Map Manager: Longitudinal image analysis with online editing and sharing.

地图管理器：纵向图像分析，在线编辑和共享。

• 批准号: 10365810
• 负责人: Robert Harry Cudmore
• 金额: $ 115.85万
• 依托单位: UNIVERSITY OF CALIFORNIA AT DAVIS
• 依托单位国家: 美国
• 财政年份: 2021
• 资助国家: 美国
• 起止时间: 2021-09-15 至 2024-09-14
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10365810
• 关键词: 3-Dimensional; 4D Imaging; Adoption; Algorithmic Software; Astronomy; Attention; Axon; BRAIN initiative; Back; Brain imaging; COVID-19 pandemic; Cells; Client; Collaborations; Communities; Computer software; Coupled; Data; Data Analyses; Data Set; Dendrites; Dendritic Spines; Development; Discipline; Documentation; Ecosystem; Encapsulated; Ensure; Environment; Event; Evolution; FAIR principles; Feedback; Fostering; Foundations; Functional Imaging; Funding; Goals; Image; Image Analysis; Individual; Internet; Intuition; Kinetics; Libraries; Light; Link; Machine Learning; Maps; Metadata; Microscope; Modeling; Modernization; Morphology; Neurons; Neurosciences; Neurosciences Research; Nomenclature; Online Systems; Parents; Physiological; Proteins; Publishing; Pythons; Recipe; Reporter; Reproducibility; Research Methodology; Research Personnel; Running; Schedule; Scientific Inquiry; Seeds; Series; Site Visit; Software Design; Software Tools; Structure; System; Three-Dimensional Image; Three-Dimensional Imaging; Time; Vertebral column; Visual; Visualization; Vocabulary; Work; analysis pipeline; automated algorithm; base; catalyst; cell type; cloud based; data mining; data modeling; data sharing; density; design; file format; flexibility; graphical user interface; high resolution imaging; image archival system; interest; interoperability; large datasets; longitudinal analysis; longitudinal design; model building; online repository; programs; serial imaging; software development; tool; two-photon; usability; virtual; web interface

The increasing availability and ease of use of confocal, two-photon, and light-sheet microscopes coupled with rapid developments in fluorescent protein reporters have made 3D and functional imaging and its analysis a central component of modern Neuroscience research. Yet, the ease of acquiring 3D and functional images is creating progressively larger datasets, prompting the need for high-throughput image analysis algorithms and software that can be both rapid and accurate. Although software to analyze single time-point images has received substantial attention, tools to analyze multiple time-point longitudinal imaging datasets is currently lacking. This lack of longitudinal image analysis tools is a major barrier to scientific inquiry with individual labs devising their own analysis strategies creating a situation where it is difficult for others to verify and reproduce this analysis. What is needed is a community agreed upon longitudinal image analysis standard that promotes sharing. Here, we propose to develop software to create and curate annotations in longitudinal imaging datasets. This software will solve a major problem by providing the needed rigor and reproducibility while making it easy for researchers to distribute their data and analysis. Making these important datasets findable, accessible, interoperable, and reusable. To achieve these goals, we propose to build intuitive web-browser and desktop graphical-user-interfaces (GUIs) that will work with cloud based data and analysis. These GUIs will be driven by a Python advanced-programming-interface (API) that is scriptable. For online editing and sharing we will work with the BRAIN funded Brain Image Library (BIL), and for interoperability with Neurodata Without Borders (NWB) and Neuroscience Data Interface. We will utilize the BRAIN Initiative NeuroMorpho.Org and Defining Our Research Methodology (DORY), to ensure our annotations of morphology, connectivity, and physiological signatures include accepted meta-data nomenclatures and vocabularies. We will work closely with a group of "seed" BRAIN funded labs to obtain feedback and make rapid improvements in the functionality and usability of the front-end GUIs and the back-end API. This will be achieved by online forums, site visits, and a hack-a-thon hosted at UC Davis. During the Covid pandemic we have learned that these events work extremely well when done virtually and are prepared to continue this model. We are committed to providing thorough documentation for the web-browser, desktop GUIs, and Python API as well as constantly refined and simple to follow recipes with interactive web-based use cases. To ensure community adoption and use, this proposal also includes working with a number of "seed" labs to run their data through the entire pipeline from analysis to online sharing. The long range goal is to have Map Manager act as a catalyst for data analysis, exploration, and sharing. Effectively creating a community based approach, akin to other disciplines such as astronomy, where data is widely and publicly shared allowing effective data mining and model building to advance new discoveries.

共焦、双光子和光片显微镜的可用性和易用性不断提高， 荧光蛋白报告基因的快速发展使得 3D 和功能成像及其分析成为可能 现代神经科学研究的核心组成部分。然而，获取 3D 和功能图像的便捷性是 创建越来越大的数据集，促使需要高通量图像分析算法和 既快速又准确的软件。尽管分析单时间点图像的软件已经获得 由于缺乏足够的关注，目前缺乏分析多时间点纵向成像数据集的工具。这 缺乏纵向图像分析工具是各个实验室设计其科学探究的主要障碍 自己的分析策略造成了其他人难以验证和重现此分析的情况。 我们需要的是一个社区商定的纵向图像分析标准，以促进共享。 在这里，我们建议开发软件来创建和管理纵向成像数据集中的注释。 该软件将通过提供所需的严谨性和可重复性来解决主要问题，同时使其变得简单 供研究人员分发他们的数据和分析。使这些重要的数据集可查找、可访问、 可互操作、可重用。为了实现这些目标，我们建议构建直观的网络浏览器和桌面 将与基于云的数据和分析一起使用的图形用户界面（GUI）。这些 GUI 将被驱动 通过可编写脚本的 Python 高级编程接口 (API)。对于在线编辑和共享，我们将 与 BRAIN 资助的脑图像库 (BIL) 合作，并实现与 Neurodata Without 的互操作性 边界 (NWB) 和神经科学数据接口。我们将利用 BRAIN Initiative NeuroMorpho.Org 定义我们的研究方法（DORY），以确保我们对形态、连接性、 生理特征包括公认的元数据术语和词汇。 我们将与一组由 BRAIN 资助的“种子”实验室密切合作，获取反馈并快速做出 前端 GUI 和后端 API 的功能和可用性的改进。这将是 这是通过在线论坛、网站访问和加州大学戴维斯分校举办的黑客马拉松来实现的。新冠疫情期间我们 我们了解到这些活动在虚拟环境下效果非常好，并准备继续这样做 模型。我们致力于为 Web 浏览器、桌面 GUI 和 Python API 以及不断完善且易于遵循的配方以及基于 Web 的交互式用例。到 确保社区采用和使用，该提案还包括与一些“种子”实验室合作来运行 他们的数据贯穿从分析到在线共享的整个流程。 长期目标是让地图管理器充当数据分析、探索和共享的催化剂。 有效地创建基于社区的方法，类似于天文学等其他学科，其中数据 广泛和公开共享允许有效的数据挖掘和模型构建以推进新发现。