Automated Plaque Detection and Classification using MRI

使用 MRI 进行自动斑块检测和分类

• 批准号: 6951380
• 负责人: RUSSELL WAYNE ANDERSON
• 金额: $ 37.5万
• 依托单位: ISCHEM CORPORATION
• 依托单位国家: 美国
• 财政年份: 2002
• 资助国家: 美国
• 起止时间: 2002-06-01 至 2007-08-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/6951380
• 关键词: artificial intelligence; atherosclerotic plaque; bioengineering /biomedical engineering; bioimaging /biomedical imaging; biomedical automation; cardiovascular disorder diagnosis; clinical research; diagnosis design /evaluation; human data; image processing; informatics; magnetic resonance imaging

DESCRIPTION (provided by applicant): Acute thrombus formation on disrupted/eroded human atherosclerotic lesions plays a critical role on the onset of acute coronary syndromes and progression of atherosclerosis. Pathological evidence has clearly established that it is plaque composition rather than stenotic severity that modulates plaque vulnerability and thrombogenicity. Therefore, the possibility of detecting and characterizing atherosclerotic lesions would have significant clinical implications. Among the different imaging modalities, MRI seems to be the most promising in terms of discrimination and due to its non-invasive nature. We propose to develop an automated MRI image analysis system for detecting, measuring, and classifying atherosclerotic plaques. The proposed project will be conducted by a highly experienced team of experts in signal processing, pattern recognition, statistics, and product development in close collaboration with key cardiovascular researchers, with specific expertise in MRI imaging and atherosclerosis. Automation would establish a fast, objective (observer-independent), and standard diagnostic measure of plaque burden, allowing for comparison of results between laboratories, throughout longitudinal studies, and across different imaging equipment. In a clinical setting, this system would greatly reduce the diagnostic costs involved in measuring the degree of stenosis and detecting thrombosis-prone plaques.

描述（由申请人提供）： 破坏/侵蚀的人类动脉粥样硬化病变上的急性血栓形成对于急性冠状动脉综合征的发作和动脉粥样硬化的进展起着关键作用。病理学证据明确表明，调节斑块脆弱性和血栓形成的是斑块组成，而不是狭窄严重程度。因此，检测和表征动脉粥样硬化病变的可能性将具有重要的临床意义。在不同的成像方式中，MRI 似乎在区分度和非侵入性方面最有前途。我们建议开发一种自动 MRI 图像分析系统，用于检测、测量和分类动脉粥样硬化斑块。拟议的项目将由信号处理、模式识别、统计和产品开发方面经验丰富的专家团队与主要心血管研究人员密切合作进行，这些研究人员在 MRI 成像和动脉粥样硬化方面具有特定的专业知识。自动化将建立一种快速、客观（独立于观察者）和标准的斑块负荷诊断测量方法，允许在实验室之间、整个纵向研究和不同成像设备之间比较结果。在临床环境中，该系统将大大降低测量狭窄程度和检测易形成血栓的斑块所涉及的诊断成本。