ULTRA-HIGH RESOLUTION MODELLING OF THE HUMAN BRAIN USING MULTISCALE, MULTIMODAL IMAGING AND AI

1. 研究背景包括潜在影响伦敦大学学院的研究人员开发了一种新的x射线成像方法，称为分层相衬断层扫描（HiP-CT），可以对整个人体器官进行详细成像，直至细胞水平。HiP-CT是一种x射线相位传播技术，它使用欧洲同步辐射设施（ESRF）极光源，在整个人体器官的任何位置进行非破坏性的三维（3D）扫描，其分辨率逐级增加，在感兴趣的体积中低至1微米/体素。因此，这些成像数据不仅提供了人类大脑复杂层次结构的新视角，而且还可以提供对难以诊断的神经退行性疾病中起作用的生物学过程的详细了解。该博士将开发先进的人工智能工具，用于数据处理，有意义的解释和分析，并将这些高分辨率图像转化为临床应用。目的和目标本项目的目的是增强我们对健康个体和神经退行性疾病患者大脑白质束和血管结构的理解。这将通过开发先进的计算技术，如大规模人工智能，来分析从同步加速器获得的多分辨率图像，然后将这些图像与MRI图像相关联，以实现在评估神经退行性疾病（如阿尔茨海默病）中的临床应用。研究方法的新颖性HiP-CT成像方法使人脑成像的分辨率比目前最高分辨率的MRI扫描高300倍，因此需要开发定制的计算方法来模拟大脑结构。这些成像数据在世界范围内是独一无二的，将与临床MRI图像相关联，以改进目前在健康和神经退行性病例中白质束定向和血管结构的模型。与EPSRC的战略和研究领域保持一致该项目与EPSRC医疗保健技术主题的优先事项保持一致，特别是围绕新型成像技术。这项研究将支持先进的人工智能工具的发展，用于数据处理，理解复杂的成像数据，理解潜在的病理和监测神经退行性疾病的进展。任何公司或合作者都参与了“陈-扎克伯格倡议”

1. Context of the research including potential impactResearchers at UCL have developed a new X-ray imaging method called Hierarchical Phase-Contrast Tomography (HiP-CT), which allows imaging whole human organs in detail up to the cellular level. HiP-CT is an X-ray phase propagation technique that uses the European Synchrotron Radiation Facility (ESRF) Extremely Brilliant Source to perform non-destructive, three-dimensional (3D) scans with hierarchically increasing resolution at any location in whole human organs, down to 1 um/voxel in Volumes of Interest. As such, these imaging data not only provide a novel view of the complex hierarchical structure of the human brain but could also provide a detailed understanding of the biological processes at work during neurodegenerative diseases which are difficult to diagnose.This PhD will develop advanced AI tools for data processing, meaningful interpretation and analysis, and translation of these high-resolution images to clinical use.2. Aims and objectivesThe goal of this project is to enhance our understanding of the structure of white matter tracts and blood vessels in the brain, both in healthy individuals and in those with neurodegenerative conditions. This will be done by developing advanced computational techniques, such as large-scale AI, to analyze multi-resolution images obtained from a synchrotron and then correlating these images with MRI images to realise clinical use in the assessment of a neurodegenerative disease such as Alzheimer's.3. Novelty of the research methodologyThe HiP-CT imaging method enables imaging of the human brain with a resolution of up to 300 times greater than the current highest resolution MRI scans and so requires bespoke computational methods to be developed for modelling brain structure. These imaging data are unique worldwide and will be correlated to clinical MRI images to improve the current models of white matter tract orientation and vascular architecture, in healthy and neurodegenerative cases.4. Alignment to EPSRC's strategies and research areasThe project aligns with the priorities of the EPSRC Healthcare Technologies Theme, specifically around Novel imaging technologies. This research will support the development of advanced AI tools for data processing, making sense of complex imaging data, understanding underlying pathology and monitoring progression of neurodegenerative diseases.5.Any companies or collaborators involvedChan Zuckerberg Initiative