Supercomputer Instrumentation for Biomedical Image Analysis and Simulation

描述（由申请人提供）：项目摘要：我们提出了一个高端仪器生物医学图像分析超级计算机（BIAS）在北卡罗来纳州的几个跨学科的中心和研究所，具有强大的研究计划耦合到生物医学研究成像的重要组成部分。这是为了使实验科学家能够将他们的测量直接耦合到实时图像分析超级计算机，这种操作模式比标准批处理超级计算机提供了显着的性能增强。大多数建模和仿真计算主要是由单独的小组使用先前收集的数据完成的。通过进行实时图像分析，将在数据收集时间范围内进行重要的额外测量，与现在使用的效率较低的收集然后建模模式相比，这将导致新的发现。我们的目标是为生物医学实验学家和理论家提供这种高性能超级计算机的交钥匙操作。这将通过以下方式实现：（1）直接与来自现有中心的工作人员（他们本身是程序员）合作，以提供针对他们的合作者的需要定制的应用程序，并通过中心自己的接口访问系统，（2）提供操作员以在系统上安装和配置其他模拟和分析工具，并教导用户如何在他们的数据集上运行它们，以及（3）为其他NIH资助的研究人员提供上述应用程序的基于Web的CGI包装器接口。该系统提供了足够的计算资源，以实现模拟与运行实验的实时耦合，从而优化模型拟合，验证给定模型是否与实验结果详细匹配。相关性：该设备是专门设计用于支持一套有针对性的NIH赞助的生物医学显微镜和医学图像分析研究项目。它将用于加速和扩展现有的算法，并开发更快，更强大的算法，用于分析这些正在进行的生物实验。目标应用范围从通过细胞有丝分裂和凝血障碍（如中风）的蛋白质相互作用的单分子模拟到囊性纤维化中的肺防御和MRI图谱形成以诊断疾病。

DESCRIPTION (provided by applicant): Project Summary: We propose a High-End Instrumentation Biomedical Image Analysis Supercomputer (BIAS) at the University of North Carolina as a critical component of several interdisciplinary Centers and Institutes that have strong research programs coupled to biomedical research imaging. This is to enable experimental scientists to directly couple their measurements to a real-time image analysis supercomputer, a mode of operation that offers significant performance enhancement over standard batch-processed supercomputers. Most modeling and simulation computation is done mainly by separate groups with data previously collected. By having real-time image analysis important additional measurements will be made during the data collection time frame which will lead to new discoveries as compared to the less efficient collect and then model mode now being used. Our goal is to provide turn-key operation of this high- performance supercomputer to biomedical experimentalists and theorists. This will be accomplished by: (1) working directly with staff from existing Centers, who are themselves programmers, to provide applications customized to the needs of their collaborators and to access the system through the Centers' own interfaces, (2) providing an operator to install and configure other simulation and analysis tools on the system and teach users how to run them on their data sets, and (3) providing web-based CGI wrapper interfaces to the above applications for other NIH-sponsored researchers. The system provides sufficient computational resources to enable real-time coupling of simulations to running experiments to optimize model fits, validating whether a given model matches experimental results in detail. Relevance: The equipment is being specifically designed to support a set of targeted NIH-sponsored research projects in biomedical microscopy and medical image analysis. It will be used to accelerate and extend existing algorithms and develop faster and more robust algorithms for the analysis of these ongoing biological experiments. Target applications range from single-molecule simulation of protein interactions through cellular mitosis and clotting disorders such as stroke to lung defense in Cystic Fibrosis and MRI atlas formation to diagnose disease.