Nobrainer: A robust and validated neural network tool suite for imagers

Nobrainer：适用于成像仪的强大且经过验证的神经网络工具套件

• 批准号: 10021957
• 负责人: Satrajit Sujit Ghosh
• 金额: $ 241.99万
• 依托单位: MASSACHUSETTS INSTITUTE OF TECHNOLOGY
• 依托单位国家: 美国
• 财政年份: 2020
• 资助国家: 美国
• 起止时间: 2020-08-01 至 2024-07-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10021957
• 关键词: Address; Adolescent; Architecture; Archives; Artificial Intelligence; BRAIN initiative; Bayesian neural network; Brain; Brain imaging; Brain scan; Cognitive; Collaborations; Collection; Communities; Computer software; Consumption; Data; Data Set; Data Sources; Development; Diagnosis; Disease; Educational workshop; Engineering; Ethics; Evaluation; Failure; Human; Image; Individual; Informatics; Institution; Label; Learning; Legal; Libraries; Longevity; Memory; Mental Health; Methods; Modeling; Nervous system structure; Network-based; Neural Network Simulation; Output; Pattern; Population; Population Heterogeneity; Privacy; Privatization; Process; Psychological Transfer; Pythons; Quality Control; Research; Research Personnel; Sampling; Scientist; Series; Software Engineering; Software Framework; Structure; Students; Technology; Testing; Three-Dimensional Imaging; Time; Training; Training Support; Trust; Uncertainty; Visualization; Work; algorithm training; base; citizen science; cognitive development; college; computerized data processing; connectome; data reuse; deep learning; distributed data; diverse data; high school; image processing; imager; improved; insight; large datasets; learning strategy; model development; network models; neural network; neuroimaging; neurophysiology; open source; software infrastructure; statistical learning; tool; user-friendly

There is an increasing need for efficient and robust software to process, integrate, and offer insight across the diversity of population imaging efforts underway across the BRAIN Initiative and other projects. Advances in statistical learning offer a set of technologies that can address many research applications using the extensive and varied data being produced by the projects. This can transform how we analyze and integrate new data. We propose using Nobrainer, an open source Python library that leverages these new learning technologies, as a platform that greatly simplifies integrating deep learning into neuroimaging research. Using this library, we are building and distributing user-friendly and cloud enabled end-user applications for the neuroimaging community. In Aim 1, we provide neural network models. We will create robust, pre-trained neural networks for brain segmentation and time series processing using brain scans from over 65000 individuals. Once trained, these models can then be used as the basis for many other applications, especially in reducing time of processing. We will subsequently use these base networks to perform image processing, image correction, and quality control. In Aim 2, we address the ability to train on private datasets. We will use Bayesian neural network models, which support principled use of prior information. We will use these networks to help detect when the models are expected to fail on an input, and provide visualizations to better understand how the model is working. In Aim 3, we focus on the engineering needed to maintain the software infrastructure, improve efficiency, and increase the scalability of our training methods. Here, we will extend, maintain, and disseminate Nobrainer, our open source software framework, together with training materials and ready to use, cloud-friendly, applications. We will also create much faster, neural network equivalents of time consuming image processing tasks (e.g., registration, segmentation, and annotation). The Nobrainer tools developed through these aims will allow users to find and apply the most pertinent applications and developers to extend the framework to support new architectures and disseminate new models and applications. We expect these tools to be used by any neuroimaging researcher through integration with BRAIN archives and popular software packages. These tools will significantly reduce data processing and new model development time, thus allowing faster exploration of hypotheses using public data and increase reusability of data through greater trust in model outputs.

人们越来越需要高效、强大的软件来处理、集成和提供跨领域的洞察力。 BRAIN Initiative 和其他项目正在进行多样化的人口成像工作。进展 统计学习提供了一套技术，可以使用广泛的数据来解决许多研究应用 以及项目产生的各种数据。这可以改变我们分析和整合新数据的方式。我们 建议使用 Nobrainer（一个利用这些新学习技术的开源 Python 库）作为 该平台大大简化了将深度学习集成到神经影像研究中的过程。使用这个库，我们 为神经影像构建和分发用户友好且支持云的最终用户应用程序 社区。在目标 1 中，我们提供神经网络模型。我们将创建强大的、预先训练的神经网络 使用超过 65000 人的脑部扫描进行脑部分割和时间序列处理。一次 经过训练后，这些模型可以用作许多其他应用程序的基础，特别是在减少 加工。我们随后将使用这些基础网络来执行图像处理、图像校正和 质量控制。在目标 2 中，我们解决了在私有数据集上进行训练的能力。我们将使用贝叶斯神经网络 网络模型，支持先验信息的原则性使用。我们将使用这些网络来帮助检测 当预计模型在输入上失败时，并提供可视化以更好地理解 模型正在工作。在目标 3 中，我们专注于维护软件基础设施所需的工程， 提高效率，并增加我们训练方法的可扩展性。在这里，我们将扩展、维护和 传播我们的开源软件框架 Nobrainer 以及可供使用的培训材料， 云友好的应用程序。我们还将创建速度更快的、相当于耗时的神经网络 图像处理任务（例如配准、分割和注释）。开发的 Nobrainer 工具 通过这些目标将允许用户找到并应用最相关的应用程序并让开发人员扩展 支持新架构并传播新模型和应用程序的框架。我们期待这些 通过与 BRAIN 档案和流行的集成，任何神经影像研究人员都可以使用的工具 软件包。这些工具将显着减少数据处理和新模型开发时间，从而 允许使用公共数据更快地探索假设，并通过更大的范围提高数据的可重用性 信任模型输出。

项目成果

TemplateFlow: FAIR-sharing of multi-scale, multi-species brain models.

• DOI: 10.1038/s41592-022-01681-2
• 发表时间: 2022-12
• 期刊: NATURE METHODS
• 影响因子: 48
• 作者: Ciric, Rastko;Thompson, William H. H.;Lorenz, Romy;Goncalves, Mathias;MacNicol, Eilidh E. E.;Markiewicz, Christopher J. J.;Halchenko, Yaroslav O. O.;Ghosh, Satrajit S. S.;Gorgolewski, Krzysztof J. J.;Poldrack, Russell A. A.;Esteban, Oscar
• 通讯作者: Esteban, Oscar

Privacy-preserving quality control of neuroimaging datasets in federated environments.

• DOI: 10.1002/hbm.25788
• 发表时间: 2022-05
• 期刊: Human brain mapping
• 影响因子: 4.8
• 作者: Saha DK;Calhoun VD;Du Y;Fu Z;Kwon SM;Sarwate AD;Panta SR;Plis SM
• 通讯作者: Plis SM

NeuroCrypt: Machine Learning Over Encrypted Distributed Neuroimaging Data.

• DOI: 10.1007/s12021-021-09525-8
• 发表时间: 2022-01
• 期刊: Neuroinformatics
• 影响因子: 3
• 作者: Senanayake N;Podschwadt R;Takabi D;Calhoun VD;Plis SM
• 通讯作者: Plis SM

GLACIER: GLASS-BOX TRANSFORMER FOR INTERPRETABLE DYNAMIC NEUROIMAGING.

Glacier：用于可解释动态神经成像的玻璃盒变压器。

• DOI: 10.1109/icassp49357.2023.10097126
• 发表时间: 2023
• 期刊: Proceedings of the ... IEEE International Conference on Acoustics, Speech, and Signal Processing. ICASSP (Conference)
• 作者: Mahmood,Usman;Fu,Zening;Calhoun,Vince;Plis,Sergey
• 通讯作者: Plis,Sergey