HIGH PERFORMANCE COMPUTING (HPC): CLINICAL APPLICATION & MEDICAL IMAGE ANALYSIS

这一分提案的目的有两个：发展高性能混凝土 经过验证的分割、配准和可视化的实现 技术；开发新的图像采集、分割、 使用HPC实现注册和可视化功能 技术 NAC将采用高性能计算，以推动快速 和/或基于大图像的数据集的交互分析。 具体来说：1.我们将应用并行算法的实用程序 并使用专用超级计算硬件(MPP、SMP和 集群机器)来设计和开发许多应用程序，这些应用程序 现已在临床上常规使用。2.我们将利用最新的硬件 和软件技术的发展，使高性能计算大大降低 价格昂贵，更容易应用于医学图像分析问题。我们 将利用这些技术改进，通过开发对称 针对映像的多处理器和集群优化应用 获取、分割、配准和可视化。这些 实现将实现高性能、高度可移植和 易于传播。

The aim of this sub-proposal is twofold: to develop HPC implementations of proven segmentation, registration and visualization techniques; to develop new image acquisition, segmentation, registration and visualization capabilities using HPC as an enabling technology. High Performance Computing will be used in the NAC to drive rapid and/or interactive analysis of large image based data-sets. Specifically: 1. We will apply the utility of parallel algorithms and make use of dedicated supercomputing hardware (MPP, SMP and cluster machines) to design and develop a number of applications which are now in routine clinical use. 2. We will utilize recent hardware and software technology developments which have made HPC much less expensive and easier to apply to medical image analysis problem. We will exploit these technology improvements by developing symmetric multi-processor and cluster optimized applications for image acquisition, segmentation, registration and visualization. These implementations will achieve high performance, be highly portable and easy to disseminate.