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 中风是世界范围内死亡的主要原因和永久性残疾的主要原因。缺血性中风是 占主导地位的中风品种，占美国中风的约80+%。来限定特定 缺血性卒中的基础病理生理学对于个性化二级预防治疗至关重要， 最大限度地降低复发事件风险的目标。然而，即使在目前广泛的诊断检查中， 在临床实践中，很大一部分缺血性卒中被归类为来源不明的栓塞性卒中 （ESUS），使这些患者没有针对其特定病理生理学的最佳治疗。最近 文献已经证明，在诊断为ESUS的受试者中，可能存在未被检测到的 颅内/外动脉粥样硬化或心脏病理学导致房颤，即所谓的 心房性心脏病这意味着有机会提高病原学诊断的敏感性和准确性。 诊断，以减少缺血性中风分类到ESUS类别，允许更有针对性，个性化 二级预防措施。颅内外磁共振成像（MRI）的新进展 动脉粥样硬化和心房性心脏病可能提供新的机会来检测这些目前未被检测到的疾病， 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 从而形成集中的治疗策略。 然而，在实践中使用这些成像方法仍然存在重大挑战：1）专门的血管 壁和心脏MRI（ESUS成像）和图像分析算法需要集成到标准中， 中风患者的护理工作流程; 2）将需要基于模型的分析， 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对卒中亚型重新分类的影响 前瞻性地。在本提案中，我们计划开发一个基于模型的分析，重点是ESUS成像和测试 假设在当前临床检查诊断为ESUS的急性缺血性卒中受试者中， 从ESUS成像中提取的一组新的RF将允许将ESUS子集重新分类为大动脉 动脉粥样硬化或心源性栓塞类别。具体目标为：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， 来自标准卒中检查和新ESUS成像的信息，以识别可以重新分类的一组RF 缺血性中风病因学;和3）评估模型对缺血性中风亚型重新分类的影响。 如果成功的话，这一建议将有助于建立一个临床工作流程，包括缺血性卒中的ESUS成像 检查并提供基于模型的算法以帮助将来的中风亚型分类。

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.