User-friendly Analysis Platform for Decentralized Multi-site Diffusion MRI Studies

用于分散式多站点扩散 MRI 研究的用户友好分析平台

• 批准号: 10724720
• 负责人: Kang Ik Cho
• 金额: $ 8.95万
• 依托单位: BRIGHAM AND WOMEN'S HOSPITAL
• 依托单位国家: 美国
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-07-06 至 2025-06-30
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10724720
• 关键词: Address; Antisocial Personality Disorder; Automatic Data Processing; Automation; Back; Bipolar Disorder; Brain; Brain Diseases; Collaborations; Computer software; Conduct Disorder; Consumption; Data; Data Collection; Data Pooling; Data Protection; Data Security; Data Set; Decentralization; Dependence; Development; Diffusion Magnetic Resonance Imaging; Digital Imaging and Communications in Medicine; Disease; Electronic Mail; Epilepsy; Etiology; Genetic; Goals; Heterogeneity; Human; Information Protection; Infrastructure; Instruction; Internet; Intervention; Intuition; Investigation; Knowledge; Link; Major Depressive Disorder; Manuals; Measures; Meta-Analysis; Methodology; Methods; Monitor; Nature; Neurologic; Operating System; Outcome; Outcome Measure; Output; Pathologic; Personally Identifiable Information; Post-Traumatic Stress Disorders; Process; Quality Control; Reporting; Reproducibility; Research; Resources; Running; Sample Size; Schizophrenia; Scientist; Seminal; Site; Sleep; Sleeplessness; Software Tools; Statistical Data Interpretation; Structure; System; Techniques; Time; Visualization; autism spectrum disorder; automated analysis; clinical center; clinical research site; computer infrastructure; computer science; computerized data processing; dashboard; data collection site; data quality; design; experience; image processing; improved; in vivo imaging; innovation; interest; nervous system disorder; neuroimaging; operation; response; skills; tool; user-friendly; web server; white matter; white matter change

Project Summary Diffusion MRI (dMRI) is a leading in-vivo imaging methodology for investigating subtle microstructural changes in the brain’s white matter, which often requires multi-site collaborations for data collection. Decentralized processing approaches are becoming the preferred approach for multi-site data collection because they support improved personal data protection and are scalable. Using a decentralized approach, The Enhancing Neuro Imaging Genetics through Meta Analysis (ENIGMA) studies produced seminal findings across different disorders. However, the dMRI pipeline established for decentralized ENIGMA dMRI studies is not automatic, not user friendly, and requires each site to have strong programming and dMRI expertise, which is not the case in many clinical sites. The goal of this project is to help alleviate the site-level technical burdens, which would promote decentralized studies by enabling the addition of more sites. Decentralized multi-site studies are needed for rigorous and statistically robust identification of subtle neurological changes, but the technical difficulties which make decentralized dMRI analysis highly resource- intensive, are impeding research sites from participating. The technical difficulties include the installation of multiple software tools and their software dependencies, analysis instructions that include multiple scripting steps that are hard to follow for novice users, and quality control (QC) steps that require expertise in dMRI. The central goal of this application is to develop a user-friendly dMRI analysis platform for decentralized studies which will facilitate the processing and QC by pursuing two specific aims: 1) develop an automated dMRI processing pipeline with extensive data QC steps; and 2) provide dashboard function and containerize the pipeline. Under the first aim, various neuroimaging and image processing utilities will be linked together to automatize the dMRI processing pipeline while estimating extensive QC measures, which will be curated for non-dMRI experts to help intuitively understand the data quality. For the second aim, a web-server will be developed to provide a dashboard that is used to interact with the pipeline and visualize the outputs. Also, all the components will be containerized as a single package for easier dissemination and deployment using docker. The innovative approaches proposed in this application will simplify the dMRI analysis, enable optimal user experience requiring minimal manual user-interventions in the installation and operation of the containerized pipeline, and will facilitate a dashboard for intuitive user interaction. Our new and substantively different approach is significant because it is expected to resolve much of the technical burden impeding the collaborative investigation of brain changes in a decentralized approach, encouraging more sites to participate in multi-site dMRI studies.

项目摘要 扩散MRI（DMRI）是研究微妙的微观结构变化的领先的体内成像方法 在大脑的白质中，通常需要多站点的数据来收集数据。分散的 处理方法已成为多站点数据收集的首选方法，因为它们支持 改进了个人数据保护，可扩展。使用分散的方法，增强神经 通过元分析（Enigma）研究成像遗传学，在不同的不同 疾病。但是，为分散的谜DMRI研究建立的DMRI管道不是自动的，不是 用户友好，要求每个站点具有强大的编程和DMRI专业知识，情况并非如此 许多临床部位。该项目的目的是帮助减轻网站级的技术Burnens，这将 通过促进更多站点来促进分散研究。 需要分散的多站点研究才能进行严格且统计上强大的微妙识别 神经系统变化，但是使分散DMRI分析的技术困难高度资源 - 密集型，阻碍了研究站点。技术困难包括安装 多个软件工具及其软件依赖性，包括多个脚本步骤的分析说明 对于喷嘴用户而言，很难遵循，以及需要DMRI专业知识的质量控制（QC）步骤。 该应用程序的核心目标是为分散研究开发一个用户友好的DMRI分析平台 这将通过追求两个具体目标来促进处理和质量控制：1）开发自动化的DMRI 通过广泛的数据QC步骤处理管道； 2）提供仪表板功能并容器化 管道。在第一个目标下，各种神经影像学和图像处理公用事业将将 自动化DMRI处理管道，同时估算广泛的QC测量结果，该测量将策划 非DMRI专家，可以帮助直观地了解数据质量。对于第二个目标，网络服务器将是 开发的是提供用于与管道交互并可视化输出的仪表板。另外，所有人 组件将作为单个软件包将容器化，以便于使用Docker进行更轻松的传播和部署。 本应用程序中提出的创新方法将简化DMRI分析，启用最佳用户 在容器的安装和操作中需要最小手动用户干扰的经验 管道，并将促进仪表板进行直观的用户交互。我们的新方法和实质性不同 之所以重要，是因为预计它将解决大部分的技术Burnen阻碍了协作 研究分散方法中大脑变化的研究，鼓励更多的站点参与多站点 DMRI研究。