Quantitative model-based ESUS reclassification using cardiac and cerebral vessel wall MRI

使用心脑血管壁 MRI 进行基于定量模型的 ESUS 重新分类

• 批准号: 10708032
• 负责人: Nazem Akoum
• 金额: $ 75.73万
• 依托单位: UNIVERSITY OF WASHINGTON
• 依托单位国家: 美国
• 财政年份: 2022
• 资助国家: 美国
• 起止时间: 2022-09-21 至 2027-08-31
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10708032
• 关键词: Acute; Algorithmic Analysis; Algorithms; Arrhythmia; Arteries; Artificial Intelligence; Atherosclerosis; Atrial Fibrillation; Cardiac; Categories; Cause of Death; Cephalic; Cerebral Infarction; Cerebrum; Classification; Clinical; Coagulation Process; Development; Diagnosis; Diagnostic; Diagnostic Procedure; Diagnostic tests; Etiology; Event; Functional disorder; Future; Goals; Healthcare Systems; Heart; Heart Atrium; Hospitals; Image; Image Analysis; Imaging Techniques; Intervention Trial; Ischemic Stroke; Left; Lesion; Literature; Machine Learning; Magnetic Resonance Imaging; Medicine; Modeling; Pathology; Patients; Pattern; Prevention Measures; Prospective Studies; Protocols documentation; Public Health; Randomized; Recurrence; Risk; Scheme; Secondary Prevention; Source; Statistical Models; Stenosis; Stroke; Subgroup; Techniques; Testing; Training; United States; Universities; Validation; Washington; cardiac magnetic resonance imaging; clinical imaging; clinical practice; deep learning; disability; embolic stroke; follow-up; heart rhythm; imaging modality; improved; individualized medicine; novel; optimal treatments; personalized medicine; predictive modeling; risk minimization; standard of care; stroke patient; targeted treatment; treatment strategy; treatment trial

Quantitative model-based ESUS reclassification using cardiac and cerebral vessel wall MRI Stroke is a major cause of death and the leading cause of permanent disability worldwide. Ischemic stroke is the dominant stroke variety, representing approximately 80+% of strokes in the United States. Defining the specific underlying pathophysiology of ischemic strokes is critical for personalized secondary prevention treatments with the goal of minimizing the risk of recurrent events. However, even with extensive diagnostic workup in current clinical practice, a large portion of ischemic strokes are classified as embolic stroke of undetermined source (ESUS), leaving these patients without optimal treatment tailored to their specific pathophysiology. Recent literature has demonstrated that among subjects diagnosed with ESUS, there may be under-detected lesions of atherosclerosis in intra/extracranial arteries or cardiac pathology on a path towards atrial fibrillation, a so called “atrial cardiopathy”. This implies that there are opportunities to improve the sensitivity and accuracy of etiologic diagnosis to reduce ischemic strokes classified into the ESUS category, allowing for more targeted, personalized secondary prevention measures. New developments in magnetic resonance imaging (MRI) of intra/extracranial atherosclerosis and atrial cardiopathy may provide new opportunities to detect these currently under-detected lesions and allow reclassification of ESUS patients into large-artery atherosclerosis or cardioembolic categories leading to focused treatment strategies. However, there are still significant challenges to using these imaging methods in practice: 1) Specialized vessel wall and cardiac MRI (ESUS-imaging) and image analysis algorithms need to be integrated into the standard of care workflow of stroke patients; 2) A model-based analysis will be needed that combines new findings from ESUS-imaging and findings from existing clinical workup so that new “risk features (RFs)” can be defined for reclassification; and 3) The impact of using these RFs on stroke subtype reclassification needs to be studied prospectively. In this proposal, we plan to develop a model-based analysis focused on ESUS-imaging and test the hypothesis that among acute ischemic stroke subjects diagnosed as ESUS under current clinical workup, a new set of RFs drawn from ESUS-imaging will allow reclassification of a subset of ESUS into large-artery atherosclerosis or cardioembolic categories. The specific aims will: 1) establish new vessel wall and cardiac MRI (ESUS-imaging) and image analysis techniques; 2) develop a multiparametric statistical model that combines information from the standard stroke workup and new ESUS-imaging to identify a set of RFs that can reclassify ischemic stroke etiology; and 3) evaluate the impact of the model on ischemic stroke subtype re-classification. If successful, this proposal will help to establish a clinical workflow that includes ESUS-imaging in ischemic stroke workup and provide a model-based algorithm to assist in future stroke subtype classification.

使用心脑血管壁 MRI 进行基于定量模型的 ESUS 重新分类 中风是全世界死亡的主要原因，也是导致永久性残疾的主要原因。缺血性中风是 占主导地位的中风类型，约占美国中风的 80% 以上。定义具体 缺血性中风的潜在病理生理学对于个性化二级预防治疗至关重要 最大限度地减少事件重复发生的风险的目标。然而，即使在当前的情况下进行了广泛的诊断检查 临床实践中，很大一部分缺血性中风被归类为来源不明的栓塞性中风 （ESUS），使这些患者无法获得适合其特定病理生理学的最佳治疗。最近的 文献表明，在诊断为 ESUS 的受试者中，可能存在未被发现的病变 颅内/颅外动脉的动脉粥样硬化或导致心房颤动的心脏病理学，即所谓的心房颤动 “心房性心脏病”。这意味着有机会提高病因学的敏感性和准确性 诊断以减少分类为 ESUS 类别的缺血性中风，从而实现更有针对性、个性化的诊断 二级预防措施。颅内/颅外磁共振成像（MRI）新进展 动脉粥样硬化和心房性心脏病可能为检测目前检测不足的这些疾病提供新的机会 病变并允许将 ESUS 患者重新分类为大动脉粥样硬化或心源性栓塞类别 从而制定有针对性的治疗策略。 然而，在实践中使用这些成像方法仍然存在重大挑战：1）专用容器 壁和心脏 MRI（ESUS 成像）和图像分析算法需要集成到以下标准中： 中风患者的护理流程； 2）需要基于模型的分析，结合来自 ESUS 成像和现有临床检查的结果，以便可以定义新的“风险特征 (RF)” 重新分类； 3) 需要研究使用这些 RF 对中风亚型重新分类的影响 前瞻性地。在本提案中，我们计划开发一种基于模型的分析，重点关注 ESUS 成像和测试 假设在当前临床检查下诊断为 ESUS 的急性缺血性中风受试者中， 从 ESUS 成像中提取的新 RF 集将允许将 ESUS 的子集重新分类为大动脉 动脉粥样硬化或心源性栓塞类别。具体目标是：1）建立新的血管壁和心脏 MRI （ESUS-成像）和图像分析技术； 2）开发一个多参数统计模型，结合 来自标准中风检查和新 ESUS 成像的信息可识别一组可重新分类的 RF 缺血性中风病因学； 3）评估模型对缺血性中风亚型重新分类的影响。 如果成功，该提案将有助于建立一个临床工作流程，其中包括缺血性中风的 ESUS 成像 检查并提供基于模型的算法以协助未来的中风亚型分类。