Analysis Technology Testing Translating Trees Visual Visual Aid Work base clinically relevant cohort computer aided detection convolutional neural network coronary lesion cost deep learning deep neural network diagnostic accuracy group intervention image processing imaging study improved innovation novel prospective recursive neural network spatiotemporal standard of care tool treatment planning usability

项目摘要

Enhanced x-ray angiography analysis and interpretation using deep learning Over 1 Million diagnostic X-ray angiograms are performed annually in the US to guide treatment of coronary artery disease (CAD) and cost over $12 billion. Despite being the clinical standard of care, visual interpretation is prone to inter- and intra-observer variability. Recently as part of the NHLBI supported Prospective Multicenter Imaging Study for Evaluation of Chest Pain (PROMISE) trial, our research team showed that cardiologists misinterpreted over 19% of angiograms obstructive CAD (greater than 50% vessel stenosis). Given the centrality of angiographic interpretation to the development of a treatment plan, reduced accuracy can lead to unnecessary poor outcomes and increased costs to our healthcare system. The potential impact is significant given that increasing interpretation accuracy by 1% could positively benefit over 10,000 patients each year in the US alone. Thus, our team is developing an X-ray angiographic analysis system (DeepAngio) driven by deep learning technology to enhance physician interpretation. In Phase I, the PROMISE dataset of over 1,000 angiograms was used to build our Convolutional Neural Network (CNN) based deep learning model. We achieved a 0.89 Area Under the Receiving Operating Characteristic (AUROC) for identifying obstructive CAD in images with expert scored ground truth (exceeding our proposed Phase I milestone of >0.85 AUROC). Now in Phase II, we present an innovative image learning pipeline to incorporate anatomical and spatiotemporal information from video sequences (similar to a cardiologist reader). A full end to end X-ray angiography video processing pipeline will be developed and tested in a new cohort of 10,000 patient angiograms with normal and graded abnormal CAD. Our patch-based frame analysis model will advance to CNN full frame-based classification of angiographic views (left heart vs. right heart) and segmentation of coronary vessels (LAD, LCx, and RCA). A multiple frame analysis approach enabled by a Recursive Neural Network (RNN) will equip our model with dynamic temporal information to estimate lesion presence accurately. Our goal for Phase II is to improve reading specificity and translate our Phase I proof of concept research findings into a clinically meaningful tool. A multi-reader, multi-case evaluation by a group of interventional cardiologists interpreting with and without DeepAngio predictions will assess clinical usability to improve coronary stenosis estimation. In the long term, we hope the combination of a cardiologist with DeepAngio as an assistive tool will improve the clinical accuracy of angiographic interpretation.

增强的 X 射线血管造影分析和使用深度学习进行解释美国每年进行超过 100 万次诊断性 X 射线血管造影，以指导冠状动脉的治疗动脉疾病 (CAD) 的费用超过 120 亿美元。尽管视觉解读是临床护理标准容易出现观察者之间和观察者内部的变异性。最近作为 NHLBI 支持的前瞻性多中心的一部分评估胸痛的影像学研究 (PROMISE) 试验，我们的研究团队表明，心脏病专家超过 19% 的血管造影被误解为阻塞性 CAD（血管狭窄超过 50%）。鉴于中心地位血管造影解释到治疗计划制定的过程中，准确性降低可能会导致不必要的结果不佳，我们的医疗保健系统成本增加。考虑到这一点，潜在的影响是巨大的仅在美国，将判读准确性提高 1% 每年就能使超过 10,000 名患者受益。因此，我们的团队正在开发由深度学习驱动的X射线血管造影分析系统（DeepAngio）技术来增强医生的解释。在第一阶段，PROMISE 数据集包含超过 1,000 个血管造影照片用于构建基于卷积神经网络 (CNN) 的深度学习模型。我们实现了 0.89 面积根据接收操作特性 (AUROC)，与专家一起识别图像中的阻塞性 CAD 获得了地面真实值（超过了我们提议的 >0.85 AUROC 的第一阶段里程碑）。现在在第二阶段，我们提出了一种创新的图像学习管道，将解剖学和时空学结合起来来自视频序列的信息（类似于心脏病专家阅读器）。完整的端到端 X 射线血管造影视频处理流程将在由 10,000 名正常和不正常患者的血管造影组成的新队列中进行开发和测试分级异常 CAD。我们基于补丁的帧分析模型将发展为基于 CNN 全帧的模型血管造影视图的分类（左心与右心）和冠状血管分割（LAD、LCx、和RCA）。由递归神经网络 (RNN) 支持的多帧分析方法将装备我们具有动态时间信息的模型可以准确估计病变的存在。我们第二阶段的目标是提高阅读特异性并将我们的第一阶段概念验证研究结果转化为临床有意义的工具。由一组介入心脏病专家进行多读者、多病例评估如果没有 DeepAngio，预测将评估临床可用性以改善冠状动脉狭窄估计。从长远来看，我们希望心脏病专家与 DeepAngio 作为辅助工具的结合将改善血管造影解释的临床准确性。