Decision Support Systems for MAG3 Renography

描述（由申请人提供）：当前的医疗保健系统要求医生掌握不断扩展的知识库，而掌握该知识库并将其应用于特定任务（例如，图像解释，看一个新的病人）正在稳步缩小。不断扩大的知识基础和不断升级的时间限制的融合带来了一个严重的问题，这将不可避免地导致医生的错误。 这个问题的解决方案需要开发新的工具和过程，以节省医生的时间，并促进应用广泛和复杂的知识库，以具体的临床问题。我们的长期目标是改善肾泌尿科患者的护理，我们建议通过开发新的工具（决策支持系统）来实现这一目标，以（1）帮助和教育医生和受训者正确执行和解释Tc-99 m巯基乙酰三甘氨酸（MAG 3）肾图，（2）处理，检查质量控制和实际解释MAG 3肾图。特别是，我们假设我们可以开发决策支持系统来解释成人MAG 3扫描以及专家小组。我们选择开发用于MAG 3肾造影的决策支持系统，因为（1）在美国每年进行的估计400，000次放射性核素肾扫描中，绝大多数由缺乏时间发展所需专业知识的放射科医生解释，（2）超过70%的研究使用MAG 3。我们建议开发两个独立的决策支持系统，使用完全不同的方法，启发式模型（基于知识的专家系统）和预测统计系统。这些系统将从3500-4000个MAG 3研究的数据库中开发，并将被设计为获取研究，从动态放射性核素肾图中生成图像和曲线，检查错误，提取相关的定量数据，然后使用这些数据来解释研究。此外，用户将能够查询这些系统，以获得证明某一具体解释合理性的支持数据和规则。启发式和预测性统计方法将阐明每种方法的相对优点，并将为今后类似的诊断和教育辅助工具的开发和应用提供一个模板。 最后，决策支持系统将允许我们计算每个参数对诊断准确性的增量贡献，删除多余的测量，最大限度地减少获取、处理和解释肾扫描所需的时间和精力，提供肾图数据的客观分析，促进解释标准的广泛传播，促进标准化解释，教导医生和受训者像专家一样胜任地执行和解释肾脏扫描，鼓励更适当的使用，最重要的是，提高MAG 3扫描解释的准确性，从而加强对肾泌尿科患者的护理。

DESCRIPTION (provided by applicant): The current health care system requires physicians to master an ever-expanding knowledge base while the hours available to master this knowledge base and apply it to specific tasks (e.g., image interpretation, seeing a new patient) are steadily shrinking. The convergence of an expanding knowledge base and escalating time constraints poses a serious problem that will inevitably lead to physician errors. Solutions to this problem require the development of new tools and processes to save physician time and facilitate the application of a broad and sophisticated knowledge base to speci5c clinical problems. Our long-term objective is to improve the care of nephrourology patients and we propose to meet this objective by developing new tools (Decision Support Systems) to (1) assist and educate physicians and trainees to appropriately perform and interpret Tc-99m mercaptoacetyltriglycine (MAG3) renography and (2) to process, check quality control and actually interpret MAG3 renograms. Specijkally, we hypothesize that we can develop Decision Support Systems to interpret adult MAG3 scans as well as a panel of experts. We have chosen to develop Decision Support Systems for MAG3 renography because (1) the vast majority of the estimated 400,000 radionuclide renal scans performed annually in the United States are interpreted by radiologists who lack the time to develop the needed expertise and (2) MAG3 is used in over 70% of studies. We propose to develop two separate Decision Support Systems using totally different approaches, a heuristic model (knowledge based expert system) and a predictive statistical system. These Systems will be developed from a database of 3500-4000 MAG3 studies and will be designed to acquire the study, generate images and curves from a dynamic radionuclide renogram, check for errors, extract the relevant quantitative data and then use these data to interpret the study. Moreover, the user will be able to query the Systems to obtain the supporting data and rules justifying a specific interpretation. The heuristic and predictive statistical approaches will clarify the relative merits of each approach and will provide a template for the development and application of similar diagnostic and educational aids in the future. Finally, the Decision Support Systems will allow us to calculate the incremental contribution each parameter makes to the diagnostic accuracy, delete superfluous measurements, minimize the time and effort required to acquire, process and interpret renal scans, provide an objective analysis of renogram data, facilitate wide dissemination of interpretative criteria, foster standardized interpretation, teach physicians and trainees to perform and interpret renal scans as competently as experts, encourage more appropriate utilization and, most of all, improve the accuracy of MAG3 scan interpretation and thereby enhance the care of nephro-urology patients.