Improving Magnetic Resonance Imaging Technologies for the Study of Brain Aging

我们测量和评估大脑老化生理的能力依赖于成像技术，如磁共振成像(MRI)。磁共振成像的改进继续显示出对一系列广泛应用的新用途，改进我们收集和解释来自成像技术的数据的方式将对加拿大和世界其他地区产生非常大的社会经济影响。与其他神经成像方法(如正电子发射断层扫描(PET))相比，MRI具有更实惠和更容易获得的优势，允许以数百种对比模式进行大脑成像，并且对研究参与者的侵入性较小。我的研究计划的长期目标是为收集大型联盟数据库做出贡献，这些数据库用于研究大脑老化的新序列，并使用这些数据库提取新的发现以及与高级数据分析和机器学习的关联。短期内，我的目标是开发一种MRI序列，在流动幻影和人类参与者中得到验证，以成像血脑屏障对水的渗透性；并开发特定的机器学习模型，以丰富联盟数据库；以及模型，分析从几个大型寿命联盟数据库收集的一系列成像数据。该领域的最新发展表明，使用先进的核磁共振序列方法，能够测量通过血脑屏障的水的流量。然而，这些新兴的方法需要更强大，并在老龄化人口中发挥作用。负责促进水通过血脑屏障的分子被称为水通道蛋白，越来越多的证据表明，水通道蛋白的减少可能会限制淀粉样蛋白等神经废物的清除，淀粉样蛋白是一种与大脑加速衰老有关的分子。我的程序利用了许多大型联盟的数据库。在我的计划中，我详细介绍了两个大数据和机器学习目标：1)建立一个图像转换模型，可以从成本更低、侵入性更小的结构MRI中预测位置发射断层扫描图像，以及2)使用多对比MRI对大脑老化过程进行高级建模。这项工作用新的信息扩展了现有的队列数据库，并研究了提取大数据的方法。该项目将取得突破性的技术进步，并带来帮助解决大脑老龄化的技术。大脑老龄化对加拿大和世界其他地区来说是一个极其负面的社会经济问题，估计每年的成本将远远超过1万亿美元。我们在图像翻译工作方面的技术进步可以产生价值1.5亿美元的信息。该项目将培训高素质的人员，掌握先进的方法，如磁共振脉冲测序、图像处理、生物物理建模、大数据处理、高性能计算和机器学习。

Our ability to measure and assess brain aging physiology relies upon imaging technologies, such as Magnetic Resonance Imaging (MRI). MRI improvements continue to show new utility for a broad set of applications, and improving how we gather and interpret data from imaging technologies is positioned to have a very large socio-economic impact for Canada and the rest of the world. Neuroimaging comes in several broad modalities that range in cost and imaging specificity, relative to other neuroimaging approaches, such as positron emission tomography (PET), MRI has the advantage of being more affordable and accessible, allowing imaging of the brain with hundreds of contrast modes and is less invasive on the research participant. The long-term goals of my research program are to contribute to gather large consortium databases used to study brain aging with novel sequences, and use these databases to distill new findings and associations with advanced data analytics and machine learning. In the short term, my objective is to develop a MRI sequence, validated in flow phantoms and human participants, to image the blood brain barrier permeability to water; and to develop specific machine learning models to enrich consortium databases; and models to analyze a range of imaging data collected from several large lifespan consortium databases. Recent developments in the field have shown the preliminary evidence of the ability to measure the flow of water across the blood brain barrier, using advanced MRI sequence methodology. However, these emerging methods need to be more robust and work in aging populations. The molecule responsible for the facilitating the passage of water through the blood brain barrier is called aquaporin, and evidence is mounting that reduced aquaporin could limit the clearing of neural waste products like amyloid, which is a molecule associated with accelerated brain aging. My program takes advantage of many large consortium databases. In my program I detail two big data and machine learning aims: 1) to build an image translation model that can predict position emission tomography images from less costly and less invasive structural MRI, and 2) to conduct advanced modelling of the brain aging process using multi-contrast MRI. This work expands existing cohort databases with new information and investigates methodology for distilling large data. The program will have ground-breaking technological advances and lead to technology to help address brain aging, which has an extremely negative socio-economic problem to Canada and the rest of the world, with a cost estimated to be well over a trillion dollars per year. Our technology advances for image translation work could yield information worth $150M. This program will train highly qualified personnel in advanced methods such as MR pulse sequencing, image processing, biophysical modelling, big data handling, high performance computing, and machine learning.