Ultrafast Quantitative pH MRI for Acute Ischemic Stroke Patients

摘要 缺血性中风是美国发病率和死亡率的主要原因之一。急性缺血性中风的目标 中风疗法是通过使用再灌流来挽救有梗死风险但仍可存活的组织。 战略。目前的再灌注治疗受到治疗时间窗口紧和潜在风险的限制。 脑出血的症状。使用这种基于时间的方法，只有有限数量的中风患者有资格 治疗。超过标准治疗时间窗口的患者可以从以下治疗中受益 基于多模式磁共振识别；然而，精确和准确地识别可挽救的组织是 这是必要的，因为治疗的潜在有益效果必须与出血风险进行权衡。 尽管弥散/灌注MRI不匹配已被建议作为识别存在的指南 并作为溶栓的选择标记物，使用这种方法的临床试验结果 标准一直没有定论，部分原因是包括了半影区的少血症区域，这 高估了危险组织的大小。酰胺质子转移(APT)核磁共振成像在检测方面显示出良好的前景 在动物模型和人类中风患者中，这种基于酸中毒的缺血半暗带。然而，大多数 目前使用的APT成像方案不是非常实用，并且没有针对 PH效应引起的信号变化。更定量的APT-MRI通常需要更长的扫描时间 由于使用多个RF饱和频率、多次采集和长的RF饱和脉冲而产生的时间 (或脉冲串)，由于卒中发作之间的时间窗口非常小，所有这些都阻碍了临床转换 以及可能的溶栓治疗。我们的长期目标是开发一种超快pH成像技术 常规临床应用指导超急性期卒中患者不同治疗时间的再灌注治疗 窗口，以及预测急性缺血性中风后出血性转化(HT)的风险。第一 临床假说是，与动物研究类似，缺血性组织酸中毒引起的pH成像半影区 预测如果没有开始再灌注，最终的最大梗塞面积。我们的第二个临床假设是 在缺血核心出现严重的组织酸中毒与继发的可能性增加有关 太好了。我们的假设将通过三个具体目标进行验证：1)开发和优化超快定量 PH显像法；2)验证该技术并评估酸中毒的诊断准确性。 临床环境中的缺血半暗带；以及3)开发一种新的深度学习模型来预测HT 急性缺血性卒中后，并量化pH成像的敏感性和特异性。这项工作是意料之中的 以加速将APT-MRI转化为临床可行和可靠的方法。增加了pH成像 标准磁共振成像方案有望更好地显示真正的缺血半暗带，因此 改善对临床结果的预测，降低高血压的发生率。

ABSTRACT Ischemic stroke is one of the leading causes of morbidity and mortality in the U.S. The goal of acute ischemic stroke therapy is to salvage tissue that is at risk of infarction, but still viable, through the use of reperfusion strategies. Current reperfusion therapies are limited by a tight time window for treatment and by the potential risk of brain hemorrhage. Using this time-based approach, only a limited number of stroke patients are eligible for treatment. Patients who present beyond the standard treatment time windows can benefit from therapy when identified based on multimodal MRI; however, precise and accurate identification of the salvageable tissue is essential, as the potential beneficial effect of treatment must be weighed against the risk of hemorrhage. Although a diffusion/perfusion MRI mismatch has been suggested as a guide with which to identify the presence of salvageable tissues and to serve as a selection marker for thrombolysis, the results of clinical trials using this criterion have been inconclusive, in part because of the inclusion of regions of oligemia in the penumbra, which overestimates the size of the tissue at risk. Amide proton transfer (APT) MRI has shown promise in detecting such an acidosis-based ischemic penumbra in animal models and in human stroke patients. However, most currently used APT imaging protocols are not very practical and not optimized with respect to the magnitude of signal changes caused by the pH effect. More quantitative APT-MRI typically would require an even longer scan time due to the use of multiple RF saturation frequencies, multiple acquisitions, and a long RF saturation pulse (or pulse train), all of which hamper clinical translation due to the very small time-window between stroke onset and possible thrombolysis treatment. Our long-term goal is to develop an ultrafast pH imaging technique for routine clinical use to guide reperfusion therapies for hyperacute stroke patients at various therapeutic time windows, as well as predict the risk of hemorrhagic transformation (HT) following acute ischemic stroke. The first clinical hypothesis is that, similar to animal studies, the pH imaging penumbra due to ischemic tissue acidosis predicts the maximum final infarction size if no reperfusion is initiated. Our second clinical hypothesis is that the presence of severe tissue acidosis in the ischemic core is associated with an increased probability of secondary HT. Our hypotheses will be tested through three specific aims: 1) to develop and optimize an ultrafast quantitative pH imaging method; 2) to validate this technique and assess the diagnostic accuracy of the acidosis-based ischemic penumbra in a clinical setting; and 3) to develop a novel deep-learning model with which to predict HT following acute ischemic stroke, and quantify the sensitivity and specificity of pH imaging. This work is expected to accelerate the translation of APT-MRI into a clinically viable and robust method. The addition of pH imaging to the standard MRI protocol is expected to enable better visualization of the true ischemic penumbra, thus improving predictions of clinical outcome and reducing the incidence of HT.