Volumetric and morphological analysis of the memory circuit in healthy ageing

健康衰老过程中记忆回路的体积和形态分析

• 批准号: RGPIN-2014-04034
• 负责人: Chakravarty, Mallar
• 金额: $ 2.26万
• 依托单位: McGill University
• 依托单位国家: 加拿大
• 项目类别: Discovery Grants Program - Individual
• 财政年份: 2014
• 资助国家: 加拿大
• 起止时间: 2014-01-01 至 2015-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=566253
• 关键词: Volumetric; morphological; analysis; memory; circuit

The hippocampus is located in the medial temporal of the brain and has long been associated with episodic memory in humans and has long been associated with several neuropsychiatric disorders such as Alzheimer’s disease and schizophrenia. Further, there is a well-known relationship between healthy ageing and the overall decline in volume of the hippocampus. In modern magnetic resonance imaging (MRI) studies, the hippocampus is typically treated as a single homogeneous structure with uniform function. However, the hippocampus is truly the combination of several intricately patterned subfields, with complex three-dimensional geometry and inter-spatial relationships. Further, there are very few analyses of the entire memory network that include the hippocampus, entorhinal cortex, the fornix, and mamillary bodies. Further, although many groups are using these software packages as a means to assess hippocampal subfield volume, there is little to no knowledge in the literature regarding the trajectories of hippocampal subfield volumes or the neuroanatomy of the memory network through the course of healthy ageing. The goal of this research proposal is to map the volumetric and morphometric trajectories of the hippocampus in males and females through the course of healthy ageing using a robust MRI image acquisition and segmentation pipeline. My group has recently published a detailed MRI acquisition and manual segmentation protocol that allows for T1- and T2-weighted high-resolution scanning (300 micron isotropic voxel dimensions) of the entire brain on a 3T MRI scanner (available to most research centres) that we have paired with a detailed manual segmentation protocol. These acquisitions, when paired with a detailed segmentation protocol, allows for the robust and reliable identification of the cornu ammonus (CA1), CA2/3, CA4/Dentate Gyrus, the molecular layers, and the subiculum. Despite it’s robustness, the manual segmentation protocol is time consuming even for the most skilled manual rater. Thus, this proposal will pair our manual segmentations with a sophisticated image processing pipeline, developed in my group, that uses a small number of manually labeled imaged as inputs and then bootstraps the segmentation process using the subject set being analyzed to provide a final set of accurate segmentation level, even at the level of the hippocampal subfields. The study design will be as follows. We will recruit 62 subjects between the ages of 18 and 80. High-resolution T1- and T2-weighted image volumes will be acquired for each study participant and automatically segmented using the image segmentation software developed in my group. Sexually dimorphic and age related tends in overall hippocampal volume and shape will be assessed. Shape analysis will be carried out relative to a surface-based population mean representation. Local vertex-wise inward and outward displacements will be used as the metric of interest for assessing shape using an extension of the segmentation pipeline previously described. The result of this work will be the first comprehensive analysis of the memory network through the course healthy ageing and will provide a baseline for other studies that are interested in the relationship between the subfields and cognitive abilities. As is the ethic of my research team, we propose not only to develop these imaging protocols and automatically generated segmentations, but to also make all our data publicly available (as we have done with our previous atlas work and algorithmic work; see http://imaging-genetics.camh.ca/Hippocampus and https://github.com/pipitone/MAGeTbrain respectively).

海马体位于大脑内侧颞叶，长期以来与人类的情景记忆有关，并且与阿尔茨海默病和精神分裂症等多种神经精神疾病有关。此外，健康衰老与海马体体积整体下降之间存在着众所周知的关系。在现代磁共振成像（MRI）研究中，海马体通常被视为具有统一功能的单一均匀结构。然而，海马体确实是几个错综复杂的子区域的组合，具有复杂的三维几何形状和空间关系。此外，对包括海马、内嗅皮层、穹窿和乳头体在内的整个记忆网络的分析也很少。此外，尽管许多团体正在使用这些软件包作为评估海马子区域体积的手段，但文献中几乎没有关于健康衰老过程中海马子区域体积的轨迹或记忆网络的神经解剖学的知识。本研究提案的目标是使用强大的 MRI 图像采集和分割流程来绘制男性和女性海马体在健康衰老过程中的体积和形态轨迹。我的小组最近发布了详细的 MRI 采集和手动分割协议，允许在 3T MRI 扫描仪（大多数研究中心可用）上对整个大脑进行 T1 和 T2 加权高分辨率扫描（300 微米各向同性体素尺寸），我们已将其与详细的手动分割协议配对。这些采集与详细的分割方案配合使用，可以对角部 (CA1)、CA2/3、CA4/齿状回、分子层和下托进行稳健且可靠的识别。尽管它很稳健，但即使对于最熟练的手动评分者来说，手动分段协议也很耗时。因此，该提案将我们的手动分割与我的团队开发的复杂图像处理管道配对，该管道使用少量手动标记的图像作为输入，然后使用正在分析的主题集引导分割过程，以提供最终的准确分割级别集，甚至在海马子区域的级别。研究设计如下。我们将招募 62 名年龄在 18 岁至 80 岁之间的受试者。将为每个研究参与者获取高分辨率 T1 和 T2 加权图像体积，并使用我小组开发的图像分割软件自动分割。将评估总体海马体积和形状的性别二态性和年龄相关倾向。形状分析将相对于基于表面的总体平均表示进行。局部顶点向内和向外位移将用作使用先前描述的分割管道的扩展来评估形状的感兴趣度量。这项工作的结果将是通过健康老龄化课程对记忆网络进行首次全面分析，并将为对子领域和认知能力之间关系感兴趣的其他研究提供基线。按照我的研究团队的道德准则，我们建议不仅开发这些成像协议和自动生成的分割，而且还公开我们的所有数据（正如我们之前的图谱工作和算法工作所做的那样；分别参见 http://imaging-genics.camh.ca/Hippocampus 和 https://github.com/pipitone/MAGeTbrain）。