US-French Research Proposal: Hippocampal Layers: Advanced Ccomputational Anatomy Using Very High Resolution MRI at 7 Tesla in Humans

磁共振成像（MRI）通过允许临床医生在体内可视化大脑变化，在评估大脑疾病中起着关键作用。例如，在局灶性癫痫中，它允许检测引起癫痫发作的病变，随后可以在存在耐药性癫痫的患者中进行手术治疗。这种揭开病灶的能力对于获得良好的手术结果至关重要，并且可以限制或避免使用脑内电极进行侵入性探查。然而，标准MRI技术具有有限的空间分辨率，这导致检测细微结构改变的灵敏度有限。这在海马体的情况下尤其如此，海马体是一个相对较小的大脑结构，经常涉及成人和青少年颞叶癫痫以及其他大脑疾病。事实上，海马体是由一组复杂的内部结构组成的，其典型尺寸低于常规MRI的分辨率。该项目旨在开发新的技术来成像海马体，通过结合尖端的MRI采集技术，利用在7特斯拉的超高磁场下的更高信噪比，以及先进的数学建模技术。这种新方法将在颞叶癫痫患者中进行评估。预计充分利用其高分辨率结构MR图像将允许揭示目前在常规放射学评价中未检测到的脑病变。此外，通过对海马结构提供前所未有的洞察，这项研究将有助于开发新的患者分类和新的理论基础，以指导颞叶癫痫的治疗选择。预计该方法还将提供关键信息，以促进我们对其他大脑疾病的理解，包括阿尔茨海默病和抑郁症，这是主要的公共卫生问题。最终，这将为诊断和预后的新生物标志物铺平道路，并有助于开发新的治疗方法。该项目的总体目标是基于7特斯拉的尖端MRI采集技术，为海马体内部结构的计算解剖学开发一个连贯的数学框架。该项目引入了一种新方法，通过将体积MRI数据和形状集成到单个框架中，使计算解剖学超越形态测量学。为了实现这一目标，研究人员将首先开发7特斯拉的MRI采集技术，以执行高分辨率和多对比度成像，包括新的技术开发，这将有助于在成人和青少年患者的临床环境中使用这些先进的方法。该项目的第二部分将致力于开发先进的计算技术，以模拟多对比度7特斯拉MRI。这种技术将基于最近的数学进步，允许模拟多尺度几何变形和联合收割机形状和强度信息在一个连贯的框架。最后，所开发的方法将被应用于颞叶癫痫患者的7T MRI采集。为此，将使用7特斯拉MRI对成人（美国）和青少年（法国）患者进行研究。这应该能够证明所开发的技术的实用性，以揭示病变是无法检测的常规手段。这将为局灶性癫痫患者带来重要益处。除了目前的项目，这些结果应该对海马体发挥关键作用的大脑疾病的诊断和治疗产生重要影响，包括阿尔茨海默病，抑郁症和精神分裂症。一个配套项目正在由法国国家研究机构（ANR）资助。

Magnetic resonance imaging (MRI) plays a pivotal role in the evaluation of brain disorders by allowing clinicians to visualize brain alterations in vivo. For instance in focal epilepsies, it allows to detect lesions that cause seizures, which can subsequently be treated surgically in patients who present drug-resistant epilepsy. This ability to unveil lesions is crucial to achieve favorable surgical outcome and may allow limiting or avoiding invasive explorations with intracerebral electrodes. However, standard MRI techniques have a limited spatial resolution, which results in limited sensitivity to detect subtle structural alterations. This is particularly true in the case of the hippocampus, a relatively small cerebral structure frequently involved in adult and adolescent temporal epilepsy, as well as in other brain disorders. Indeed, the hippocampus is composed of a complex set of internal structures whose typical size is below the resolution of conventional MRI. This project aims to develop new techniques to image the hippocampus, by combining cutting edge MRI acquisition techniques, taking advantage of higher signal to noise ratio at a ultra high magnetic field of 7 Tesla, with advanced mathematical modeling techniques. This new approach will be evaluated in patients with temporal lobe epilepsy. It is expected that exploiting to their full extent very high-resolution structural MR images will allow unveiling cerebral lesions currently undetected in conventional radiological evaluation. Furthermore, by providing unprecedented insight into hippocampal structures, this research will help developing new patient classification and new rationale to guide therapeutic choices in temporal lobe epilepsy. The proposed approach is also expected to provide critical information to advance our understanding of other brain disorders, including Alzheimer's disease and depression, which are major public health concerns. Ultimately, this will pave the way to new biomarkers for diagnosis and prognosis, and help developing new treatments.The overall goal of this project is to develop a coherent mathematical framework for computational anatomy of the internal structures of the hippocampus based on cutting edge MRI acquisition techniques at 7 Tesla. The project introduces a new approach to move computational anatomy beyond morphometry by integrating both volumetric MRI data and shape into a single framework. To achieve this goal, the researchers will first develop MRI acquisition techniques at 7 Tesla to perform high-resolution and multi-contrast imaging, including new technical developments that will facilitate the use of these advanced methods in clinical settings, for adults and teenagers patients. The second part of the project will be devoted to the development of advanced computational techniques to model multi-contrast 7 Tesla MRI. Such techniques will be based on recent mathematical advances allowing to model multi-scale geometric deformations and to combine shape and intensity information in a coherent framework. Finally, the developed approaches will be applied to 7T MRI acquisition of patients with temporal lobe epilepsy. To that purpose, adult (in the US) and teenagers (in France) patients will be studied with 7 Tesla MRI. This should enable to demonstrate the utility of the developed techniques to unveil lesions that are undetectable by conventional means. This should result in important benefits for patients with focal epilepsy. Beyond the present project, the results should have an important impact on the diagnosis and treatment of brain conditions in which the hippocampus plays a key role, including Alzheimer's disease, depression and schizophrenia. A companion project is being funded by the French National Research Agency (ANR).