Computational methods for regional hippocampal morphometry in AD

描述（由申请人提供）：PI是一名定量研究人员，具有计算机科学背景，并在图像分析方法、计算机视觉和机器学习领域发表了大量出版物。他的目标是建立自己作为一个高生产力的阿尔茨海默病（AD）和痴呆症的独立研究人员，通过补充他在定量分析与痴呆症和衰老的生物医学方面的新获得的专业知识的优势。宾夕法尼亚大学功能性神经影像中心主任、国际知名的神经学家John Detre医学博士教授已提出担任主要导师。共同发起人John Trojanowski教授，医学博士，博士和Murray Grossman教授，医学博士，EdD已经提出帮助PI获得痴呆症的病理学和生理学以及其对记忆，语言和思维的影响方面的专业知识。该研究计划涉及开发一个先进的框架，用于使用体内神经成像检测和跟踪海马和海马旁回解剖子区域的结构和功能变化。使用该框架，PI将评估AD相关变化在这些颞叶结构的解剖子区域中存在差异的假设，因此，其形态学和生理学可用于早期准确预测AD。这些假设将通过挖掘AD患者，AD风险人群和老年对照的纵向MRI图像数据的大量数据库进行测试，这些数据库由NIA/NIH最近发起的ADNI倡议产生。本案无关随着婴儿潮一代的年龄增长，预计美国AD家庭的数量以及相关的社会经济负担将急剧增加。虽然没有治愈AD的方法，但正在进行相当大的努力来开发可能阻止或减缓导致AD的神经退行性过程的药剂。为了有效，这些治疗需要早期发现。Pi的职业目标，由 本申请中提出的研究是提高早期AD诊断的准确性，并开发非侵入性分析工具，这些工具将有助于药物开发并拓宽我们对AD病理学和生理学的了解。

DESCRIPTION (provided by applicant): The PI is a quantitative researcher with a background in computer science and a record of publications in the areas of image analysis methodology, computer vision and machine learning. His objective is to establish himself as a highly productive independent researcher of Alzheimer's disease (AD) and dementia by complimenting his strengths in quantitative analysis with a newly acquired expertise in the biomedical aspects of dementia and aging. Prof. John Detre MD, the head of Penn's Center for Functional Neuroimaging and an internationally renowned neurologist, has offered to serve as the principal mentor. The co-sponsors, Prof. John Trojanowski, MD, PhD and Prof. Murray Grossman, MD, EdD have offered to help the PI acquire expertise in the pathology and physiology of dementia as well as its impact on memory, language and thinking. The research plan involves developing an advanced framework for detecting and tracking structural and functional changes in the anatomical subregions of the hippocampus and parahippocampal gyrus using in vivo neuroimaging. Using this framework, the PI will evaluate the hypotheses that AD-related changes differ across the anatomical subregions of these temporal lobe structures and, consequently, that their morphology and physiology can be used to predict AD early and accurately. These hypotheses will be tested by mining the massive database of longitudinal MRI image data of AD patients, people at risk for AD and elderly controls, which is being generated by the ADNI initiative launched recently by the NIA/NIH. RELEVANCE. As the baby boom generation ages, the number of US families devastated by AD and the associated financial burden on the society is expected to increase dramatically. While no cure for AD is known, a sizable effort is underway in developing pharmaceutical agents that may halt or slow down the neurodegenerative processes that cause AD. To be effective, these treatments will require early detection. The Pi's career objectives, addressed by the research proposed in this application, are to improve the accuracy of early AD diagnosis and to develop non-invasive analytic tools that would aid drug development and broaden what we know about the pathology and physiology of AD.