RESEARCH IN VLSI SYSTEMS FOR BIOMEDICAL APPLICATIONS

目标是发展设计和建造专门的 生物医学计算系统，利用商业上可获得的和 定制设计超大规模(VLSI)集成电路，并制造 该技术作为华盛顿大学资源的一部分提供给 生物医学计算。 四个具体目标是：1)获得VLSI设计工具和技术 并将它们扩充和集成到一个设计辅助系统中 适用于生物医学应用；2)在设计中开展研究 以及使用在VLSI中实现的异步模块，作为一种便利的手段 实验系统的建立和修改；3)研究使用 在前景看好的应用领域中通过与 华盛顿大学的生物医学研究人员；以及4)识别和 通过以下方式消除轻松广泛使用我们的新功能的障碍 生物医学科学家。 一个困难但非常重要的目标是使 VLSI技术有可能分为早期和中期，以及后期 需要大量资金的重要生物医学研究项目的阶段 可访问的计算。目前这类项目的例子包括 改进的PET图像的最大似然估计 (正电子发射断层扫描)仪器和高精度计算 放射治疗计划中的剂量。实现这一目标 需要快速指定、设计和制造专用芯片 和系统，并非常仔细地注意专业的整合 将处理器转化为灵活且快速发展的研究计算系统。 到目前为止，我们的工作在促进芯片方面取得了很好的进展 规格和设计；当前时期的主要推动力 研究正在解决快速和灵活的系统演化问题。

The objective is to develop the capability to design and build specialized biomedical computing systems that make use of Commercially available and custom designed Very Large Scale (VLSI) integrated circuits, and to make this technology available as part of the Washington University Resource for Biomedical Computing. Four specific goals are: 1) to acquire VLSI design tools and techniques from others, and to augment and integrate them into a design aid system useful for biomedical applications; 2) to carry out research in the design and use of asynchronous modules realized in VLSI, as a means to facilitate the building and modification of experimental systems; 3) to study use of the new tools in promising application areas through collaboration with biomedical researchers at Washington University; and 4) to identify and remove the obstacles to easy and widespread use of our new capabilities by biomedical scientists. A difficult but very important objective is to make the introduction of VLSI technology possible into the early and middle, as well as the late stages of important biomedical research projects that require large amounts of accessible computation. Current examples of such projects include maximum-likelihood estimation of images from advanced PETT (Positron-Emission Tomography) instruments, and high-accuracy calculation of dosage in radiation treatment planning. Accomplishing this objective requires rapid specification, design, and fabrication of specialized chips and systems, and very careful attention to the integration of specialized processors into flexible and rapidly evolving research computing systems. Our work to date has made excellent progress in facilitating chip specification and design; the major thrust during the current period of research is solving the problems of rapid and flexible system evolution.