SOFTWARE ARCHITECTURE FOR GUIDELINE DIRECTED THERAPY

指南指导治疗的软件架构

• 批准号: 2897373
• 负责人: Mark A Musen
• 金额: $ 54.69万
• 依托单位: STANFORD UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 1995
• 资助国家: 美国
• 起止时间: 1995-04-01 至 2001-06-30
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/2897373
• 关键词: AIDS; computer assisted medical decision making; computer program /software; computer system design /evaluation; human data; hypertension; oncology; problem solving

EON is a modern, component-based architecture that developers can use to build robust decision-support systems that reason about guideline- directed care. The system comprises a number of modules, including (1) problem solvers that plan medical therapy and determine whether patients are eligible for particular protocols, (2) a database mediator that handles all queries posed by the problem solvers to return point patient data or temporal abstractions of those data in a transparent fashion, and (3) a shared knowledge based of clinical guidelines and basic medical knowledge. EON is not a standalone program, but rather constitutes middleware that is intended to be embedded within a given clinical information system. In this competing renewal application, we propose to develop and evaluate new computational methods to improve the EON architecture in a variety of ways to make the components more useful, explainable, and maintainable. Our specific aims include (1) creation of a set of interactive interfaces that will allow clinicians to visualize the behavior of the EON components and the data that they process, (2) extension of the capabilities of the EON database mediator to answer more complex time-dependent queries, (3) reformulation of the knowledge bases on which the EON components operate as CORBA-compliant servers, and (4) extensions to the problem-solving components that automate specific tasks related to guideline-directed therapy. In particular, we will develop CORBA-based mapping objects that allow problem solvers and knowledge bases of clinical protocols to interoperate over the Internet. We will evaluate the usability of the new EON components both within our laboratory and at another clinical site outside of Stanford. We will use EON to encode and execute protocols for breast cancer, AIDS, hypertension, and diabetes. Our goals are to develop EON into a set of middleware components that are sufficiently robust to handle the complexities of an extremely wide range of clinical protocols and guidelines, to demonstrate the effectiveness of a comprehensive decision-support system that is compatible with modern component-based software architectures, to provide better insight into the basis for EON's therapy recommendations, and to refine EON's knowledge-acquisition methodology so that clinicians can author and maintain electronic knowledge bases of protocols and guidelines more confidently and independently.

EON是一种现代的、基于组件的架构，开发人员可以使用 建立强大的决策支持系统， 定向护理。 该系统包括多个模块，包括（1） 问题解决者，他们计划医学治疗，并确定患者是否 符合特定协议，（2）数据库中介， 处理问题解决者提出的所有问题，以返回点患者 数据或以透明的方式对这些数据进行时间抽象， 和（3）基于临床指南和基本 医学知识。 EON不是一个独立的程序，而是 构成旨在嵌入给定的 临床信息系统 在这个竞争性的更新申请中，我们建议开发和 评估新的计算方法，以改善EON架构， 多种方法使组件更有用，更易于解释， 维护 我们的具体目标包括：（1）建立一套 交互式界面，将允许临床医生可视化 EON组件的行为及其处理的数据，（2） 扩展EON数据库中介器的功能， 更复杂的时间相关查询，（3）知识的重新表达 EON组件作为CORBA兼容服务器操作的基础， 以及（4）对自动化解决问题的组件的扩展 与指南指导治疗相关的特定任务。 特别是， 我们将开发基于CORBA的映射对象， 和临床协议的知识库， 互联网 我们将评估新EON组件的可用性， 在我们的实验室和斯坦福大学以外的另一个临床基地。 我们将使用EON来编码和执行乳腺癌，艾滋病， 高血压和糖尿病。 我们的目标是将EON开发成一组中间件组件， 足够强大，可以处理极其广泛的复杂性， 一系列临床方案和指南，以证明 一个全面的决策支持系统， 与现代基于组件的软件体系结构兼容， 为EON的治疗建议提供更好的依据， 并完善EON的知识获取方法， 能够编写和维护协议的电子知识库， 更加自信和独立的指导方针。