User-ready tools and scalable workflows for INCLUDE datasets in the cloud: advancing brain imaging data management and analytics

用于云中 INCLUDE 数据集的用户就绪工具和可扩展工作流程：推进脑成像数据管理和分析

• 批准号: 10406678
• 负责人: Robert J Carroll
• 金额: $ 27.36万
• 依托单位: CHILDREN'S HOSP OF PHILADELPHIA
• 依托单位国家: 美国
• 财政年份: 2020
• 资助国家: 美国
• 起止时间: 2020-09-26 至 2025-08-31
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10406678
• 关键词: Address; Algorithmic Analysis; Alzheimer' s disease brain; Alzheimer’s disease biomarker; Architecture; Award; Bioinformatics; Brain Pathology; Brain imaging; Clinical; Clinical Data; Clinical Research; Clinical Trials; Cloud Computing; Collaborations; Communities; Computer software; Computing Methodologies; Data; Data Analyses; Data Analytics; Data Reporting; Data Set; Development; Diffusion; Down Syndrome; Ecosystem; Environment; Goals; Grant; Image; Image Analysis; Infrastructure; Institution; Label; Libraries; Magnetic Resonance Imaging; Malignant neoplasm of brain; Medical; Methods; Modality; Molecular; Multi-Institutional Clinical Trial; Neurosciences; Occupations; Ontology; Operating System; Perfusion; Process; Provider; Research; Research Methodology; Research Personnel; Resources; Scanning; Semantics; Spin Labels; Standardization; Structure; System; United States National Institutes of Health; Work; bioimaging; cloud based; computer science; computerized data processing; data handling; data harmonization; data hosting; data interoperability; data management; data pipeline; data repository; data sharing; data standards; flexibility; imaging modality; imaging platform; imaging system; indexing; interoperability; large scale data; member; molecular imaging; multimodal data; neuroimaging; open source; privacy protection; repository; software development; tool; usability; Address; Algorithmic Analysis; Alzheimer' s disease brain; Alzheimer’s disease biomarker; Architecture; Award; Bioinformatics; Brain Pathology; Brain imaging; Clinical; Clinical Data; Clinical Research; Clinical Trials; Cloud Computing; Collaborations; Communities; Computer software; Computing Methodologies; Data; Data Analyses; Data Analytics; Data Reporting; Data Set; Development; Diffusion; Down Syndrome; Ecosystem; Environment; Goals; Grant; Image; Image Analysis; Infrastructure; Institution; Label; Libraries; Magnetic Resonance Imaging; Malignant neoplasm of brain; Medical; Methods; Modality; Molecular; Multi-Institutional Clinical Trial; Neurosciences; Occupations; Ontology; Operating System; Perfusion; Process; Provider; Research; Research Methodology; Research Personnel; Resources; Scanning; Semantics; Spin Labels; Standardization; Structure; System; United States National Institutes of Health; Work; bioimaging; cloud based; computer science; computerized data processing; data handling; data harmonization; data hosting; data interoperability; data management; data pipeline; data repository; data sharing; data standards; flexibility; imaging modality; imaging platform; imaging system; indexing; interoperability; large scale data; member; molecular imaging; multimodal data; neuroimaging; open source; privacy protection; repository; software development; tool; usability

PROJECT SUMMARY With rapidly growing medical datasets and data repositories, there is a critical need for technical infrastructures to make data sharing and usability feasible. Cloud ecosystems provide computationally efficient resources for large-scale data handling and analytics. While cloud-based platforms have been developed for molecular and clinical data processing, there is no single end-to-end platform for processing biomedical imaging data. Moreover, there is no standardized system to enable interoperable work between molecular, clinical, and imaging data pipelines. We propose to (Aim 1) integrate existing open-source neuroimaging software into a centralized, cloud-based platform for brain imaging data (Flywheel). This will provide scalable, automated, and publicly accessible tools to process and analyze data from standard imaging modalities, and supply streamlined pipelines for standardization and privacy-protection (de-identification) for submission to long-term repositories. In addition, we will (Aim 2) establish data representation, ontology, and workflows to allow for interoperability between currently incompatible molecular and imaging systems (CAVATICA and Flywheel). This will allow the use of rich multi-modal datasets in scientific research on brain pathology and development. The present proposal leverages collaboration between inter-disciplinary teams of software developers and biomedical analysts and researchers, and their complementary expertise in computer science, cloud computing, and neuroimaging and molecular research. Overall, this project will provide the infrastructure and methods to support large-scale data hosting and management for the neuroscience and bioinformatics communities, which will support scientific efforts including multi-institutional research and clinical trials.

项目摘要 凭借快速增长的医疗数据集和数据存储库，对技术基础架构的迫切需要 使数据共享和可用性可行。云生态系统为计算有效的资源 大规模数据处理和分析。虽然已经开发了基于云的平台用于分子和 临床数据处理，没有用于处理生物医学成像数据的单一端到端平台。 此外，没有标准化系统可以在分子，临床和 成像数据管道。我们建议（AIM 1）将现有的开源神经成像软件整合到一个 集中式，基于云的平台用于大脑成像数据（飞轮）。这将提供可扩展，自动化和 公共可访问的工具来处理和分析来自标准成像方式的数据，并简化了供应 标准化和隐私保护的管道（去识别），以提交长期存储库。 此外，我们将（AIM 2）建立数据表示，本体论和工作流程以允许互操作性 目前不兼容的分子和成像系统（Cavatica和Flywheel）之间。这将允许 在脑病理学和发育科学研究中使用丰富的多模式数据集。现在 提案利用软件开发人员与生物医学的跨学科团队之间的合作 分析师和研究人员及其计算机科学，云计算和 神经影像学和分子研究。总体而言，该项目将为基础架构和方法提供 支持神经科学和生物信息学社区的大规模数据托管和管理 将支持科学工作，包括多机构研究和临床试验。